AP-T336-18 Design and Performance of Foamed Bitumen Stabilised Pavements-Reduced

Technical Report
AP-T336-18
Design and Performance of Foamed

Bitumen Stabilised Pavements
Design and Performance of Foamed Bitumen Stabilised Pavements
Prepared by Publisher
Geoff Jameson Austroads Ltd.

Level 9, 287 Elizabeth Street
Sydney NSW 2000 Australia
Project Manager Phone: +61 2 8265 3300
[email protected]
Richard Yeo
www.austroads.com.au
Abstract About Austroads

There is a need to improve the Austroads procedures for mix design Austroads is the peak organisation of Australasian road
and structural design of foamed bitumen stabilised (FBS) materials. transport and traffic agencies.
This final report summarises the mix design test methods developed
within the context of an Austroads mix design process. The report Austroads’ purpose is to support our member organisations to
details the construction and performance monitoring of six trial deliver an improved Australasian road transport network. To
sections. It was observed that cracking was the predominant distress succeed in this task, we undertake leading-edge road and
type and that pavement thickness design needs to consider FBS transport research which underpins our input to policy
fatigue. The observed fatigue lives were reasonably consistent with development and published guidance on the design,
the predictions of the Austroads interim method. Generally, rutting construction and management of the road network and its
and shape loss are not dominating performance, even when the host associated infrastructure.
material stabilised includes 20% recycled asphalt. Austroads provides a collective approach that delivers value
for money, encourages shared knowledge and drives
consistency for road users.
Keywords Austroads is governed by a Board consisting of senior

executive representatives from each of its eleven member
Pavement design, rehabilitation, foamed bitumen, stabilisation, organisations:
recycling, mix design, construction, field performance, validation
• Roads and Maritime Services New South Wales
• Roads Corporation Victoria
ISBN 978-1-925671-64-3 • Queensland Department of Transport and Main Roads
Austroads Project No. TT1825 • Main Roads Western Australia
• Department of Planning, Transport and Infrastructure
Austroads Publication No. AP-T336-18
South Australia
Publication date July 2018 • Department of State Growth Tasmania
Pages 303 • Department of Infrastructure, Planning and Logistics
Northern Territory
• Transport Canberra and City Services Directorate,
Australian Capital Territory
© Austroads 2018
• The Department of Infrastructure, Regional Development
This work is copyright. Apart from any use as permitted under the and Cities
Copyright Act 1968, no part may be reproduced by any process • Australian Local Government Association
without the prior written permission of Austroads.
• New Zealand Transport Agency.
This report has been prepared for Austroads as part of its work to promote improved Australian and New Zealand transport outcomes by
providing expert technical input on road and road transport issues.
Individual road agencies will determine their response to this report following consideration of their legislative or administrative
arrangements, available funding, as well as local circumstances and priorities.
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.
Summary
In July 2012 Austroads commissioned ARRB to conduct research project TT1825 Mix Design and Field
Evaluation of Foamed Bitumen Stabilised Pavements. The objectives of this project were to:
• improve the Austroads procedures for the structural design of foamed bitumen stabilised materials for
new pavements and structural rehabilitation treatments
• identify distress modes of bitumen stabilised pavements from the trial sites
• improve and harmonise national mix design procedures for bitumen stabilised materials.
This final project report:

• summarises the improved test methods developed in conjunction with road agencies and industry
• investigates the influence of compaction method, moisture content and other factors on the mix design
moduli
• compares laboratory mix design moduli with those from field cores
• describes the distress modes observed at the six trial sites
• investigates improvements to the Austroads structural design method.
The major project conclusions were:

• The mix design methods of foamed bitumen stabilised (FBS) materials have been improved and
hamonised.
• Mix design indirect tensile moduli are similar to very early-life values and well below cured field values.
• Fatigue cracking is a distress mode for FBS pavement layers and needs to be considered in structural
thickness design.
• Despite identified deficiencies, the fatigue life predictions using Austroads interim thickness design
method appear reasonably consistent with the observed field performance based on the limited
information available from field trials. The use of very conservative FBS design moduli with the asphalt
fatigue relationship, together with empirical guidance on particle size distribution, plasticity and maximum
binder contents, appears to result in FBS pavement layer depth based on the limited performance
information available.
Austroads 2018 | page i

Contents
Summary ......................................................................................................................................................... i
1. Introduction ............................................................................................................................................ 1
2. Mix Design Procedures ......................................................................................................................... 2
2.1 Introduction ....................................................................................................................................... 2
2.2 Background to Indirect Tensile Modulus Requirements .................................................................. 2
2.2.1 Pioneering Work in the 1990s ............................................................................................. 2
2.2.2 Early Development by TMR ................................................................................................ 3
2.2.3 Current TMR Test Methods and Modulus Requirements ................................................... 5
2.3 Development of Austroads Mix Design Test Methods ..................................................................... 6
2.3.1 Introduction .......................................................................................................................... 6
2.3.2 Determination of Foaming Properties.................................................................................. 7
2.3.3 Mixing of Foamed Bitumen Materials .................................................................................. 7
2.3.4 Compaction of Test Cylinders ............................................................................................. 7
2.3.5 Curing of Modulus Specimens ............................................................................................ 8
2.3.6 Resilient Modulus Measurement ......................................................................................... 9
2.4 Description of Procedure to Determine Binder Contents Framework .............................................. 9
3. Effect of Moisture Content and Lime Type on Resilient Modulus .................................................. 10
3.1 Introduction ..................................................................................................................................... 10
3.2 Influence of Moisture Content on Newell Highway Trial Mix .......................................................... 11
3.2.1 Method ............................................................................................................................... 11
3.2.2 Compactability ................................................................................................................... 11
3.2.3 Density and Modulus ......................................................................................................... 12
3.3 Influence of Moisture Content on Western Freeway Trial Mix ....................................................... 15
3.4 Influence of Moisture Content on Calder Freeway Trial Mix .......................................................... 16
3.5 Effect of Lime Type ........................................................................................................................ 17
3.6 Summary ........................................................................................................................................ 19
4. Effect of Laboratory Compaction Method on Modulus .................................................................... 20
4.1 Introduction ..................................................................................................................................... 20
4.2 Effect of Specimen Size ................................................................................................................. 21
4.2.1 Marshall Specimens .......................................................................................................... 21
4.2.2 Gyratory Specimens .......................................................................................................... 22
4.3 Compaction Method ....................................................................................................................... 23
4.3.1 Comparative Testing of Newell Highway Material ............................................................. 23
4.3.2 Comparative Testing of Calder Freeway Material ............................................................. 24
4.3.3 Additional TMR Testing ..................................................................................................... 25
4.4 Summary ........................................................................................................................................ 26
5. Comparison of Laboratory-mixed and Compacted Moduli with Field Core Moduli ...................... 27
5.1 Introduction ..................................................................................................................................... 27
5.2 Newell Highway, Bellata ................................................................................................................. 27
5.3 Calder Freeway, Woodend............................................................................................................. 29
Austroads 2018 | page i

5.4 Western Freeway, Ballan ............................................................................................................... 31

5.5 Combining the Project Results ....................................................................................................... 34
5.6 Summary ........................................................................................................................................ 34
6. Testing of FBS Mixes during Construction ....................................................................................... 36
6.1 Introduction ..................................................................................................................................... 36
6.2 Comparison of Moduli of Laboratory-mixed and Field-mixed Materials ......................................... 36
6.3 Effect of Time Delay between Mixing and Compaction ................................................................. 39
6.4 Summary ........................................................................................................................................ 41
7. Overview of Field Trial Observations of Distress ............................................................................. 42
7.1 Introduction ..................................................................................................................................... 42
7.2 Cracking ......................................................................................................................................... 43
7.3 Rutting ............................................................................................................................................ 45
7.4 Flushing/Embedment ..................................................................................................................... 45
8. Austroads Thickness Design Method ................................................................................................ 46
8.1 Introduction ..................................................................................................................................... 46
8.2 Background to the Interim Method ................................................................................................. 46
8.2.1 Early use of Mechanistic-empirical Method ....................................................................... 46
8.2.2 Laboratory Fatigue Life of Field Beams ............................................................................ 46
8.2.3 TMR Development of the Method ..................................................................................... 48
8.3 Interim Austroads Method .............................................................................................................. 49
8.3.1 Introduction ........................................................................................................................ 49
8.3.2 Method to Determine FBS Design Modulus ...................................................................... 49
8.3.3 Fatigue Relationship .......................................................................................................... 51
8.3.4 Surfacing Requirements .................................................................................................... 51
8.4 Initial Comparison of Predicted and Observed Performance ......................................................... 51
8.4.1 Introduction ........................................................................................................................ 51
8.4.2 FBS Design Moduli ............................................................................................................ 51
8.4.3 Adjusted FBS Fatigue Relationship used in the Analysis ................................................. 52
8.4.4 Predicted Allowable Traffic Loadings ................................................................................ 52
8.4.5 Comparison of Observed and Predicted Fatigue Lives..................................................... 53
8.5 Review of FBS Design Modulus ..................................................................................................... 53
8.5.1 Definition of Design Modulus ............................................................................................. 53
8.5.2 Use of Mix Design Moduli .................................................................................................. 54
8.5.3 Maximum Design Modulus ................................................................................................ 54
8.6 Fatigue Relationship....................................................................................................................... 54
8.7 Proposed Amendment to Interim Method ...................................................................................... 55
8.8 Summary ........................................................................................................................................ 55
9. Summary and Conclusions................................................................................................................. 57
9.1 Mix Design ...................................................................................................................................... 57
9.1.1 Test Methods ..................................................................................................................... 57
9.1.2 Laboratory Characterisation .............................................................................................. 57
9.1.3 Mix Design Moduli Compared to In Situ Values ................................................................ 58
9.1.4 Compliance Testing of Field-mixed Materials ................................................................... 58
9.2 Distress Types ................................................................................................................................ 59
9.3 Austroads Structural Thickness Design Method ............................................................................ 59
References ................................................................................................................................................... 61
Austroads 2018 | page ii

Appendix A Kwinana Freeway, Baldivis ............................................................................................ 65

Appendix B Kewdale Road, Welshpool .............................................................................................. 75
Appendix C Port Wakefield Road, Virginia ........................................................................................ 87
Appendix D Calder Freeway, Woodend ........................................................................................... 111
Appendix E Newell Highway, Bellata ............................................................................................... 163
Appendix F Western Freeway, Ballan .............................................................................................. 206
Appendix G Moduli Back-calculated from Measured FWD Deflections ........................................ 264
Tables
Table 2.1: TMR early minimum modulus requirements ................................................................................ 3

Table 2.2: TMR modulus requirements in 2004 ............................................................................................ 4
Table 2.3: TMR initial modulus mix design limits for foamed bitumen stabilised materials .......................... 5
Table 2.4: TMR cured modulus mix design limits for a foamed bitumen stabilised base ............................. 5
Table 2.5: TMR cured modulus mix design limits for a foamed bitumen stabilised subbase ....................... 5
Table 2.6: Austroads test methods for FBS mix design ................................................................................ 6
Table 2.7: Selecting mix components ........................................................................................................... 9
Table 2.8: Determination of design bitumen content .................................................................................... 9
Table 3.1: Variation in moisture contents used in mix designs ................................................................... 10
Table 3.2: Influence of moisture on density and modulus of Newell Highway material .............................. 13
Table 3.3: Mix design testing post-construction ......................................................................................... 16
Table 3.4: Effect of lime type on IDT moduli ............................................................................................... 18
Table 4.1: Effect of Marshall mould and hammer size ................................................................................ 21
Table 4.2: Effect of gyratory compaction specimen size............................................................................. 23
Table 4.3: Effect of compaction method on moduli of Newell Highway material ........................................ 24
Table 4.4: Effect of compaction method on modulus, Calder Freeway field-mixed material ...................... 25
Figures
Figure 3.1: Influence of moisture on the Newell Highway material compactability ...................................... 11
Figure 3.2: Variation of mensuration voids at the first compaction cycle against wheel-tracking rate for
asphalt ....................................................................................................................................... 12
Figure 3.3: Influence of moisture and compaction method on density and modulus ................................... 14
Figure 3.4: Influence of moisture and bitumen content on density and modulus ......................................... 15
Figure 3.5: Effect of moisture on density and modulus for Western Freeway trial mixes ............................ 16
Figure 3.6: Calder Freeway trial mix variation of moduli with density and moisture .................................... 17
Figure 3.7: Variation in dry moduli with lime type and dry density ............................................................... 18
Figure 4.1: Effect of Marshall hammer and mould size ................................................................................ 22
Figure 5.1: Comparison of Newell Highway mix design moduli with untrafficked field core values ............. 28
Figure 5.2: Comparison of Newell Highway mix design with untrafficked field core moduli after
adjustment for density test method ........................................................................................... 29
Figure 5.3: Comparison of Calder Freeway mix design moduli with untrafficked field core values ............. 31
Figure 5.4: Comparison of Western Freeway field core and laboratory measured moduli .......................... 33
Figure 5.5: Comparison of mix design moduli and untrafficked field core values after 1–2 years in
service ....................................................................................................................................... 35
Figure 6.1: Modulus variation with density of field-mixed materials, field moisture content marked ........... 37
Figure 6.2: Comparison of field-mixed specimens and laboratory-mixed specimens, field moisture
content marked .......................................................................................................................... 38
Figure 6.3: Effect of time delay between laboratory mixing and compaction ............................................... 40
Figure 7.1: Mechanism causing longitudinal traffic-induced cracks to occur at a transverse shrinkage
crack in a cement treated base ................................................................................................. 44
Figure 8.1: Laboratory-measured fatigue lives of field beams ..................................................................... 47
Austroads 2018 | page iii

1. Introduction
The Guide to Pavement Technology Part 5: Pavement Evaluation and Treatment Design (Austroads 2011a)
includes an interim procedure for the thickness design of foamed bitumen stabilised (FBS) pavements.
However, there has been a concern that the interim FBS design procedure is not as well founded as those
for conventional treatments due to lack of performance data. The design method needed to be
verified/modified with field performance data.
In July 2012, Austroads commissioned ARRB Group to undertake research project TT1825 Mix Design and
Field Evaluation of Foamed Bitumen Stabilised Pavements. The objectives of this project were to:
• improve the Austroads procedures for the design of FBS materials for new pavements and structural
rehabilitation treatments
• identify distress modes of FBS pavements from in-service trial sites
• improve and harmonise national mix design procedures for FBS materials.
Progress reports (Austroads 2013, 2014 and 2015) have been published.
This final report for project TT1825 is structured as follows:

• Section 2 summarises the Austroads test methods developed and mix design framework for inclusion in
a future edition of the Austroads Guide to Pavement Technology Part 4D: Stabilised Materials.
• Section 3 describes the research findings on the effect on mix design indirect tensile modulus of mixing
moisture content and lime type.
• Section 4 reports the extent to which the modulus varies due to the laboratory compaction process.
• Section 5 compares the modulus of laboratory-manufactured specimens with field cores.
• Section 6 investigates several issues associated with modulus testing of treated materials sampled from
the roadbed during construction.
• Section 7 summarises the six trial sites constructed to obtain performance data and describes the
distress types observed during the performance monitoring.
• Section 8 discusses changes to the interim Austroads structural thickness design method.
• Section 9 provides a summary of research findings and the conclusions.
• Appendix A to Appendix F provide details of design, construction and performance monitoring of the six
trial sites.
• Appendix G lists the back-calculated moduli of selected trial sites.
Austroads 2018 | page 1

2. Mix Design Procedures
2.1 Introduction
As described in Section 1, a key project objective was to improve and harmonise national mix design
procedures for FBS materials. Specifically, the project included revision of the Guide to Pavement
Technology (GPT) Part 4D: Stabilised Materials (Austroads 2006) and development of an associated test
method.
Under the direction of the Austroads Pavement Task Force, and with guidance from the Austroads Pavement
Structures Working Group and a Mix Design Working Group, emphasis was placed on building on existing
Australian mix design practices rather than research into the most appropriate performance properties to
measure in mix design.
As described in Section 2.2, for well over 10 years indirect tensile modulus values (initial, cured dry and
cured wet) have been used as performance-related properties. There was a need to harmonise the modulus
test methods that have been developed by road agencies and industry. Such harmonised Austroads test
methods would lead to improved understanding of the characteristics of FBS materials through pooling
national experience and expertise.
Consideration was given as to whether the modulus testing needed to be supplemented by deformation
testing using a small wheel-tracking test for heavily trafficked roads. It was agreed that this was not
necessary at this stage. The principal performance tests related to early-life rut resistance would be grading,
particle shape and initial modulus, with optional use of the wheel-tracking test method.
Section 2.2 provides background to the Australian use of indirect tensile modulus values in mix design.
Section 2.3 provides background to the development of test methods used to prepare and test modulus
specimens. Section 2.4 provides the step-by-step procedures for the determination of design bitumen and
secondary binder contents based on modulus measurement.
The revised text of Part 4D in relation to the mix design foamed bitumen stabilised materials was developed
in conjunction with experienced road agency and industry practitioners. This text and the associated test
method were forwarded separately to Austroads for publication.
2.2 Background to Indirect Tensile Modulus Requirements
2.2.1 Pioneering Work in the 1990s
Mobil Oil first carried out FBS in Australia in the 1960s. Bowering (1973) and Bowering and Martin (1976)
described the early use of foamed bitumen stabilisation in Australia. Due to the oil crisis in the 1970s and
resulting high bitumen costs the use of FBS waned in Australia until the early 1990s. The developments in
the FBS design methods since the early 1990s is summarised below.

Maccarrone et al. (1994) reported the increasing interest in FBS after a hiatus since the 1970s. The authors
described the use of measured indirect tensile modulus for FBS mixes, reflecting the increase in use of that
test method for asphalt. FBS mixtures were compacted using gyratory compaction, dried for 3 days at a
temperature of 60 °C. At that time, the ‘dry’ modulus was measured at a temperature of 25 °C and a rise time
of 50 milliseconds (ms). The ‘wet’ modulus was then measured after soaking for 24 hours in a water batch at
60 °C. Importantly, the mix design process included testing an FBS mix at varying moisture contents to
determine the optimum moisture content (OMC) and maximum dry density (MDD) of the mixture. At OMC the
dry and wet moduli were close to maximum values.
VicRoads trialled the use of FBS on three roads in Metropolitan Melbourne (Lancaster, Mc Arthur &
Warwick 1994). At that time, VicRoads did not have a standard specification for FBS, so various specification
clauses were prepared to describe the requirements for each trial. For one trial, Lancaster, Mc Arthur and
Warwick reported the following indirect tensile modulus requirements at a temperature of 25 °C and with a
rise time of 50 ms:
• a minimum dry modulus of 6000 MPa, after specimens were dried for 3 days at a temperature of 60 °C
• a minimum wet modulus of 1500 MPa, after 24 hours soaking at 60 °C.
A detailed description of the method of mixing and compacting the test specimens was not reported.
2.2.2 Early Development by TMR
Ramanujam and Jones (2000) reported on the design and construction of an FBS project in South East
Queensland. The authors provided the following details in use at that time in relation to the indirect tensile
modulus testing used in the selection of binder contents:
• The host materials were initially mixed with water to a moisture content of 70% standard Proctor OMC of
the untreated materials. Note that this was a departure from the earlier work by Maccarrone et al. (1994),
which included determination of the OMC of the FBS mixture rather than the untreated material.
• 2% hydrated lime with bitumen contents of 2%, 3% and 4% were then mixed into the host material.
• The treated material was then compacted into Marshall moulds using 50 blows of a Marshall hammer.
(Details of the mould diameter and hammer weight were not reported).
• The initial modulus of each specimen was then measured.
• Each specimen was then dried for 3 days at a temperature of 60 °C and the dry modulus measured.
• Each specimen was then soaked for 24 hours in a water bath at a temperature of 60 °C and the soaked
modulus measured.
• The retained modulus after soaking was calculated by dividing the soaked modulus by the dry modulus.
Table 2.1 lists the Queensland Department of Transport and Main Roads (TMR) modulus requirements at
that time, presumably related to measurements at a temperature of 25 °C and a rise time of 40 ms. The
authors reported that, generally, materials were being treated with 3.5% to 4% bitumen and either
1.5% quicklime or 2% hydrated lime.
Table 2.1: TMR early minimum modulus requirements
Parameter Minimum requirement

Initial modulus 500–700 MPa
Soaked modulus 1500 MPa
Retained modulus ratio 0.50
Source: Adapted from Ramanujam and Jones (2000).

In developing these early modulus requirements, TMR built on the performance of foamed bitumen stabilised
pavements constructed by Brisbane City Council (BCC). BCC experience was that when FBS mixes without
lime or cement were opened to traffic shortly after construction, they rutted. The initial modulus values of
mixes without lime were in the range 100–400 MPa. Using this information TMR proposed new design
requirements to inhibit rutting due to early trafficking, namely, minimum initial modulus of 500 MPa for lightly
trafficked roads and 700 MPa for heavily trafficked roads.
The specification of a minimum retained modulus ratio (wet modulus/dry modulus) of 0.50 was built in part
based on the concept of a retained tensile strength used for stripping of asphalt. More importantly, the mixes
used in FBS pavements that survived early trafficking prior to surfacing when inundation by water or
prolonged rainfall had retained ratios exceeding 0.50.
In 2004, Jones and Ramanujam described the TMR method further and refined the modulus requirements
(Table 2.2) to:
• vary the modulus requirements with the design traffic
• increase the soaked modulus requirements for moderately trafficked roads
• increase the cured and soaked modulus values for heavily trafficked roads
• reduce the retained modulus ratio requirements for lightly trafficked roads
• change the description from ‘dry modulus’ to ‘cured modulus’.
Table 2.2: TMR modulus requirements in 2004
Minimum requirements
Design
traffic (ESA) Initial modulus Cured modulus Soaked modulus Retained modulus
(MPa) (MPa) (MPa) ratio (%)
< 106 500 2500 1500 40
106 – 107 700 3000 1800 45
> 107 700 4000 2000 50
Source: Jones and Ramanujam (2004).
It is surmised that this change to increase the modulus requirements with increasing heavy traffic was based
on engineering judgement and the assumption that the higher the modulus the better the rutting resistance.
The associated higher design moduli also allowed a reduction in FBS thickness. Note the significant increase
in the modulus requirements for heavily-trafficked roads (i.e. minimum cured dry modulus 4000 MPa). In
providing this increase, TMR considered the very high in situ moduli of field cores and the moduli
back-calculated from measured surface deflections.
Jones and Ramanujam (2004) provided the following additional details of the TMR test method:
• The temperature at which the test specimens were dried was reduced to 40 °C from 60 °C used earlier.
Presumably, this reduction was made because of concerns that at 60 °C, which is close to the softening
point of bitumen, the bitumen dispersion within the compacted specimens may be affected. The authors
did not discuss whether this change in drying temperature was considered in setting the Table 2.2 limits.
• Prior to the soaked modulus testing, the dried specimens were soaked in water for 10 minutes under a
vacuum. Previously, the specimens had been soaked for 24 hours without application of a vacuum.
Whereas earlier the water was at a temperature of 60 °C, no mention was made of the temperature and it
is assumed to be 20–25 °C. The authors did not discuss whether this change in the test method was
considered in setting the Table 2.2 limits.
• The initial modulus was measured three hours after completion of compaction of each specimen.

• The FBS mixtures were compacted into 150 mm diameter Marshall moulds rather than the commonly
used 100 mm diameter mould used for asphalt. The hammer weight used and number of blows per end
were not described. However, Jones and Ramanujam (2008) later clarified that the loose FBS mixtures
were being compacted into 150 mm diameter Marshall moulds using 50 blows of the impact hammer to
each specimen end. Although not stated, it is understood that a heavy 9.9 kg Marshall hammer was used
by this time.
2.2.3 Current TMR Test Methods and Modulus Requirements
By 2008 over 100 mix designs had been completed by TMR, and Jones and Ramanujam (2008) clarified the
reason for the use of the Marshall hammer to compact test specimens rather than gyratory compaction.
Laboratory-cured samples prepared by applying 50 blows of the Marshall hammer to the end of the
specimen ‘can achieve similar resilient modulus values to the upper 50% of cores extracted from the
pavement following 12–24 months field curing’. This comment presumably related to cores taken in the
wheel paths rather than from untrafficked areas of the pavement.
In addition, Jones and Ramanujam (2008) modified the modulus requirements according to the daily average
Equivalent Standard Axles (ESA) of traffic in the first year and whether the mix was for use as base or
subbase. These limits are those currently in use (TMR 2017b) and are given in Table 2.3, Table 2.4 and
Table 2.5. The indirect tensile modulus values are measured at a temperature of 25 °C and a rise time of 40
milliseconds.
Table 2.3: TMR initial modulus mix design limits for foamed bitumen stabilised materials
Average daily ESA in design year of opening Minimum initial modulus(1) (MPa)
< 100 500
≥ 100 700(2)
1 Initial sample curing time of 3 hours at 25 ºC ±5 ºC required prior to initial resilient modulus testing being completed.
2 It is recommended that wheel tracker testing be completed to confirm the curing time required where the average
daily ESA in design year of opening ≥ 1000.
Source: TMR (2017b).
Table 2.4: TMR cured modulus mix design limits for a foamed bitumen stabilised base
Minimum ‘cured’ indirect Minimum ‘soaked’

Average daily ESA in Minimum retained
tensile modulus indirect tensile modulus
design year of opening modulus ratio(1)
(MPa) (MPa)
< 100 2500 1500 0.40
100–1000 3000 1800 0.45
> 1000 4000 2000 0.50
1 Retained modulus ratio = ‘soaked’ cured modulus divided by the ‘cured’ modulus.
Note: Base is defined as a stabilised pavement layer with a bituminous seal surfacing or less than 100 mm of asphalt
overlying it.
Source: Adapted from TMR (2017b).
Table 2.5: TMR cured modulus mix design limits for a foamed bitumen stabilised subbase
Minimum ‘cured’ indirect Minimum ‘soaked’

Average daily ESA in Minimum retained
tensile modulus indirect tensile modulus
design year of opening modulus ratio(1)
(MPa) (MPa)
< 100 2500 1500 0.40
100–1000 2500 1500 0.45
> 1000 2500 1500 0.50
1 Retained modulus ratio = ‘soaked’ cured modulus divided by the ‘dry’ cured modulus.
Note: Subbase is defined as a stabilised pavement layer with at least 100 mm of asphalt overlying it.
Source: Adapted from TMR (2017b).

Jones and Ramanujam (2008) and TMR (2012) clarified the reasons three modulus values were specified:
• Initial modulus at 3 hours is used to provide an indication of the susceptibility of the material to permanent
deformation (due to trafficking) early in the pavement life. The initial modulus provides an indication of
whether the stabilised materials can be trafficked early in their life (e.g. at the end of a construction work
period). According to TMR (2012) ‘the applicability of the initial modulus tests will need to be assessed
based on the traffic conditions of, and the construction process to be adopted for, each individual project.’
• Cured modulus provides an indication of the medium-term (3–6 months) modulus of the material.
• Soaked modulus provides an indication of the stabilised material’s susceptibility to weakening by water
penetration (e.g. if the pavement becomes inundated).
The modulus criterion given in Table 2.3, Table 2.4 and Table 2.5 were developed by TMR based on
experience in Queensland. According to TMR (2012):
They are appropriate for typical road pavement moisture and traffic conditions in
Queensland. They may be inappropriate where the pavement conditions are atypical (e.g.
the pavement is subjected to severe loading or extreme moisture conditions, an example of
the latter being a floodway). Judgement is required when considering the appropriateness
of these limits for atypical situations. Equally, in situations where the pavement is not
anticipated to ‘wet up’ or be inundated during the design period, some relaxation or either
the soaked modulus (say up to 500 MPa less than the values given in Table 2.4 and
Table 2.5 or retained modulus ratio (say up to 10% absolute) less than the values given in
Table 2.4 and Table 2.5 limits) may be considered acceptable.
TMR’s experience with bituminous mixes has indicated that higher bitumen contents result
in superior fatigue performance, but excessive bitumen can lead to poor rut resistance. For
most suitable materials, 3.0% to 3.5% bitumen and 1% to 2% hydrated lime, both by dry
mass, has been found appropriate.
2.3 Development of Austroads Mix Design Test Methods
2.3.1 Introduction
As mentioned in Section 2.1, at the outset of the project it was decided to build on past Australian mix design
experience which for many years has used indirect tensile modulus as the performance-related property.
Table 2.6 lists the mix design test methods developed in conjunction with road agencies and industry. These
methods closely align with the following TMR test methods:
• Q138 Preparation and Testing of Foamed Bitumen Materials, January 2017 (TMR 2017c)
• Q139 Resilient Modulus of Stabilised Materials (indirect tensile method), January 2017 (TMR 2017d).
The basis for these methods is described below.
Table 2.6: Austroads test methods for FBS mix design
Number Title
AGPT-T301 Determination of Foaming Properties of Bitumen
AGPT-T302 Mixing of Foamed Bitumen Stabilised Materials
AGPT-T303 Compaction of Test Cylinders of Foamed Bitumen Stabilised Mixtures: Part 1 Dynamic Compaction
Using Marshall Drop Hammer
AGPT-T305 Resilient Modulus of Foamed Bitumen Stabilised Mixtures

2.3.2 Determination of Foaming Properties
GPT Part 4D (2006) describes the use of laboratory foaming apparatus to determine the foaming
characteristics of bitumen. The expansion ratio (increase in volume due to foaming) and the half-life (time for
the expanded volume to collapse to half the maximum expansion volume) are the characteristics measured
currently.
The Working Group agreed that testing for expansion ratio and half-life was still appropriate.
Austroads test method AGPT-T301 was developed in consultation with the Working Group from methods
being used by road agencies and industry.
2.3.3 Mixing of Foamed Bitumen Materials
Austroads test method AGPT-T302 was developed in consultation with the Working Group, principally from
TMR test method Q138 Preparation and Testing of Foamed Bitumen Materials, January 2017. During the
development of AGPT-T302, Working Group members met with TMR laboratory staff and resolved many
outstanding differences between existing procedures.
It is well known that the moisture content of the mixture affects the distribution of the foamed bitumen
through the host material and also the density to which the mixture can be compacted. Results obtained in
the project (Section 3) reinforced the significance of moisture content to the achieved densities and moduli.
As described in Section 2.2.1, in the initial development of the mix design process in Australia, the mix
design process included modulus testing over a range of moisture contents to identify the optimum moisture
content (OMC) under the compactive effect applied in the laboratory. Later, TMR decided to simplify the mix
design process by determination of the OMC of the untreated host materials under standard Proctor
compactive effort. Based on their experience, a moisture content of 70% of the untreated material OMC was
used in FBS mix design. Similarly, others use 80–85% of the modified Proctor OMC. These varying FBS mix
design practices reflect the variation in use of standard Proctor or modified Proctor for unbound granular
materials across Australian jurisdictions.
Given the significant influence of moisture content on modulus, a significant challenge was to harmonise the
mixing and compaction moisture contents given entrenched use of different Proctor methods that have
different OMCs. Consideration was given to determination of the OMC of the FBS mixture using the Marshall
hammer, in other words revert to including a range of moisture content in the mix design process as
originally used. Whilst there was interest in this approach, given the considerable experience with the use of
the untreated Proctor OMC values, there was no agreement to implement this method at present.
In summary, there remains a need to harmonise on the method to determining the moisture used in the mix
design.
In the past, both planetary mixers (e.g. Hobart mixer) and Wirtgen pugmills have been used to mix the
binders with the host materials. Given that most if not all the limited number of laboratories capable of FBS
mix design now use pugmills, it was agreed to delete mention of planetary mixers from the test method. The
use of pugmills by all laboratories should improve the reproducibility of the modulus testing.
2.3.4 Compaction of Test Cylinders
One key area where harmonisation was of national benefit was the method used to compact FBS test
cylinders for modulus testing. GPT Part 4D (Austroads 2006) mentions that modulus test specimens are
compacted using either gyratory compaction (80 cycles) or Marshall drop hammer (50 blows per end).
Where the untreated materials include those over 20 mm in size, the Guide recommended the use 150 mm
internal diameter moulds.

In summary at the commencement of the project there were three methods is use:
• gyratory compaction using either 100 mm or 150 mm diameter moulds depending on material size
• dynamic compaction using a standard 4.5 kg Marshall hammer in 100 mm diameter moulds
• dynamic compaction using a heavy 9.9 kg Marshall hammer in 150 mm diameter moulds.
At the third Working Group meeting in August 2014, the findings of Progress Report 2 (Austroads 2014) were
discussed, particularly the difference in modulus due to variation in the specimen size and compaction
methods (Section 4).
It was agreed that the Austroads mix design method should specify one compaction method. The TMR
compaction method using a heavy Marshall hammer was selected due to their considerable mix design
experience with this method. Accordingly, test method T303 provides exclusive use of 152.4 mm (6 inch)
diameter specimens compacted by applying 50 blows to each end of the specimen using a 10.2 kg Marshall
hammer. Note that although TMR had used 150 mm diameter moulds and a 9.9 kg Marshall hammer in the
past, this equipment was purpose-built and not readily available. It was decided to adopt a similar mould
(152.4 mm) and hammer (10.2 kg) as specified in ASTM D5581 07a (2013) Standard Test Method for
Resistance to Plastic Flow of Bituminous Mixtures Using Marshall Apparatus (6-inch-diameter specimen).
Furthermore, it was agreed that research should continue to relate the moduli of gyratory and Marshall
specimens as this may enable a future change to the gyratory method.
This agreement on a single method of specimen compaction was a significant step towards harmonisation
and reproducibility in modulus determination.
2.3.5 Curing of Modulus Specimens
The GPT Part 4D (Austroads 2006) design procedure includes the following indirect tensile modulus testing
based in part on the TMR early use (Section 2.2.2):
• initial modulus testing (uncured, unsoaked)
• specimens are dried for three days at 60 °C and the cured dry modulus is measured
• specimens are then either soaked for 24 hours or soaked in a vacuum chamber for 10 minutes at
0.95 kPa vacuum and the cured wet modulus is measured.
As discussed in Section 2.2.2, over at least the last 10 years the TMR mix design method has been changed
so that:
• initial indirect tensile modulus testing is conducted after three hours curing at 25 °C
• specimens are dried for three days at 40 °C and the cured dry modulus is measured
• specimens are then soaked in water in a vacuum chamber under a partial vacuum of 13 kPa absolute
pressure or less for 10 minutes and the cured wet modulus is measured.
The change in curing temperature from 60 °C to 40 °C was implemented due to concerns that 60 °C was
above the softening point of the bituminous binder which allows its mobilisation and possible absorption into
the aggregate. This bitumen mobilisation was considered a fundamental change to the material’s property
and therefore not representative of the appearance of FBS mixes in situ. The curing process in Austroads
test method T305 aligns with this updated TMR practice.
From testing field cores and analysis of measured deflections, it was concluded (Section 5) that the drying
specimens for three days at 40 °C does not provide an indication of cured in situ modulus other than in the
first few months after construction. A need was identified to include longer-term curing in the mix design
process. It is noted that for plant-mixed materials TMR (2017b) have recently included additional testing after
7 days and 14 days curing to assess modulus development (TMR 2017b). The revised text of Part 4D
mentions this additional curing process.

2.3.6 Resilient Modulus Measurement
In terms of the measurement of modulus, Austroads test method AGPT-T305 references the use of the
method used for asphalt, AS/NZS 2891.13.2 Determination of the Resilient Modulus of Asphalt: Indirect
Tensile Method.
2.4 Description of Procedure to Determine Binder Contents

Framework
The procedure to determine the bitumen and secondary binder contents based on measurement of indirect
tensile modulus values is given in Table 2.7 and Table 2.8.
Table 2.7: Selecting mix components
Step Activity
1 Obtain a representative sample of the untreated material to be stabilised, including any recycled surfacing
materials.
2 Measure particle size distribution and assess whether it is necessary to improve the gradings by adding
other granular materials.
3 Measure liquid limit, plastic limit and plasticity index, and assess whether the material needs to be
pre-treated.
4 Obtain a representative sample of Class 170 bitumen and determine the expansion ratio and half-life of
samples with 2, 3 and 4% water content.
5 Compare measured expansion ratio and half-life against limits. If the limits cannot be achieved at any
moisture, assess whether a foaming agent or another bitumen is needed and repeat step 4. Otherwise,
select the moisture content for foaming and proceed to step 6.
6 Select whether the secondary binder is lime or cement, obtain a representative sample, and check for
compliance.
Table 2.8: Determination of design bitumen content
Step Activity
7 From the specification for construction, determine whether standard or modified Proctor maximum dry
density (MDD) will be used in assessing field compaction and determine the OMC of the untreated
material. Based on this testing, select a range of moisture contents for use in testing (e.g. 70% standard
Proctor or 80–85% modified Proctor optimum moisture content).
8 Select the trial bitumen content and secondary binder content.
9 Add water to the untreated granular material to bring it to the selected moisture content. It may be
necessary to adjust the moisture content to allow for hydration of the secondary binder.
10 Mix the granular materials and secondary binders at a mixing moisture content and leave the mixture to
cure for 45 minutes.
11 Thoroughly mix the foamed bitumen into the mixture.
12 Compact at least three test cylinders by placing the loose mixture into 152 mm diameter moulds and
compact by applying 50 blows per face with a 10.2 kg Marshall hammer.
13 Cure the specimens for three hours at 25 °C and measure the initial indirect tensile modulus – Mi.
14 Dry the specimens for three days at 40 °C and measure the dry indirect tensile modulus – Md.
15 Soak the specimens in a vacuum chamber for 10 minutes at a partial vacuum and measure the wet
indirect tensile modulus – Mw.
16 Measure the specimen dry densities.
17 Calculate the ratio of wet to dry modulus.
18 Compare the measured results of steps 13–17 with the specified values. If the results are below the
requirements, increase the bitumen content and/or secondary binder and repeat steps 8–17. If the results
are above the requirements, decrease the bitumen content and repeat steps 8–17.
19 Select the design bitumen and secondary binder contents.

3. Effect of Moisture Content and Lime Type on

Resilient Modulus
3.1 Introduction
The moisture content of the untreated material to which the foamed bitumen is added affects the density to
which the foamed stabilised material can be compacted, and possibly the dispersion of the foamed bitumen.
The selection of the mixing and compaction moisture content has a significant effect on laboratory modulus
values and also on in situ properties and performance.
Existing guidelines vary for selecting the moisture in mix design (Austroads 2011b). For example, the TMR
method (TMR 2017b) states that the moisture content is usually 70% standard Proctor OMC. At least one
industry laboratory has used 80 to 85% OMC modified Proctor compaction on some projects.
The differences in Australian practices were apparent from the mix design testing undertaken for the Newell
Highway (Section E.3) and Calder Freeway (Section D.4) projects. The variations in moisture contents used
by the two test laboratories are shown in Table 3.1 compared to the field moisture contents during
compaction. As described below, a 1% difference in moisture content can have a substantial effect on the
mix design moduli and hence mix acceptance. In addition to the differences between laboratories,
differences occur between the moistures used in the laboratory mix design and those in field construction.
Table 3.1: Variation in moisture contents used in mix designs
Moisture content (%)
Laboratory Newell Highway Western Freeway

Calder Freeway project
project project
(Appendix D)
(Appendix E) (Appendix F)
TMR (150 mm Marshall mould) 4.5–4.6 7.1–7.3
Downer (150 mm Marshall mould) 5.5 and 5.8
Downer (Gyratory compaction) 3.9–4.3 7.6–7.9
Field compaction 3.9(1) 8.0 4.6
1 Field moisture may be underestimated as the sample was taken after roller compaction, which was at least two hours
after mixing. The air temperature at that time exceeded 35 °C.
It was of interest to investigate the extent to which such moisture variations influence the laboratory density
and modulus values measured in mix design. Using the laboratory testing associated with three of the trial
sections this issue is investigated in Section 3.2 to Section 3.4.
In addition, there are differences in mix design practices between laboratories in lime type use. Due to
occupational health and safety (OHS) concerns in handling quicklime, some laboratories always use
hydrated lime regardless of the lime type used in construction. Other laboratories have suitable OHS
procedures to enable either lime type to be used, which enables alignment of the lime type used in the mix
design with that proposed in construction. It was of interest to investigate whether moduli of laboratory mixed
and compacted specimens vary with lime type (Section 3.5).

3.2 Influence of Moisture Content on Newell Highway Trial Mix
3.2.1 Method
Material sampled from the Newell Highway trial section (Section E.5) was used for the investigation. The
OMC of the untreated material was about 10.5% using the standard Proctor compaction method
(Appendix E.3.3). Hence, the mix design would commonly be undertaken at a mixing moisture content of
7.3%. In addition, the mean moisture content during field compaction was 8.0% (Table E 11). Based on
these values, it was decided to evaluate the effect of the mixing and compacting moisture contents of 7.4,
8.2 and 8.8%, spanning the laboratory and field moisture contents with the value of 8.8% selected to assess
whether if overly wet the distribution of foamed bitumen would be adversely affected.
Each test specimen was prepared by mixing the untreated materials with 3.5% bitumen and 1.5% quicklime,
consistent with the Newell Highway target field values.
The influence of moisture was investigated using two laboratory compaction methods:
• specimens compacted in 100 mm diameter Marshall moulds using a 4.54 kg Marshall drop hammer, with
50 hammer blows to each end, and
• specimens compacted in 150 mm diameter moulds using 120 cycles of gyratory compaction using a
vertical stress of 235 kPa and an angle of gyration of 3°.
3.2.2 Compactability
As part of the preparation of the gyratory specimens, the variations in specimen densities with cycles of
compaction were determined as a measure of the variation of material compactability. The results for three
specimens are shown in Figure 3.1. As expected, the moisture content has a very significant influence on
compactability and hence density and modulus. Note that the density at moisture contents of 8.2% and 8.8%
was similar and greater than around 70% standard Proctor OMC (7.4%).
Figure 3.1: Influence of moisture on the Newell Highway material compactability
2.10
2.05
2.00
1.95
1.90
Dry density 7.4% Moisture content
(t/m3)
1.85 8.2% Moisture content
8.8% Moisture content

1.80
1.75
1.70
0 10 20 30 40 50 60 70 80 90 100 110 120
Cycles of gyratory compaction

Note that for dense graded asphalt, the higher the air voids after one cycle of gyratory compaction the more
rut-resistant the mix, as measured by the small-scale wheel tracker (Figure 3.2). A similar effect may occur
for FBS materials and hence provide insight into FBS mix rut resistance.
Figure 3.2: Variation of mensuration voids at the first compaction cycle against wheel-tracking rate for
asphalt
Source: Austroads (2008).
3.2.3 Density and Modulus
Table 3.2 lists the density and modulus test results, which are plotted in Figure 3.3.
As expected, moisture content has a major influence on the specimen density and hence modulus. Clearly
the choice of moisture content used in the mix design needs careful consideration, as a variation in moisture
content of 0.5–1.0% or less can affect the assessment of whether the material meets modulus criteria and
hence decisions on whether materials are suitable for FBS.
As mentioned above, it is common to use a moisture content of 70% standard Proctor OMC in the mix
design, which for this host material results in mix design moisture content of 7.3%. However, for this mix, the
mix design moduli would have been greater if a higher moisture content was used. This suggests the
distribution of foamed bitumen in the mix was not adversely affected by mixing moisture contents above
7.3%.Figure 3.4 shows the Table 3.2 results for the 100 mm diameter specimens compacted using a 4.54 kg
Marshall hammer, compared to the 150 mm diameter specimens compacted using a 9.92 kg Marshall
hammer for bitumen contents of 2, 3 and 4% (Appendix E.3.3). What is of particular interest is the high
sensitivity to moisture content compared to bitumen content above a bitumen content of 3%. That is, when
the bitumen content increases from 3 to 4%, the modulus increases about 7% (2720 MPa to 2910 MPa)
whereas when the moisture content is increased by 0.8% (7.4 to 8.2%), the modulus increases by about
22% (2880 MPa to 3520 MPa) (Note 70% of standard Proctor OMC was 7.3%).
These results reinforce the need to harmonise on the methods used to select mixing and compacting
moisture contents.

Table 3.2: Influence of moisture on density and modulus of Newell Highway material
Moisture Cured dry Cured wet

Dry density(1) Ratio wet/dry
Compaction method content Sample modulus modulus
(t/m3) modulus
(%) (MPa) (MPa)
100 mm diameter 7.4 M7a 1.946 2730 740 0.27
Marshall
M7b 1.944 2980 900 0.30
M7c 1.941 2930 940 0.32
Mean 1.94 2880 860 0.30
8.2 M8a 2.000 3430 1310 0.38
M8b 2.029 3550 1620 0.46
M8c 2.027 3570 1650 0.46
Mean 2.02 3520 1530 0.43
8.8 M9a 2.082 3820 1850 0.48
M9b 2.078 3730 1940 0.52
M9c 2.087 3730 1970 0.53
Mean 2.08 3760 1920 0.51
150 mm gyratory 7.4 S7a 1.939 1760 580 0.33
S7b 1.945 2250 940 0.42
S7c 1.944 2240 1040 0.46
Mean 1.94 2080 860 0.40
8.2 S8a 2.037 3360 1490 0.44
S8b 2.037 3350 1530 0.46
S8c 2.043 3910 1730 0.44
Mean 2.04 3540 1580 0.45
8.8 S9a 2.075 3700 1800 0.48
S9b 2.070 3630 1790 0.49
S9c 2.067 3560 1890 0.53
Mean 2.07 3630 1820 0.50
1 AS/NZS 2891.9.3:2014 Mensuration Method.

Figure 3.3: Influence of moisture and compaction method on density and modulus
4200
4000
3800
3600
3400
3200
Cured
3000
dry
modulus 2800
(MPa)
2600 7.4% moisture, 100 mm mould Marshall
8.2% moisture, 100 mm mould Marshall
7.4% moisture, 150 mm mould Gyratory
2200 8.2% moisture, 150 mm mould gyratory
8.8% moisture, 150 mm mould gyratory
2000
1800
1600
1.92 1.94 1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10
Dry density (t/m3)
2200
2000
1800
1600
1400
Wet 1200
dry
modulus
1000
(MPa)
600 7.4% moisture, 150 mm mould Gyratory
400
200
0
1.92 1.94 1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10
Dry density (t/m3)

0.6
0.5
0.4
Modulus
0.3
ratio

0.2 8.8% moisture, 100 mm mould Marshall
7.4% moisture, 150 mm mould Gyratory
0.1
0
1.92 1.94 1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10
Dry density (t/m3)
Figure 3.4: Influence of moisture and bitumen content on density and modulus
4400 3.5% bitumen,7.4% moisture, 100 mm mould Marshall
3.5% bitumen,8.2% moisture, 100 mm mould Marshall
4200
3.5% bitumen,8.8% moisture, 100 mm mould Marshall
4000 2% bitumen, 7.3% moisture 150 mm mould TMR mix design
3800 3% bitumen, 7.1% moisture 150 mm mould TMR mix design

4% bitumen, 7.2% moisture 150 mm mould TMR mix design
3600
3400
Cured
dry 3200
modulus
(MPa) 3000
2800
2600
2400
2200
2000
1.92 1.94 1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10
Dry density (t/m3)
3.3 Influence of Moisture Content on Western Freeway Trial Mix

Material sampled from the Western Freeway Ballan trial during construction (Appendix F.4.2) was also used
for the moisture content investigation. The Ballan trial investigated the field performance of a crushed
rock/asphalt planings mixture stabilised with the following two mix types:
• target 3% bitumen and target 1.5% hydrated lime
• target 2.5% bitumen and target 1.5% hydrated lime.

For all modulus test specimens, the untreated materials were mixed at the field moisture content of 4.6% and
a higher moisture content of about 80% modified Proctor OMC of the untreated material. Next, the target
bitumen and lime contents were incorporated. The mixes were then compacted into a 152 mm diameter
Marshall mould using a 10.2 kg Marshall drop hammer to apply 50 hammer blows to each end of the
specimen.
The results are listed in Table F 5 and plotted in Figure 3.5.
This testing confirmed the findings of the Newell Highway material testing (Section 3.2) that the moisture
content at which the FBS mixes are mixed and compacted has a highly significant influence on the density
and modulus.
Figure 3.5: Effect of moisture on density and modulus for Western Freeway trial mixes
3.4 Influence of Moisture Content on Calder Freeway Trial Mix
Table 3.3: Mix design testing post-construction
Mean indirect tensile modulus at

Hydrated 25 °C (MPa)
Bitumen Moisture Mean dry Ratio of
Sample lime
content content density(4) wet/dry
ID content Initial Cured dry Cured wet
(%) (%) (t/m3) modulus
(%) modulus(1) modulus(2) modulus(3)
39A,39B,
3 1.5 3.5 2.119 765 4510 1940 0.43
39C
A, B, C 3 1.5 3.7 2.122 1055 4310 2240 0.52
50A,50B,
3 1.5 4.8 2.124 700 5200 2720 0.52
50C
1 Cured unsealed for three hours at 25 °C.
2 Cured unsealed for three days at 40 °C.
3 Soaked in water under partial vacuum for 10 minutes at 25 °C.

Figure 3.6: Calder Freeway trial mix variation of moduli with density and moisture
5500
5000
Cured dry 4500

indirect
tensile
modulus
(MPa) 4000
3.5% moisture
3.7% moisture
4.8% moisture
3500
3000
2.08 2.09 2.10 2.11 2.12 2.13 2.14 2.15 2.16
Dry density (t/m3)
3000
2500
Cured wet
indirect
tensile 2000
modulus
(MPa)
3.5% moisture
3.7% moisture
1500
4.8% moisture
1000
2.08 2.09 2.10 2.11 2.12 2.13 2.14 2.15 2.16
Dry density (t/m3)
3.5 Effect of Lime Type

As mentioned in Section 3.1, there are differences in mix design practices between laboratories in the lime
type used, specifically quicklime or hydrated lime.
To investigate the influence of lime type on mix design moduli, a high-quality crushed rock base from
Oakland Junction quarry in Victoria was used in the testing. This crushed rock was stabilised using
3.5% bitumen and either 2% hydrated lime or 1.6% quicklime. These lime contents were selected as they
included the same active lime contents. Mistakenly, no additional moisture was added to slake the quicklime.

The moduli and density results are listed in Table 3.4 and the dry moduli are plotted in Figure 3.7. Resilient
moduli were measured under the standard indirect tensile (IDT) testing conditions.
Table 3.4: Effect of lime type on IDT moduli
Lime
Moisture Dry Initial Cured dry Cured wet
activity
Mix Sample content density(1) modulus modulus modulus
as
(%) (t/m3) (MPa) (MPa) (MPa)
Ca(OH)2
3.5% bitumen 93 H1 4.7 2.117 549 4138 2529
and
H2 4.7 2.114 518 4334 2410
2% hydrated
lime H3 4.7 2.111 571 3780 2241
Mean 4.7 2.11 546 4084 2393
3.5% bitumen 119 Q1 4.8 2.058 275 1916 1367
and
Q2 4.8 2.056 271 1980 1511
1.6% quicklime
Q3 4.8 2.059 257 1948 1439
Mean 4.8 2.06 268 1948 1439
1 AS/NZS 2891.9.3:2014b Mensuration Method.
Figure 3.7: Variation in dry moduli with lime type and dry density
5000
4500
Hydrated lime
4000 Quicklime
3500
Cured dry
modulus 3000
(MPa)
2500
2000
1500
1000
2.05 2.06 2.07 2.08 2.09 2.1 2.11 2.12 2.13
Dry density (t/m3)
The mean cured dry modulus of the quicklime specimens is about half the value of the hydrated lime
specimens. It was noted that the dry densities of the quicklime specimens were on average about 3% lower
than the hydrated lime specimens. This may have been due to moisture consumed in slaking the quicklime.
Using the Section 4 findings, a 3% reduction in density produces about a 30–40% reduction in dry modulus.
This suggests that about half the reduction in modulus associated with the use of quicklime is due to the
reduction in density. After allowing for this difference in density, the results tend to indicate hydrated lime
specimens still have higher moduli.

The results tend to suggest that the lime type used in the mix design influences the moduli. Until further
information clarifies this issue, it is proposed that the Austroads mix design method be based on a single
lime type i.e. hydrated lime. Quicklime is not used in some laboratories due to occupational health and safety
(OHS) concerns.
3.6 Summary
Based on the testing, it was concluded that mixing and compaction moisture content has a significant impact
on:
• the dry density to which FBS materials can be compacted
• the modulus.
The results to date suggest that the Austroads mix design processes need to include greater consideration of
the influence of mixing and compaction moisture on mix properties. As discussed in Section 2.3.3, the
process to determine the moisture varies between jurisdictions. To improve the reproducibility of mix design
outcomes a single procedure to select moisture content is required.
Importantly, the results suggest the currently used mix design moisture contents are limiting cured dry
modulus and cured wet modulus values. The design process would be improved by modulus testing over a
range of moisture contents are discussed in Section 2.3.3.
Preliminary testing indicates that the lime type (quicklime, hydrated lime) used in the mix design does
influence the mix design assessment. The Working Group agreed to harmonise on the use of hydrated lime
in the Austroads mix design method.

4. Effect of Laboratory Compaction Method on

Modulus
4.1 Introduction
At the commencement of the project, there were varying processes used in Australian laboratories to
compact modulus test cylinders. The variations included using:
• 150 mm diameter moulds and compacting the FBS material by applying 50 blows per specimen end
using a 9.92 kg Marshall drop hammer with a 150 mm diameter face (later this method was varied slightly
to 152 mm mould and 10.2 kg hammer)
using a 4.54 kg Marshall drop hammer with a 100 mm diameter face
• either 100 mm or 150 mm diameter moulds and compacting the FBS using gyratory (Servopac)
compaction and either adjusting the number of compaction cycles to produce specimens at anticipated
field density or using 80 cycles.
It was of interest to investigate the extent to which such variations in test specimen preparation affect the
measured indirect tensile moduli. The effects of specimen size and compaction method are discussed in
Section 4.2 and Section 4.3, respectively.
Comparison of the moduli obtained by different compaction methods required consideration of the dry
density to which the specimens were compacted. Two test methods are being used which produce
significantly different dry densities:
• AS/NZS 2891.9.3:2014 Methods of Sampling and Testing Asphalt: Determination of Bulk Density of
Compacted Asphalt: Mensuration Method. In this method, specimen volumes are calculated from
measurements of their diameters and heights assuming the specimens are cylindrical. The measured
volume includes the volume of any voids on the surface. For high surface air void mixes or irregularly
shaped specimens, this method can lead to an overestimation of volume and hence underestimation of
density.
• ASTM C642-13, Standard Test Method for Density, Absorption and Voids in Hardened Concrete. The dry
densities obtained by this method are the same as those measured with AS/NZS 2891.9.2:2014 Methods
of Sampling and Testing Asphalt: Determination of Bulk Density of Compacted Asphalt: Presaturation
Method. In this method, specimen volumes are determined from the mass of water displaced when
immersed in water. In this case, the measured volume excludes the volume of voids on the surface. In
this report, regardless of which of these two methods is used, the method is referred to as the
presaturation method.
Section 4.3.3 discusses the differences in densities measured by these two methods.
As the laboratory compaction method affects the modulus, it was also of interest to assess which test
method produces moduli in better agreement with the moduli of field cores. The findings are discussed in
Section 5.

4.2 Effect of Specimen Size
4.2.1 Marshall Specimens
Testing to investigate mould size was undertaken using untreated material sampled from the roadbed of the
Calder Freeway trial (Appendix D.7.5). Laboratory specimens were compacted using:
using a heavy 10.2 kg Marshall drop hammer with a 150 mm diameter face
using a 4.54 kg Marshall drop hammer with a 100 mm diameter face.
The mean modulus and density results are summarised in Table 4.1 and the individual specimen results
plotted in Figure 4.1. The test results are variable and hence additional data is required to establish the effect
of Marshall mould size and hammer weight.
Table 4.1: Effect of Marshall mould and hammer size
Mean indirect tensile

Hydrated modulus (MPa) at 25 °C
Bitumen Moisture Mean dry Ratio
Mould size and lime
hammer mass content Cured dry Cured wet
(%) modulus(1) modulus(2)
Mould 100 mm
diameter, 4.5 kg 3.0 1.5 4.0 2.11 4830 1690 0.35
hammer
Mould 152 mm
diameter, 3.0 1.5 4.0 2.10 3950 2090 0.52
10.2 kg hammer

Figure 4.1: Effect of Marshall hammer and mould size
5500
150 mm mould, 10.2 kg hammer
5000
4500
Cured dry
modulus
(MPa)
4000
3500
3000
2.07 2.08 2.09 2.1 2.11 2.12 2.13 2.14
Dry density (t/m3)
3000
2500 100 mm mould, 4.5 kg hammer
2000
Cured wet
modulus 1500
(MPa)
1000
500
0
2.07 2.08 2.09 2.1 2.11 2.12 2.13 2.14
Dry density (t/m3)
4.2.2 Gyratory Specimens
As part of the laboratory testing of field-mixed material sampled from the Calder Freeway trial
(Appendix D.7.5), FBS material specimens were compacted using gyratory compaction in
100 mm and 150 mm diameter moulds with the number of cycles adjusted to achieve a target wet density of
2.26 t/m3.

The dry densities and moduli are given in Table 4.2.
Table 4.2: Effect of gyratory compaction specimen size
Indirect tensile modulus (MPa)

Moisture Dry at 25 °C
Sample
Compaction method content density(3)
number Cured dry Cured wet
(%) (t/m3)
modulus(1) modulus(2)
100 mm diameter mould, gyratory S1 4.0 2.17 3260 1660
compaction to a target wet density
S2 4.0 2.19 2360 1230
of 2.26 t/m3
S3 4.0 2.19 2660 1670
S4 4.0 2.20 2770 1730
Mean 4.0 2.19 2760 1570
150 mm diameter mould, gyratory 1 4.0 2.26 5900 3240
compaction to a target wet density
2 4.0 2.23 5740 3040
of 2.26 t/m3
3 4.0 2.23 5710 2930
4 4.0 2.23 5620 3130
5 4.0 2.23 5380 2920
6 4.0 2.23 5290 2910
Mean 4.0 2.24 5610 3030
Mean of gyratory results adjusted to
4430 2390
2.19 t/m3
One issue that limits evaluation of the effect of specimen size is the average 2% difference in dry density
between the 100 mm and 150 mm specimens. Using the Figure 6.1 results, pending further research, it is
assumed that both the wet and dry moduli increase 12% for a 1% increase in dry density – a large
adjustment compared to that used for dense graded asphalt (Austroads 2012). Using this density
dependency, the mean modulus of the 150 mm diameter specimens was adjusted to a value at the mean dry
density of the 100 mm diameter specimens, as listed in Table 4.2. This analysis indicates that for a given dry
density, the 150 mm diameter specimens are significantly higher in modulus than the 100 mm diameter
specimens.
4.3 Compaction Method

It was of interest to assess the difference in moduli of the Marshall and gyratory compacted specimens. This
assessment was made without the additional influence of the difference in specimen density that commonly
results in differences between the two compaction methods. This was achieved by adjusting the number of
cycles of gyratory compaction to give similar dry densities to the Marshall specimens.
4.3.1 Comparative Testing of Newell Highway Material
As part of the mix design for the Newell Highway trial (Appendix E.3.2), testing was undertaken by two
laboratories, one at TMR and the other at Downer Infrastructure.
The TMR laboratory testing enabled a comparison of the moduli of 150 mm diameter gyratory specimens
compacted to the same mean dry density as the 150 mm diameter Marshall specimens (Table 4.3). This
data indicates that the mean moduli of gyratory specimens are less than half the Marshall specimen values.

The Downer laboratory testing of the 100 mm diameter specimens also indicated that gyratory specimens
were lower in moduli, but to a lesser extent than observed in the TMR testing of 150 mm diameter
specimens.
Table 4.3: Effect of compaction method on moduli of Newell Highway material
Mean Mean Mean indirect tensile

Bitumen Lime Moisture modulus (MPa) at 25 °C
dry dry
Compaction method content content content
density density(4) Cured dry Cured wet
(%) (%) (%)
(t/m3)(3) (t/m3) modulus(1) modulus(2)
TMR testing (150 mm diameter specimens)
Marshall 9.92 kg 3.0 2.0 7.1 – 1.99 2720 1090
hammer (hyd.
lime)
Gyratory (Servopac) 3.0 2.0 7.1 – 1.98 1110 480
(hyd.
lime)
Downer testing (100 mm diameter specimens)
Marshall 4.5 kg hammer 3.5 1.5 7.7 2.11 – 2570 1710
(quicklime)
Gyratory (Servopac) 3.5 1.5 7.7 2.11 – 2220 1390
(quicklime)
4 ASTM C642-13 similar to AS/NZS 2891.9.2:2014 Presaturation Method.
4.3.2 Comparative Testing of Calder Freeway Material
As part of the Calder Freeway trial (Appendix D), samples of the foamed stabilised crushed rock base were
obtained from the roadbed prior to field compaction (Appendix D.7.5). The samples were transported to
Downer’s laboratory and the following specimens were prepared the following day (after a 24-hour delay
from field mixing):
• four specimens compacted into 100 mm diameter Marshall moulds using 50 blows of a 4.54 kg Marshall
hammer to each end of the specimen
• four specimens compacted into 100 mm diameter moulds using gyratory (Servopac) compaction to a
target wet density of 2.26 t/m3.
The measured initial, cured dry and wet moduli are given in Table 4.4. The mean dry density of the gyratory
specimens was about 1% higher than the Marshall specimens. To enable a comparison at the same dry
density, the mean moduli of the gyratory specimens were reduced assuming modulus decreases by 12% for
every 1% decrease in density. Again, the mean moduli of gyratory specimens are less than half the Marshall
specimen values.

Table 4.4: Effect of compaction method on modulus, Calder Freeway field-mixed material
Dry Moisture Indirect tensile modulus (MPa) at 25 °C

Sample
Compaction method density(4) content Initial Cured dry Cured wet
number
(t/m3) (%) modulus(1) modulus(2) modulus(3)
100 mm diameter mould, M1 2.15 4.0 1090 6410 2740
Marshall hammer 50 blows per
M2 2.16 4.0 1070 6800 2780
end
M3 2.18 4.0 – 6980 3160
M4 2.15 4.0 – 6900 3280
Mean 2.16 4.0 1080 6770 2990
100 mm diameter mould, S1 2.17 4.0 700 3260 1660
gyratory compaction to a target
S2 2.19 4.0 510 2360 1230
wet density of 2.26 t/m3
S3 2.19 4.0 – 2660 1670
S4 2.20 4.0 – 2770 1730
Mean 2.19 4.0 605 2760 1570
Mean of gyratory results adjusted to 2.16 t/m3 2420 1380
4.3.3 Additional TMR Testing
In addition, TMR tested material from the upgrade of the Smith Street Motorway-Olsen Avenue Interchange,
Gold Coast Queensland. The material was stabilised with 3% bitumen and 2% hydrated lime and compacted
into 150 mm diameter moulds using a 9.92 kg Marshall hammer and gyratory compaction. The results are
given in Table 4.5.
Table 4.5: Effect of compaction method on modulus, Gold Coast Interchange project
Mean dry density Mean indirect tensile modulus (MPa)

Moisture (t/m3) at 25 °C
Sample Compaction
content
number method
(%) Mensuration(4) Presaturation(5) Initial Cured dry Cured wet
modulus(1) modulus(2) modulus(3)
Marshall 5.1 2.289 2.353 555 2830 1470
S14/137
Gyratory 5.1 2.282 2.355 270 2250 1230
Marshall 5.5 2.294 2.349 340 2890 1630
S14/138
Gyratory 5.5 2.294 2.350 175 2510 1610
Marshall 5.6 2.318 2.356 330 3290 2080
S14/139
Gyratory 5.6 2.320 2.369 165 2850 1870
5 ASTM C642-13, measured dry densities are similar to AS/NZS 2891.9.2:2014 Presaturation Method.
Again, the moduli of specimens compacted using the gyratory method were lower than Marshall specimen
values but not to the same extent as reported in Section 4.3.1 and Section 4.3.2.
Note from Table 4.5 that dry density results were measured using two methods as described in Section 4.1:
• AS/NZS 2891.9.3:2014, Methods of Sampling and Testing Asphalt: Determination of Bulk Density of
Compacted Asphalt: Mensuration Method

• ASTM C642-13, Standard Test Method for Density, Absorption and Voids in Hardened Concrete. The dry
densities obtained by this method are similar to those measured with AS/NZS 2891.9.2:2014 Methods of
Sampling and Testing Asphalt: Determination of Bulk Density of Compacted Asphalt: Presaturation
Method.
The results for the Gold Coast Interchange project indicate that the presaturation densities are about
2.5% higher in density than the more commonly used mensuration method. The presaturation densities are
higher as the surface air voids are excluded from the volume measurement. For this reason, TMR use the
presaturation method (ASTM C642-13) in mix design, as this method results in the same densities as
AS/NZS 2891.9.2:2014.
4.4 Summary
The findings were:
• Mix design moduli vary significantly with the method used to compact test specimens. It was observed
that Marshall specimen moduli can be more than twice that of gyratory specimens compacted to the
same density. It was concluded that to improve reproducibility of mix design outcomes, the Austroads mix
design process should allow only one method. The project Working Group agreed that the Austroads mix
design method should be based on a single method: 152 mm diameter specimens compacted using a
10.2 kg Marshall hammer consistent with TMR practice.
• In terms of the difference in moduli due to use of 100 mm and 150 mm diameter moulds, while a very
significant moduli difference was measured when the gyratory compaction method was used, the results
using the Marshall compaction method were too variable to determine the effect. The project Working
Group agreed that to improve consistency in mix design outcomes the Austroads mix design method
should be based on 152 mm diameter moulds regardless of the material maximum size.
• Two methods of measuring specimen dry density are being used in mix design: mensuration and
presaturation. As the measured densities differ, it is important that in comparing results of different
laboratories and when mix design moduli are compared to field core moduli, a consistent approach to
density assessment be used. If it is proposed to field test cores after construction to evaluate modulus
and density compliance, the presaturation method is recommended as the mensuration underestimates
the density due to the presence of surface voids.

5. Comparison of Laboratory-mixed and

Compacted Moduli with Field Core Moduli
5.1 Introduction
Section 4.3 indicates that for a given test density, laboratory specimens compacted using the Marshall
hammer have a higher indirect tensile modulus than gyratory compacted specimens. This raised the
question of which method produced specimens in closer agreement to the in situ FBS modulus.
In this section, the laboratory-mix design results obtained by compacting FBS mixes into 150 mm diameter
moulds using a heavy Marshall hammer are compared to early-life in situ moduli obtained by coring.
5.2 Newell Highway, Bellata

Data obtained for the Newell Highway trial (Appendix E) enabled the moduli of 150 mm diameter Marshall
specimens compacted using a 9.92 kg Marshall hammer to be compared to field core moduli. The mean data
is summarised in Table 5.1 and the individual data points plotted in Figure 5.1.
Table 5.1: Comparison of Newell Highway Marshall specimen moduli with untrafficked field core values
Mean Mean Mean IDT modulus at 25 °C

Moisture dry Bitumen (MPa)
Compaction Source of dry
content density(4) content
method material density(3) As Cured Cured
(%) (%)
(t/m3) (t/m3) received dry(1) wet(2)
Laboratory Laboratory-mixed 7.1 – 1.99 3.0 – 2720 1090
compaction and laboratory
using compacted in
Marshall 150 mm diameter
hammer mould (TMR mix
design)
Rollers in the Field-mixed and 8.1 2.00 – 3.5 7500 5330 2320
field field compacted -
sampled by coring
pavement
2 months after
construction(5)
Field-mixed and – 1.95 2.01 – 10 760 10 570 5190
field compacted -
sampled by coring
pavement
18 months after
construction(5)
4 ASTM C642-13, measures similar dry density as AS/NZS 2891.9.2 Presaturation Method.
5 Cores taken along the edge line, untrafficked area of pavement.

Figure 5.1: Comparison of Newell Highway mix design moduli with untrafficked field core values
14000
13000
12000
11000
10000
9000
Indirect
8000
tensile
modulus 7000
(MPa)
6000
Mix design cured dry modulus, 3% bitumen
5000 Mix design cured wet modulus, 3% bitumen
Field cores at 2 months-bottom half
4000
Field cores at 2 months- top half
3000 Field cores at 18 months- bottom half
Field cores at 18 months -top half
2000
1000
1.92 1.93 1.94 1.95 1.96 1.97 1.98 1.99 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.10 2.11 2.12
Dry density (t/m3)
The initial impression from Figure 5.1 is that for a given dry density, the cured dry moduli of the 150 mm
Marshall specimens are about half that of field cores taken two months after construction. In comparing the
modulus values, it should be noted that the method of measuring specimen dry density varied and needs
consideration (Section 4.3.3).
The mensuration method (AS/NZS 2891.9.3:2014) was used to obtain the specimen volume for the
two-month-old field core specimens. In TMR mix design testing, the volume is determined from the volume of
water displaced in the saturated surface bulk density in accordance with ASTM C642-13 (as this method
results in the same dry densities to the AS/NZS 2891.9.2:2014 Presaturation Method, in this report the
method is referred to as the presaturation method regardless of which test method is used). As the specimen
volume measured in the presaturation method excludes surface air voids, the volumes are lower than those
by the mensuration method, and hence the dry densities are higher.
Comparative testing of the 18-month-old field cores indicated that the mean dry density determined using the
mensuration method was about 3.2% lower than the presaturation method (Table 5.1). Using this factor, the
two-month-old mensuration dry densities were converted to estimated presaturation values as used in the
mix design and for the 18-month-old cores.
When these amended densities of two-month-old cores are plotted (Figure 5.2) the mix design cured dry
moduli were reasonably consistent with two-month field core values extrapolated to the lower densities of the
laboratory specimens. That is, the cured dry moduli adjusted to in situ densities provide a reasonable
estimate of early-life field values.
The current Australian practice (Austroads 2011a) is to determine the thickness design modulus from cured
wet modulus values. It should be noted that the mix design wet moduli were less than half the dry moduli.
Given that the cured dry modulus values provide a conservative estimate of in situ values, the use of cured
wet modulus values for dry climates, similar to Bellata, adds further to the disconnect between the design
moduli and the in situ values.

Figure 5.2: Comparison of Newell Highway mix design with untrafficked field core moduli after adjustment
for density test method
14000
13000
12000
11000
10000
9000
Indirect 8000
tensile
7000
modulus
(MPa) 6000
5000
Mix design cured dry modulus, 3% bitumen
4000
Mix design cured wet modulus, 3% bitumen
3000
Field cores at 2 months -bottom half
2000 Field cores at 2 months - top half
1000 Field cores at 18 months - bottom half
0 Field cores at 18 months - top half

1.92 1.93 1.94 1.95 1.96 1.97 1.98 1.99 2.00 2.01 2.02 2.03 2.04 2.05 2.06 2.07 2.08 2.09 2.10 2.11
Dry density using pre-saturation method (t/m3)
It should be noted that at the same density:

• The mix design cured dry moduli were about a third the 18 month untrafficked field core moduli.
• The field moduli more than doubled between 2 months and 18 months.
It was concluded that for the Newell Highway site:

• The mix design curing process provides an indication of early-life in situ values when allowance is made
for difference in densities between those in situ and the mix design specimens.
• Early-life field core moduli are more closely aligned with mix design cured dry moduli rather than cured
wet moduli.
• The mean mix design density was similar to the mean density of the bottom half of field cores taken at
18 months but well below the mean density of the top halves.
5.3 Calder Freeway, Woodend

Data obtained for the Calder Freeway trial (Appendix D) enabled the moduli of 152 mm diameter Marshall
specimens compacted using a 10.2 kg hammer to be compared to untrafficked field core moduli. The mean
data is summarised in Table 5.2 and the individual data points plotted in Figure 5.3. All dry densities were
measured using the mensuration method.

Table 5.2: Comparison of Calder Freeway Marshall specimen moduli with untrafficked field core values
Mean IDT modulus

Mean at 25 °C
Moisture Bitumen (MPa)
Compaction dry
Source of material content content
method density(3) As
(%) (%) Cured Cured
(t/m3) received
dry(1) wet(2)
from field
Laboratory Laboratory-mixed 4.8 2.12 3.0 – 5200 2720
compaction and laboratory
using Marshall compacted in
hammer 152 mm diameter
mould
Rollers in the Field-mixed and field – 2.07 – 2930 5880 3000
field compacted –
sampled by coring
pavement 8 days
after construction(4)
Field-mixed and field – 2.05 – 2560 4250 2880
compacted –
sampled by coring
pavement 2 months
after construction(4)
Field-mixed and field – 2.10 – 11 130 12 200 6950
compacted –
sampled by coring
pavement 2 years
months after
construction(4)
4 Cores taken along the edge line, untrafficked area of pavement.
As shown in Figure 5.3, the mix design cured dry and wet moduli were reasonably consistent with early-life
(8 days and 1 month) field core values at the same dry density.
The mix design cured dry moduli were less than a half the values of two year untrafficked field cores at the
same density.
The field moduli increased by about a factor of three between 1 month and 2 years.
It was concluded that for the Calder Freeway trial:

• The mix design process provides an indication of early-life in situ values similar to the conclusion of
Newell Highway trial (Section 5.2).
• Early-life field core moduli were between the mix design dry and wet moduli, different from the findings of
the Newell Highway trial.
• In contrast to the Newell Highway trial, the mean mix design density was well above the mean density of
the bottom half of field cores. The lower achieved compaction of the Calder Freeway trial was attributed
to the type and weight of the rollers used (Appendix D.7.1).

Figure 5.3: Comparison of Calder Freeway mix design moduli with untrafficked field core values
20000
Mix design cured dry modulus
Mix design cured wet modulus
18000
Field cores 8 days- bottom half
16000 Field cores at 8 days - top half

field core 1 month-bottom half
14000 Field core at 1 month- top half
Field cores at 2 years -bottom half
12000 Field cores at 2 year - top half
Indirect
tensile 10000
modulus
(MPa)
8000
6000
4000
2000
0
1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12 2.14 2.16 2.18 2.20 2.22 2.24
Dry density using mensuration method (t/m3)
5.4 Western Freeway, Ballan

The Western Freeway at Ballan also provided an opportunity to compare the moduli of laboratory-mixed and
Marshall compacted specimens (Table F 5), with field core moduli taken between wheel paths 1 month after
construction and tested for modulus at 3 months and cores taken after 1 year (Section F.9.3). The mean data
is summarised in Table 5.3 and plotted in Figure 5.4.

Table 5.3: Comparison of Western Freeway Marshall specimen moduli with untrafficked field core values
Mean IDT modulus

Mean Mean at 25 °C
Moisture
Compaction dry dry (MPa)
Source of material content
method density(3) density(4)
(%) As Cured Cured
(t/m3) (t/m3)
received dry(1) wet(2)
2.5% bitumen mix
Laboratory Laboratory-mixed and 5.8 2.16 – – 4110 2490
compaction laboratory compacted in
using Marshall 150 mm diameter mould
hammer
Rollers in the Field-mixed and 4.6 2.20 2.28 5910 6950 4950
field field-compacted –
sampled by coring
untrafficked pavement
one month after
construction tested at 90
days
Field-mixed and – 2.27 2.30 11 950 12 670 10 380
field-compacted –
sampled by coring
1 year after construction
3% bitumen mix
Laboratory Laboratory-mixed and 5.5 2.13 – – 3650 2310
compaction laboratory-compacted in
using Marshall 150 mm diameter mould
hammer
Rollers in the Field-mixed and 4.6 2.15 2.26 7020 7840 5780
field field-compacted –
sampled by coring
one month after
construction tested at
90 days
Field-mixed and – 2.24 2.27 12 200 12 750 11 620
field-compacted –
sampled by coring
4 ASTM C642-13 measures similar dry densities as AS/NZS 2891.9.2:2014 Presaturation Method.
Due to loss of material from the surface of the early-life field cores during coring and handling, in this
instance the presaturation dry densities provide a more suitable dataset to compare the moduli than the
mensuration densities. This requires the mensuration dry densities of the mix design specimens (Table F 5)
to be adjusted to estimated presaturation densities. As described in Section 4.3.3, in testing of a mix from
another project the presaturation dry density was about 2.5% higher than the mensuration. Using this factor,
in Figure 5.4 individual adjusted data points were plotted for comparison with the core data.

Figure 5.4: Comparison of Western Freeway field core and laboratory measured moduli
2.5% bitumen mix

20000
Mix design cured wet modulus
18000
field cores sampled at 1 month, bottom half
field cores sampled at 1 month, top half
16000 field cores at 1 year, bottom half
field cores at 1 year, top half
14000
Indirect 12000
tensile
modulus
(MPa) 10000
8000
6000
4000
2000
2.16 2.18 2.20 2.22 2.24 2.26 2.28 2.30 2.32 2.34 2.36
Dry density using presaturation method (t/m3)
3% bitumen mix
17000

15000 Mix design cured wet modulus
Field cores sampled at 1 month, bottom half
Field cores at 1 month, top half
13000 Field cores sampled at 1 year, bottom half
Field cores sampled at 1 year, top half
11000
Indirect
tensile 9000
modulus
(MPa)
7000
5000
3000
1000
2.14 2.16 2.18 2.20 2.22 2.24 2.26 2.28 2.30 2.32
Dry density using presatuation method (t/m3)

It was concluded:
• For this project, the in situ densities were well in excess of the densities of the mix design specimens.
Unlike the Newell Highway and Calder Freeway projects, the mean mix design density was even well
below the mean density of the bottom half of field cores taken at 1 year.
• Comparison of the mix design moduli with cores taken early-life (1 month) is hindered by the density
differences between mix design specimens and cores. For the 2.5% bitumen mix, at the same density the
moduli early-life cores appear closer to the mix design wet moduli than the dry moduli. For the 3% mix,
the mix design moduli are clearly closer to the moduli of early-life cores than 1-year-old cores.
Figure 5.4 shows that the cores sampled 1 year after construction are well above the mix design dry and wet
moduli.
5.5 Combining the Project Results

In Figure 5.5, the results from the Newell Highway, Calder Freeway and Western Freeway are summarised.
Plotted are the mix design moduli compared to the field core moduli taken from untrafficked areas and
adjusted to the same density as the mix design specimens.
It is apparent that the mix design moduli based on curing for 3 days are well below cured in situ values.
There is a case for improving the process for determining FBS design modulus used in structural thickness
design by providing an adjustment of the mix design moduli for in situ curing. This is discussed further in
Section 8.5.
5.6 Summary
In summary the findings were:
• The Marshall mix design specimens have densities both below and above the in situ densities. For
example, for the Calder Freeway trial the mix design specimens were about 3% higher in density than the
bottom halves of two-year-old field cores, whereas for the Western Freeway trial, the mix design
specimens were about 5% lower in density. Given that modulus changes by about 12% for a 1% change
in density, it is apparent that the structural design method would be improved by provision of a method to
adjust the FBS design modulus to the in situ air voids. This is discussed further in Section 8.5.
• Across the three projects investigated, the Marshall mix design moduli reflect early-life (1–3 months) in
situ values. The in situ moduli after curing for 12 months or more are well in excess of the mix design
moduli even after allowance is made for density differences.
• As the mix design moduli of gyratory compacted samples are lower than Marshall specimen values
(Section 4.3), the use of the gyratory method is not recommended at this time. Changes to the gyratory
compaction method and equipment may be required so that moduli are more closely aligned with in situ
values.

Figure 5.5: Comparison of mix design moduli and untrafficked field core values after 1–2 years in service
13000
Calder Fwy
11000
Newell Hwy
Line of equality
9000
Western Fwy
Field core 2.5% bitumen
dry modulus 7000 Western Fwy
(MPa) 3% bitumen
5000
3000
1000
1000 3000 5000 7000 9000 11000 13000
Mix design cured dry modulus (MPa)
7000
Calder Fwy
6000
Western Fwy
Western Fwy 2.5% bitumen Line of equality
3% bitumen
5000
Field core Newell Hwy

wet modulus 4000
(MPa)
3000
2000
1000
1000 2000 3000 4000 5000 6000 7000
Mix design cured wet modulus (MPa)

6. Testing of FBS Mixes during Construction
6.1 Introduction
For some construction projects, quality assurance processes require an assessment of whether the
laboratory compacted modulus of the field-mixed material complies with the mix design requirement. Issues
that arise in such testing were investigated as part of the construction of the Western Freeway trial section.
The findings are discussed in Section 6.2.
A related issue is the effect on modulus and density of time delays between mixing through the foamed
bitumen and compaction. This is particularly of interest when field-mixed materials are sampled from the
roadbed during construction to determine the laboratory maximum dry density at each field density site. In
some projects, these field-mixed samples are also used to provide verification that the specified modulus
requirements have been achieved. The findings are discussed in Section 6.3.
6.2 Comparison of Moduli of Laboratory-mixed and Field-mixed

Materials
Loose FBS crushed rock was sampled from the Western Freeway Ballan (Appendix F.7.7) roadbed on the
day of construction prior to compaction. The samples were transported to Downer’s laboratory and
compacted into 152 mm diameter Marshall moulds by applying 50 blows to each end with a 10.2 kg Marshall
hammer.
The mean measured cured dry moduli, cured wet moduli, and modulus ratio results are given in Table 6.1,
together with the delay times between field mixing and laboratory compaction. The individual cured dry
moduli results are plotted in Figure 6.1.
Note the significantly higher moduli of the 3% FBS mix than the 2.5% mix. This may in part be due to the
shorter delay between field mixing and laboratory compaction for the 3% mix, as seen from the delay times
listed in Table 6.1. The moisture contents of the 3% mix at the time of laboratory compaction were also
slightly higher.

Figure 6.1: Modulus variation with density of field-mixed materials, field moisture content marked
4000
6.4% 6.4%
3500
2.5% bitumen
3.0% bitumen
3000
4.4% 6.4%
5.3%
4.2%
Cured dry 2500 5.3%
4.7% 4.2% 4.7%
indirect tensile
modulus 4.4%
4.2% 5.3%
4.4%
(MPa) 2000
4.1% 4.1%
4.1%
1500 4.7%
3.9% 4.0% 3.9%

1000
3.9%
4.0% 4.0%
500
1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12
Dry density (t/m3)
Table 6.1: Modulus of laboratory compacted field-mixed FBS materials
Indirect tensile modulus (MPa) at 25 °C

Dry Moisture
Time delay Cured dry Cured wet Ratio
Mix Chainage density(3) content
(hours: min) modulus(1) modulus(2) wet/dry
(t/m3) (%)
modulus
2.5% 40 5:20 1.99 4.0 800 635 0.79
60 5:50 1.99 3.9 930 765 0.82
80 6:20 2.08 5.3 2490 1500 0.61
100 6:50 2.06 4.7 2100 1380 0.67
Mean 6:05 2.03 4.5 1660 1100 0.72
3.0% 180 3:05 2.05 4.4 2350 1560 0.66
200 3:35 2.06 4.2 2390 1490 0.63
220 4:05 2.04 6.4 3390 2370 0.70
240 4.35 1.98 4.1 1690 1280 0.76
Mean 3:50 2.04 4.8 2450 1670 0.69
TMR minimum base requirements (> 1000 ESA/day) for specimens
4000 2000 0.50
compacted in 150 mm diameter mould and large Marshall hammer
Figure 6.2 compares the moduli of the field-mixed specimens with laboratory-mixed and compacted
specimens (Table F 5). Note that when preparing the laboratory-mixed modulus specimens in a mix design,
laboratory compaction commences 30–60 minutes after mixing.

The moduli of the field-mixed and laboratory-compacted specimens are about half the values of the
laboratory-mixed and compacted specimens. One very significant reason for the low moduli was the
differences in dry densities, as shown in Figure 6.2 labelled with their moisture contents. The reason for the
lower densities of the field-mixed materials is likely to be related to the generally lower moisture contents and
longer time delays between mixing and compaction (Section 6.3). The delay in compacting the 2.5% bitumen
mix was about 2 hours more than the 3% bitumen mix. This may have been the reason why the 2.5%
bitumen field-mixed specimens had a mean density about 5% below that of laboratory-mixed specimens,
whereas for the 3% bitumen the density difference was only 2–3%.
Figure 6.2: Comparison of field-mixed specimens and laboratory-mixed specimens, field moisture content
marked
(a) 2.5% bitumen mix
4500
5.8%
5.8%
4000 field-mixed, with 5-7 hour delay
lab-mixed 30-60 minute delay 5.8%
4.6%
3500
4.6% 4.6%
3000
Cured dry 5.3%
indirect tensile 4.7%
modulus 2500 5.3%
4.7%
(MPa) 5.3%
2000
1500 4.7%
3.9% 4.0% 3.9%

1000
3.9%
4.0%
4.0%
500
1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12 2.14 2.16 2.18 2.20
Dry density (t/m3)
(b) 3.0% bitumen mix

4500
5.5%
field-mixed with 3-4.5 hour delay
4000 Lab-mixed, 30-60 minute delay
6.4% 6.4% 5.5%
3500
5.5%
4.6%
Cured dry 3000 4.6%
indirect tensile 4.4%
modulus 4.6%
4.2%
(MPa) 2500 6.4%
4.2%
4.4%
4.4%
2000 4.2%
4.1%
4.1%
1500
4.1%
1000
1.94 1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12 2.14 2.16
Dry density (t/m3)

To enable a comparison of the cured dry moduli field-mixed and laboratory-mixed specimens under the
same conditions, allowance needs to be made for differences in density and moisture content. Whilst the
data to date suggests a 1% increase in dry density results in about a 12% increase in cured dry modulus
(Figure 6.1), the effect of moisture (other than its influence on the compacted dry density) remains uncertain.
Hence, the field and laboratory data in Figure 6.2 need to be compared at similar moisture contents.
For the 2.5% bitumen mix, the mean modulus of laboratory-mixed specimens at 4.7% moisture was
compared to the mean modulus of field-mixed material at 4.6% moisture. The difference in mean modulus
between the laboratory-mixed and field-mixed specimens was similar to that expected from the difference in
the specimen mean densities. A similar conclusion was reached for the 3% bitumen mix when the mean
modulus of laboratory-mixed 4.4% moisture specimens was compared to field-mixed specimens at
4.6% moisture.
This suggests that to translate the modulus of field-mixed materials to equivalent laboratory-mixed results
may simply require an adjustment for density difference, provided the moisture contents are the same.
6.3 Effect of Time Delay between Mixing and Compaction

The effect on modulus and density of time delay between mixing through the foamed bitumen and
compaction was investigated. This is particularly of interest when samples of field-mixed materials are taken
from the roadbed during construction to determine the laboratory maximum dry density for use in
field-compaction assessment. In some projects, these samples are also used to provide verification that the
specified modulus requirements have been achieved.
To investigate this issue, the untreated crushed rock/asphalt planings sampled from the Western Freeway
Ballan trial (Section F.4.2) roadbed were used to mix and compact modulus test samples with and without a
24-hour delay between laboratory mixing and compaction. Specimens were prepared for two mix types:
2.5% bitumen and 3% bitumen. All specimens were mixed with 1.5% hydrated lime and 4.6% moisture prior
to mixing with foamed bitumen. For the specimens that were tested after a 24-hour delay, the mixed material
was stored in a sealed container at a temperature of 22 °C between mixing and compaction the next day.
Modulus test specimens were compacted into a 152 mm diameter mould by applying 50 blows to each
specimen end with a 10.2 kg Marshall hammer. The specimens were then tested for indirect tensile modulus
(initial, cured dry and cured wet) at a temperature of 25 °C and 40 milliseconds rise time. The results are
reported in Table 6.2 and the cured dry modulus results are plotted against dry density in Figure 6.3.
Most of the reduction in moduli associated with the 24-hour delay may be attributed to its effect on the
specimen density. However, there may be an effect over and above that due to a reduction in density.
Further testing is required to clarify this issue.

Table 6.2: Effect of time delay on density and modulus results
Mean indirect tensile modulus (MPa) at

Dry 25 °C
Compaction Specimen
Mix type density(3)
delay (hours) no. Initial Cured dry Cured wet
(t/m3)
2.5% bitumen/ <1 A 2.130 785 3490 1990
1.5% hydrated lime/
B 2.139 725 3220 1890
4.6% moisture
C 2.117 705 3340 1950
Mean 2.13 740 3350 1943
24 D 2.075 470 2070 1090
E 2.106 475 2005 960
F 2.055 500 2080 1080
Mean 2.08 480 2050 1040
3.0% bitumen/ <1 J 2.093 685 2900 1710
1.5% hydrated lime/
K 2.101 725 2740 1630
4.6% moisture
L 2.083 680 2830 1840
Mean 2.09 695 2820 1730
24 P 2.063 430 1440 880
Q 2.027 490 1640 980
R2 2.048 475 1510 830
Mean 2.05 465 1530 900
TMR minimum requirements for base with > 1000 ESA/day
700 4000 2000
compacted at 70% of standard Proctor OMC
Figure 6.3: Effect of time delay between laboratory mixing and compaction
4000
2.5% bitumen, 4.6% moisture, < 1 hour delay

2.5% bitumen, 4.6% moisture, 24 hours delay
3500 2.5% bitumen
3% bitumen, 4.6% moisture, < 1 hour delay
E = 17757DD - 34653
3% bitumen, 4.6% moisture, 24 hours delay
3000
Cured dry
indirect tensile 2500 3% bitumen
modulus E = 21192DD - 41673
(MPa)
2000
1500
1000
2.02 2.04 2.06 2.08 2.1 2.12 2.14 2.16
Dry density (t/m3)

6.4 Summary
It was concluded that the sampling and testing of field-mixed FBS to assess compliance with mix design
modulus criteria needs careful consideration.
Time delays between field mixing and laboratory compaction can result in lower test specimen densities, and
hence lower moduli than laboratory-mixed and compacted specimens. In addition, the time delay may reduce
the moduli over and above that due to the reduction in density. To address this issue, TMR require that the
compaction of the field-mixed material in the laboratory commences within three hours of field mixing.
The moisture content of the field-mixed material can differ from that used in the mix design: commonly the
field-mixed materials are drier due to the higher compactive effort used during construction compared to that
used in the mix design process. As a result, the densities of the field-mixed specimens can differ from that of
the mix design specimens with an associated effect on modulus. In such cases, the compliance assessment
needs to consider adjustment of the modulus results for differences in specimen densities.

7. Overview of Field Trial Observations of

Distress
7.1 Introduction
This project investigated the performance of six FBS field trials. Details of each of these trials is documented
in Appendix A to Appendix F.
Table 7.1 summaries the site locations with a broad description of the FBS treatment and pavement
surfacing.
Table 7.1: Summary of trial sites
Mean
Section constructed Bitumen
Year of length Lime content
Road name Surfacing FBS content
construction (%)
(m) thickness (%)
(mm)
Kwinana Freeway, Baldivis 2010 100 30 mm OGA on 150 3.0 0.8%
(Appendix A) geotextile seal quicklime
on 30 mm DGA
(A15E)
Kewdale Road, Welshpool 2011 Section 1: 30 mm DGA 115 3.5 0.8%
(Appendix B) 50 m quicklime
Section 2: 30 mm DGA 115 3.5 0.8%
50 m quicklime
Port Wakefield Road, 2011 Section 1: Sprayed seal 140 >4 1% hydrated
Virginia (Appendix C) 100 m lime
Section 2: Sprayed seal 180 3.5(1) 1% hydrated
917 m lime
Calder Freeway, Woodend 2013 100 Sprayed seal 200 2.8 1.1%
(Appendix D) quicklime
Newell Highway, Bellata 2013 100 Sprayed seal 165 3.0 1.5%
(Appendix E) quicklime
Western Freeway, Ballan 2015 138 Sprayed seal 185 2.5 1% hydrated
(Appendix F) lime
128 Sprayed seal 205 3.0
1 Recovered bitumen contents were highly variable.
One of the key objectives of the project was to gain a better understanding of the distress modes of FBS
pavements. Of particular interest was whether fatigue cracking of the FBS is the predominant distress type
as assumed in the Austroads interim structural design method (Austroads 2011a). Consequently, most of
these pavements were intentionally under-designed in terms of FBS thickness so that their fatigue cracking
characteristics could be observed within the four-year project duration.
This section summarises the findings from the performance monitoring of the trial sites.

7.2 Cracking
A principal objective of the project was to investigate whether fatigue cracking is a distress type for FBS
pavements. The need for this was driven by the fact that the interim Austroads structural design method
(Austroads 2011a) considers fatigue cracking, yet fatigue cracking is not considered in South Africa and New
Zealand. South African mixes are different from Australian mixes, with the Asphalt Academy (2009)
describes the following attributes:
• Typically, bitumen content does not exceed 3% and commonly is in the range 2.0–2.5%.
• Where cement is used as the secondary binder, the content should not exceed 1%.
• Hydrated lime is also as a secondary binder, especially where the untreated material is plastic. In such
cases, the hydrated lime content may exceed 1% if dictated by lime demand (initial lime consumption).
• The amount of fines is less than in Australian mixes (e.g. ideally 5–10% passing 75 µm sieve).
Consequently, the matrix of bound fines may be assumed to be less than for Australian mixes and the South
Africans consider their mix process results in non-continuous binding of the host fines by the bitumen and
secondary binder. As such, they are not considered susceptible to fatigue cracking.
In New Zealand (Gipenz 2014), the bitumen contents are reasonably similar to those currently used in
Australia (typically 2.7–3.5%). Cement is used as the secondary binder, with typical contents in the range
1.0–1.5%. FBS fatigue is not considered in structural thickness design: FBS mixes are considered to behave
as high modulus rut-resistant unbound granular materials. A possible reason that New Zealand FBS mixes
are less susceptible to fatigue cracking than in Australia is that the host materials stabilised generally have
lower fines and hence there is less stabilised mastic.
Given the very high in situ moduli of FBS measured in the project trial sections (Figure 5.5), there is no doubt
FBS mixes being used in Australia are fully bound materials. As with other bound pavement materials
(asphalt, cemented materials, concrete), there was an expectation at the commencement of the research
project that FBS would be susceptible to fatigue cracking.
Wheel path surface cracking was observed at four of the six sites. At one of these sites (Newell Highway,
Bellata), the presence of reactive subgrade soils casts doubts on whether longitudinal environmental
cracking has influenced the performance. In addition, the inner wheel path was overdosed with lime that may
have resulted in the FBS mix being more susceptible to fatigue. Moreover, the pavement supporting the FBS
base had back-calculated moduli well in excess of that expected for unbound granular material. A pavement
investigation may assist in clarifying whether the pavement had been previously in situ stabilised using a
cementitious binder.
However, at the remaining three sites the cracking was concluded to be traffic-load induced. In this regard,
the Western Freeway, Ballan trial of 2.5% bitumen and 3% bitumen mixes provided the most conclusive
evidence that fatigue cracking is the dominant distress type of a conforming FBS mix constructed to a high
quality. Fatigue cracking was also observed on the Port Wakefield Road trial that had bitumen contents
towards the maximum values now being used (3–4%). Fatigue cracking was also observed on the Calder
Freeway, Woodend trial; however, in some locations the top 40 mm thickness of the FBS layer delaminated
from the underlying FBS. This delamination may have been due to reworking the final compacted surface
during construction. Despite coring confirming the fatigue cracking was full-depth, this delamination may
have influenced the performance.
Of the two sites where cracking has not been observed at the completion of the research project, the
absence of cracking at the Kwinana Freeway, Perth trial may be due to:
• the thin polymer modified asphalt and geotextile-reinforced seal surfacing over the FBS base inhibits the
development of surface cracks
• an under-estimation of the support provided by the sand subgrade.

The other site where cracking has not been observed is the thin asphalt surfaced FBS base at Kewdale
Road, Welshpool. Coring of the pavement in 2017 indicated that in at least some areas the FBS layer is
being supported by bound cemented subbase. Such a subbase reduces the strains applied to the FBS layer
and this is likely to be a factor in the lack of fatigue cracking to date. Monitoring of this site has ceased.
An interesting finding was that the nature of initiation and progression of cracking varied between trial
sections. At the Port Wakefield Road and Western Freeway trials, in general the cracking appeared to
commence as transverse cracking in the wheel paths, followed by longitudinal cracking intersecting the
transverse cracks, soon after forming block cracking. This is consistent with the predictions of the interim
Austroads design method as the predicted longitudinal strains exceed the transverse strains.
For the Newell Highway and Calder Freeway sites, longitudinal wheel path surface cracking was firstly
observed rather than transverse cracking which is contrary to the predictions of the Austroads structural
design method. In the case of Newell Highway, the longitudinal cracking was influenced by the underlying
reactive subgrade rather than solely related to the applied traffic loading.
Molenaar and Pu (2008) reported the same for cement treated base pavements in the Netherlands. To
explain this unexpected finding, they proposed that firstly transverse shrinkage cracking occurs but is not
observed due to thick overlaying asphalt in the Netherlands. Load-induced longitudinal cracking then occurs
between the shrinkage cracking due to the higher strains caused by the reduced load transfer at transverse
cracks (Figure 7.1).
This same mechanism may have occurred at the Newell Highway and Calder Freeway trial sites, as fine
shrinkage cracking was not observed in the first few years after placement of a double application sprayed
seal surfacing. A hypothesis yet to be tested is whether the inclusion of lime as a secondary binder leads to a
different form of fatigue cracking from that observed for asphalt, depending on the lime content and amount
of plastic fines.
From monitoring of the six sites, it was concluded that:

• cracking is the distress type limiting the pavement life
• fatigue cracking is a distress type for Australian FBS mixes
• shrinkage cracking of the FBS may be influencing fatigue performance
• other possible causes of cracking in FBS include proximity of the wheel path to longitudinal construction
joints, reactive subgrades and reflection cracking from the underlying cemented materials.
It is recommended that the Austroads pavement design method continue to consider FBS fatigue cracking in
determining layer thicknesses. An alternative design approach considering FBS as modified granular
material not susceptible to fatigue cracking was not supported from field trial monitoring of FBS mixes
currently being designed and constructed in Australia.
Figure 7.1: Mechanism causing longitudinal traffic-induced cracks to occur at a transverse shrinkage crack
in a cement treated base
Source: Molenaar and Pu (2008).

7.3 Rutting
In general, rutting did not increase significantly with trafficking except when the pavement was opened on the
day of construction. There is no indication that the sites will need shape correction before treatment for
cracking. In general rutting and roughness only increased after severe cracking developed. It seems the
current mix design process results in high modulus, rut-resistant mixes.
However, two short lengths of trial pavements needed to be removed and replaced due to rutting:
• The Port Wakefield Road site included a 30 m long section where unknowingly an asphalt patch was FBS
treated and rutted quickly during the first summer (Appendix C.6). This reinforces the current construction
practice of removing asphalt patches prior to FBS, and possibly afterwards spreading all or part of the
asphalt millings along the project.
• The Calder Freeway trial was opened to traffic within two hours of completion of construction and prior to
resurfacing. By the next day, the surface had ravelled (Appendix D.7.2) and a 10 m length at the start of
the project had rutted such that it needed to be removed and patched with asphalt. For
moderate-to-heavily-trafficked roads, this suggests that it is good practice to at least cure the FBS
overnight and apply the surfacing before opening traffic. Proof rolling can also provide guidance on the
timing of opening to traffic.
7.4 Flushing/Embedment
At the Calder Freeway and Western Freeway sites, flushing of the bitumen or embedment of the sealing
aggregate was observed. Advice was subsequently sought from experienced TMR staff on the design of
sprayed seal surfacing. Their advice was to avoid the use of cutting oil in the bitumen as this has the
potential to soften the bitumen within the FBS mix. In addition, TMR recommends the use of an armour-coat
seal prior to the double application seal as a further measure to inhibit embedment.

8. Austroads Thickness Design Method
8.1 Introduction
A principal objective of the project was to improve the Austroads procedures for the structural design of FBS
materials for new pavements and structural rehabilitation treatments. The interim Austroads structural
thickness design method is described in the Guide to Pavement Technology Part 5: Pavement Evaluation
and Treatment Design (Austroads 2011a).
Section 8.2 provides background to this interim method which is summarised in Section 8.3. Using the
results of monitoring the trial sections, Section 8.4 compares the performance predicted using the interim
method with the observed field performance.
Section 8.5 discusses possible improvements to the method of FBS modulus determination. Similarly,
Section 8.6 discusses the limitations of the current fatigue relationships and proposed research for its
improvement.
Section 8.7 and Section 8.8 summarise the findings and the proposed changes to the interim method.
8.2 Background to the Interim Method
8.2.1 Early use of Mechanistic-empirical Method
Maccarrone et al. (1994) reported that the structural thickness design of FBS pavements was determined
using a mechanistic-empirical method adapted from that used for asphalt pavements (Austroads 1992 –
listed under superseded Austroads publications in the references). The following details were provided:
• The FBS design modulus was determined from the indirect tensile modulus measured at a rise time of
50 ms on wet conditioned samples. The authors noted ‘a conservative value of about 1500 MPa at 25 °C
can be used for designs’.
• The FBS fatigue relationship was that used for asphalt. This was decided based on research undertaken
in the United Kingdom.
Lancaster, Mc Arthur and Warwick (1994) of VicRoads confirmed the use of this mechanistic-empirical
method. FBS material was considered a bound layer and the design was based on the critical tensile strains
predicted at the bottom of the FBS layer. They recommended that both dry cured and ‘soaked’ moduli be
alternatively used in the pavement design, as a sensitivity analysis of the design. The authors mentioned the
need to determine the fatigue relationship for use in thickness design and proposed that field beams be
tested and compared to dense graded asphalt.
8.2.2 Laboratory Fatigue Life of Field Beams
Leek (2001, 2010) reported an extensive fatigue testing program of field beams sawn from City of Canning
(Western Australia) pavements. The field beams were sampled at least 3 months after construction. The
beams were cut from the upper half of the constructed depths.
The initial flexural moduli of the beams were measured at a frequency of 10 Hz and a temperature of 20 °C.
These moduli varied over a wide range (1400–9400 MPa) with a mean modulus of about 6000 MPa. As
shown in Figure 8.1, there was a large scatter of measured fatigue lives.

The following FBS fatigue relationship (Equation 1) was developed (Leek 2010).
1558 6 1
N=� �
𝜇𝜇𝜇𝜇
where
N = number of load repetitions to half the initial modulus
𝜇𝜇𝜇𝜇 = tensile strain (microstrain)
Using the applied tensile strain, the measured initial flexural modulus and the volume of bitumen of each test
beam, Leek (2010) also predicted the fatigue lives using the Austroads asphalt fatigue relationship. These
predicted fatigue lives are also plotted in Figure 8.1 (data is termed ‘predicted life by asphalt model’ and the
line of best fit the ‘MRDQ predicted life’). The fatigue life using the Austroads asphalt fatigue relationship and
the measured flexural moduli and volume of bitumen were in reasonable agreement on average with the
measured fatigue lives considering the very high scatter. Nevertheless, the Austroads asphalt relationship
did not provide a reliable method of predicting the fatigue life of individual FBS test beams.
Figure 8.1: Laboratory-measured fatigue lives of field beams
1.E+10
Test Beam Results

1.E+09
Predicted Life by Asphalt Model

1.E+08
MRDQ Predicted Life
Load Repititions
1.E+07
Asphalt model
best fit curve
1.E+06
1.E+05 N = (1558.175/µe)
5.9369
1.E+04
1.E+03
0 50 100 150 200 250 300 350 400
Strain Level (µe)
Source: Leek (2010).

8.2.3 TMR Development of the Method
Based on limited Queensland field performance data, Jones and Ramanujam (2004) concluded that the
primary distress mechanism of foamed bitumen stabilised pavements was fatigue cracking of the stabilised
layer. A mechanistic-empirical thickness design method was adapted from that used to design asphalt
pavements. It included:
• The design modulus of the FBS layer was determined from the soaked indirect tensile modulus measured
in the laboratory-mix design subject to a maximum design modulus of 2500 MPa.
• To predict the allowable traffic loading to fatigue cracking of the FBS layer, the same asphalt fatigue
relationship was used but with a volume of bitumen in the range 6–8% as calculated for the FBS mix from
the bitumen content.
Ramanujam and Jones (2007) explained the reason why the soaked indirect tensile modulus was used. FBS
moduli back-calculated from measured surface deflections and preliminary laboratory beam flexural testing
indicated flexural moduli in the range 500 MPa to 2000 MPa. These values closely resembled the soaked
indirect tensile moduli measured in the mix design. FBS design moduli of 1200 MPa to 1500 MPa were
mentioned as being used to predict FBS fatigue life. Cognisance may also have been taken of the design
moduli (1500 MPa) initially used by Maccarrone et al. (1994) (Section 8.2.1).
Jones and Ramanujam (2008) listed two load-associated distress types for FBS materials:
• Rutting of the FBS layer: this distress mode can generally be avoided by confirming the FBS material
complies with the minimum modulus values (Table 2.3, Table 2.4 and Table 2.5).
• Fatigue cracking of the FBS layer: for thickness design purposes the fatigue life is determined using the
tensile strains at the bottom of the FBS layer.
Jones and Ramanujam (2008) provided a method of adjusting the measured soaked indirect tensile moduli
from the test temperature of 25 °C to the in-service temperature (weighted mean annual pavement
temperature) and from 40 ms rise time in the testing to the heavy vehicle loading rate in situ. The process
used to determine the speed adjustment factors was not described. Again, the method for determination of
the design modulus included a maximum design modulus of 2500 MPa.
In terms of the FBS fatigue relationship, Jones and Ramanujam (2008) summarised the finding of
Leek (2001) from testing field beams (Section 8.2.2). They recommended the continued use of the asphalt
fatigue relationship with a volume of bitumen from 5% to 8% depending on the design bitumen content and
mix design cured wet indirect tensile modulus adjusted for temperature.
As discussed in Section 8.2.2, the analysis by Leek (2010) indicated that the Austroads asphalt fatigue
provided a reasonable estimate of the average FBS laboratory fatigue lives. In this evaluation, the initial FBS
flexural moduli were used. The mean flexural modulus was about 6000 MPa. These beam flexural moduli
were mostly higher than design moduli used by TMR based on the soaked indirect tensile modulus up to a
maximum of 2500 MPa. Presumably, TMR used these comparatively low FBS design moduli for all projects,
regardless of the risk of water saturation, as they resulted in FBS thicknesses with an appropriate degree of
conservatism.
The current TMR practice (TMR 2012) is similar to that described in 2006 and 2008, except that a procedure
for adjusting for loading rate is no longer provided.

8.3 Interim Austroads Method
8.3.1 Introduction
Building on the TMR design method, Austroads (2011a) published an interim structural design method as
described in Section 8.1.
This mechanistic-empirical design method is an adaptation of the method used for the design of new asphalt
pavements (Austroads 2012). Two distress types influence the pavement thickness requirements:
• fatigue cracking of the FBS layer determined using the predicted horizontal tensile strain at the bottom of
the FBS layer, and
• rutting and shape loss structure determined using the vertical compressive strain on the top of the
subgrade.
Similar to the design of asphalt pavement, the thickness design method does not include a method to predict
the rutting of the FBS layer. Rutting of the FBS layer can generally be avoided if the moduli meet the
minimum requirements (Jones & Ramanujam 2008). Early-life modulus is enhanced by the use of a
secondary binder such as lime or cement.
In this interim method, the requirement for adequate FBS fatigue life dominates the required pavement
structure. The allowable loading in terms of FBS fatigue is calculated from the predicted tensile strain at the
base of the FBS layer and the Austroads asphalt fatigue relationship (Section 8.3.3). These thickness design
procedures assume the FBS treatment:
• includes sufficient quantity of residual bitumen (and secondary stabilising agent) to produce a bound layer
with significant tensile strength and hence fatigue properties
• does not include so much secondary binder or a type of secondary binder that results in mix overly
susceptible to fatigue.
The Guide states ‘usually a minimum 3% residual bitumen with lime or cement as secondary binders is
required to produce a foamed bitumen layer that meets the above criteria.’
The development of the Austroads method is described by Gonzalez and Jameson (Austroads 2011c).
8.3.2 Method to Determine FBS Design Modulus
According to Austroads (2011a), for pavement design purposes, the appropriate design modulus is an
estimate of the value obtained from the resilient modulus measured using the standard indirect tensile
test (ITT) adjusted to the in-service temperature (weighted mean annual pavement temperature, WMAPT)
and for the rate of loading to which the layer will be subjected in the roadbed.
The interim method for determination of FBS design modulus is as follows:

• determine the indirect tensile resilient modulus of soaked specimens, following AS/NZS 2891.13.2-1995
• adjust the modulus to the in-service pavement temperature
• adjust the modulus to the in-service traffic loading rate.
Equation 2 is used to adjust the measured indirect tensile modulus from the measurement temperature of
25 °C to the in-service temperature (WMAPT). This method was based on data obtained by indirect tensile
modulus testing (Leek 2001) and FWD testing on foamed bitumen pavements (Fu & Harvey 2007).

Modulus at WMAPT 2
exp(-0.025[WMAPT-T])
=
Modulus at test temperature (T)
where
WMAPT = weighted mean annual pavement temperature (°C)
T = temperature of indirect tensile resilient modulus test (°C), commonly 25 °C
Equation 3 is used to adjust the measured indirect tensile modulus from the laboratory measurement loading
rate of 40 ms rise time to the loading in service.
Modulus at speed V 3
= 0.46V0.16
Modulus measured at 40 ms rise time
where
V = heavy vehicle design speed (km/h)
Equation 3 was derived (Austroads 2011c) using:

• indirect tensile modulus test results of Calder Freeway trial (Appendix D.15) field cores measured using a
range of loading rates (rise times of 25 ms to 100 ms)
• assuming a rise time of 40 ms in the indirect tensile test is equivalent to a flexural modulus measured at a
sinusoidal frequency of 14.8 Hz, consistent with that used for the adjustment of asphalt moduli
• assuming the heavy vehicle loading rate in service is calculated as 2π times the sinusoidal frequency
consistent with that used for the adjustment of asphalt moduli.
Hence, the loading rate adjustment also includes an adjustment from indirect tensile modulus to flexural
modulus consistent with that used for asphalt.
The FBS design modulus is determined by multiplying the mean soaked indirect tensile modulus by the
temperature adjustment ratio (Equation 2) and by the loading rate adjustment ratio (Equation 3), subject to a
maximum value of 2500 MPa.
Unlike the method for determination of asphalt modulus, a method to adjust for differences between the test
specimen and in-service air voids was not provided.

8.3.3 Fatigue Relationship
As an interim measure, GPT Part 5 (Austroads 2011a) recommends that the fatigue relationship for FBS
layers be that used for asphalt fatigue (Austroads 2012), with the appropriate design moduli and volume of
bituminous binder, namely (Equation 4):
5 4
 6918(0.856 Vb + 1.08) 
N =  
 E 0.36 με 
where
N = allowable number of repetitions of the load
µε = tensile strain produced by the load (microstrain)
Vb = percentage by volume of residual bitumen in the foamed bitumen mix (%)
E = design modulus of the foamed bitumen mix (MPa)
8.3.4 Surfacing Requirements
GPT Part 5 also provides guidance with the surfacing requirements where the pavement includes an FBS
base. Namely, where design traffic exceeds 107 ESA a minimum of 30–40 mm hot mix asphalt layer is
recommended. For lower traffic loadings, either a sprayed seal or hot mix asphalt can be used.
8.4 Initial Comparison of Predicted and Observed Performance
8.4.1 Introduction
To assist in identifying the need for improvements in the interim method, the allowable traffic loadings
predicted using the method were compared to the observed performance of the trial sites.
8.4.2 FBS Design Moduli
To determine the FBS design modulus the mean cured wet indirect tensile modulus needs to be measured
on samples prepared in accordance with the Austroads mix design method.
Cured wet modulus of 150 mm diameter specimens compacted using the heavy (9.9 kg or 10.2 kg) Marshall
hammer were used in the evaluation. For three of the six sites (Kwinana Freeway, Kewdale Road and Port
Wakefield Road) such moduli were not measured. Consequently, the analysis was limited to the remaining
three sites as detailed in Table 8.1. For three of the four mixes, the design modulus was limited to the
maximum value of 2500 MPa.

Table 8.1: FBS design moduli from measured soaked indirect tensile modulus
Weighted mean
Mean soaked annual Heavy vehicle FBS design
Trial site IDT modulus pavement design speed modulus
(MPa) temperature (km/h) (MPa)
(°C)
Calder Freeway, Woodend, Vic 2720 24 80 2500 (max)
Newell Highway, Bellata, NSW 1090 31 80 870
Western Freeway, Ballan, Vic
2490 20 80 2500 (max)
2.5 % bitumen mix
Western Freeway, Ballan, Vic
2310 20 80 2500 (max)
3% bitumen mix
8.4.3 Adjusted FBS Fatigue Relationship used in the Analysis
When Equation 4 is used to predict the allowable traffic loading in terms of asphalt fatigue, it is assumed that
95 out of 100 projects will have observed fatigue lives exceeding the Equation 4 predictions. In other words,
Equation 4 is associated with a design reliability of 95%.
For the analysis of the project findings, the best estimate or mean predicted fatigue life is more appropriate to
compare with the observed fatigue life. In the forthcoming revision of the Guide to Pavement Technology
Part 2: Pavement Structural Design (Austroads 2012), asphalt fatigue lives with a design reliability of 50%
will be estimated by multiplying the allowable traffic loadings predicted using Equation 4 by a shift factor of 6.
Accordingly, in predicting the FBS fatigue lives in Section 8.4.4, this factor of 6 was applied to the Equation 4
predictions.
8.4.4 Predicted Allowable Traffic Loadings
Using the interim method modified to predict mean life (Section 8.4.3) and the pavement structures
described in Appendix D.6.1, Appendix E.4 and Appendix F.5, the allowable traffic loadings were calculated
(Table 8.2). In all cases, these loadings were limited by the FBS fatigue cracking rather than rutting and
shape loss.
Table 8.2: Predicted allowable loadings
Mean
FBS Annual
Volume FBS tensile allowable Predicted
design Fatigue traffic
Trial site of binder strain traffic fatigue life
modulus constant loading1
(%) (microstrain) loading (years)
(MPa) (ESA)
(ESA)
Calder Freeway, 2500 6.0 2572 219 1.1 x 106 9.0 x 105 1–2
Woodend, Vic
Newell Highway, 870 7.0 4278 320 2.2 x 106 8.7 x 105 2–3
Bellata, NSW
Western Freeway,
Ballan, Vic 2500 5.2 2288 197 1.1 x 106 1.4 x 106 1
2.5% bitumen mix
Western Freeway,
Ballan, Vic 2500 6.2 2643 181 3.3 x 106 1.4 x 106 2–3
3% bitumen mix
1 Assuming ESA = SAR5/1.2.

8.4.5 Comparison of Observed and Predicted Fatigue Lives
For the Calder Freeway trial, the mean predicted fatigue life was 1–2 years. As detailed in Appendix D.12, no
cracking was observed after the first year of trafficking but after 2 years significant wheel path cracking was
observed. After a further 2 years of trafficking the cracking became widespread, but block cracking had not
yet developed to the severity that necessitates the area be treated.
The FBS mix used on the Newell Highway trial had mix design moduli well below the requirements specified
by TMR. For instance, the cured wet modulus was 1090 MPa, about half the minimum acceptable TMR
modulus. The interim method predicts fatigue of the 165 mm thick FBS layer after 2–3 years trafficking. As
detailed in Appendix E.6.5, when the pavement was inspected after 14 months trafficking no significant
cracking was observed. However, when the pavement was inspected after almost 3 years trafficking very
fine longitudinal and transverse cracking was present. If this cracking is due to FBS fatigue, there is good
agreement between the observed and predicted lives. However, it is uncertain whether the reactive subgrade
at this site or underlying cement treated materials have been influential in the development of cracking.
When the cracking develops further, the pavement will be excavated to clarify the cause.
At the Western Freeway, Ballan, there was a minor area of cracking in the 2.5% bitumen mix after 1 year of
trafficking. After a further year of trafficking significant areas of block cracking had developed in both the
2.5% bitumen and 3% bitumen mixes.
During construction, old asphalt patches were removed, and the excavated areas filled with imported
crushed rock. This untreated crushed rock had a moisture content suitable for compaction of an untreated
crushed rock, but on the wet side of the optimum moisture for compaction after FBS stabilisation. The fatigue
cracking observed was more prevalent in these reinstated patching areas than elsewhere. This may have
been because the high crushed rock moisture content affected the surface area of fines to which the foamed
bitumen adheres. Alternatively, these areas may have weaker subbase/subgrades as supported by the fact
that they previously needed asphalt patching.
Nevertheless, fatigue cracking was observed in areas that had not previously been asphalt patched, albeit to
a lesser extent and severity. After 2 years of trafficking the observed performance of these areas was not
inconsistent with the predicted lives. However, the unexpected finding was the extent and severity of
cracking in the two mixes were similar after two years despite their difference in binder content. This
performance was contrary to the prediction that the lower FBS thickness and volume of bitumen of the
2.5% mix should have resulted in it being more susceptible to fatigue cracking.
8.5 Review of FBS Design Modulus
8.5.1 Definition of Design Modulus
As stated in Section 8.3.2, the interim design method defines the FBS design modulus as follows:
For pavement design purposes, the appropriate design modulus is an estimate of the value
obtained from the resilient modulus measured using the standard indirect tensile test (ITT)
adjusted to the in-service temperature (WMAPT) and for the rate of loading to which the
layer will be subjected in the roadbed.
Although not specifically stated in the method, the adjustment for loading includes an adjustment from
indirect tensile modulus to flexural modulus determined from asphalt testing (Section 8.3.2).
This design modulus definition aligns with that for asphalt used in the design of new flexible pavements
(Austroads 2012), except that the FBS modulus is after curing and soaking.

8.5.2 Use of Mix Design Moduli
Currently the design moduli are determined from the indirect tensile moduli measured in the mix design. The
following moduli of laboratory-mixed and compacted specimens are measured:
• initial modulus 3 hours after compaction
• cured dry modulus, after 3 days drying at a temperature of 40 °C
• cured wet modulus, after soaking the specimens.
The following issues were identified about the use of mix design cured wet moduli to determine the design
modulus:
• In the mix design, the specimens are dried for 3 days at a temperature of 40 °C before testing the ‘cured
dry modulus’. Although this process increases the modulus by drying the specimens, it only appears to
simulate the curing that occurs in the roadbed in the first few months (discussed in Appendix D.11.1,
Appendix E.6.3 and Appendix F.9.3). As shown in Figure 5.5 the mix design moduli are well below that of
field core values taken after 1–2 years in service. This brings into question whether the design moduli
should continue to be based on early-life moduli estimated from the mix design testing or the method
changed to adjust the mix design moduli for curing in service.
• Limited modulus data was obtained on the modulus of field cores without the influence of water used for
saw cutting in the field and in the preparation of modulus specimens. However, the moduli of early-life
field cores without wetting or drying appeared to be more closely aligned with cured dry modulus values
than cured wet modulus. For the trial sites, which were not in high rainfall climates or subject to
inundation, the use of cured wet moduli to derive design moduli would seem to provide an additional
degree of conservativism on top of the use of early-life values.
• FBS modulus increases markedly with density. Significant differences were observed between the dry
density of the mix design specimens and in situ densities. There is sufficient information available to
provide a method to adjust for these density differences in the determination of design modulus.
It is apparent that procedures to determine FBS design modulus could be improved. However, the use of the
current very conservative design moduli together with the asphalt fatigue relationship appears to provide
fatigue life predictions in reasonable agreement with observed performance (Section 8.4.5). Increasing the
FBS design modulus values to more closely align with in situ values without changing the fatigue relationship
will result in lower FBS thicknesses and increase the likelihood of fatigue cracking. It is apparent that if the
process to determine FBS design moduli is changed, there needs to be an associated change to the fatigue
relationship.
8.5.3 Maximum Design Modulus
The interim method includes a maximum FBS design modulus of 2500 MPa based on TMR practice.
No change to this modulus limit is proposed as it reduces the risk that high modulus less flexible mixes will
be designed and used. When there is an improved understanding of the material properties that influence
FBS fatigue, this maximum modulus value should be reviewed.
8.6 Fatigue Relationship

As discussed in Section 8.3.3, the FBS fatigue relationship used in the interim method is the asphalt fatigue
relationship with the following inputs:
• FBS design moduli derived from cured wet moduli measured in the mix design
• a volume of bitumen calculated for the FBS mix.

Given that FBS mixes commonly contain lime or cement and that the foamed bitumen is distributed
differently from more uniformly coating of bitumen in asphalt, it seems unlikely that fatigue of FBS mixes
varies with FBS modulus and volume of bitumen as provided in the asphalt fatigue relationship. As FBS
mixes contain cementitious materials, flexural strength may be as influential a mix property as it is for
cemented materials. In addition, the presence of cementitious material may result in a higher strain damage
exponent than the value of 5 used for asphalt.
Austroads research project TT2046 Improved Cost-effectiveness of Foamed Bitumen Stabilised Pavements
includes an objective to improve methods to assess the FBS fatigue. It is proposed that changes to the
fatigue relationship be deferred until the findings of this research can be evaluated.
8.7 Proposed Amendment to Interim Method

Based on the limited performance data, it was concluded that use of the current very conservative FBS
design moduli together with the asphalt fatigue relationship appears to provide fatigue life predictions in
reasonable agreement with observed performance (Section 8.4.5). As discussed in Section 8.6, additional
fatigue research is required before major changes to the interim method can be made.
However, the existing method provides the following advice in relation to binder types and contents:
• The design method assumes the FBS mix includes sufficient quantity of residual bitumen (and secondary
stabilising agent) to produce a bound layer with significant tensile strength and hence fatigue properties.
• The FBS mix does not include an amount or type of secondary binder such that it is overly susceptible to
fatigue.
• Usually a minimum 3% residual bitumen with lime or cement as secondary binders is required to produce
a foamed bitumen stabilised layer that meets the above two criteria.
Since the publication of GPT Part 5 (Austroads 2011a), there has been an increase in the use of bitumen
contents in the range 2.5% to 3%. Accordingly, the minimum 3% bitumen needs to be amended. There is
also recognition that advice on suitable binder combinations should vary depending on whether the material
to be FBS treated has previously been stabilised with cementitious binders.
The scope of the thickness design method is proposed to be limited to mixes with the following binders:
• Where 50% or more of host materials is material previously cement stabilised, the binders shall be limited
as follows
– a minimum residual bitumen content of 3%
– a maximum hydrated lime content of 1.5% or maximum 1% quicklime/cement.
• Where less than 50% of host material comprises material previously cement stabilised, the binders shall
be limited as follows
– a minimum residual bitumen content of 2.5%
– a maximum hydrated lime content of 2%, maximum 1.5% quicklime/cement.
8.8 Summary
The fatigue life predictions using the interim method (Austroads 2011a) appear reasonably consistent with
the observed performance of three field trials.
The method of determining the FBS design modulus could be improved as the moduli appear overly
conservatism as the values are consistent with those early-life (1–3 months) cores without allowance for the
curing that occurs in situ over the first year.

For pavements not subject to inundation, the use of soaked moduli adds an additional degree of
conservatism in the modulus determination. Consequently, the design modulus procedure could be improved
to more closely align with in situ values. However, if the design moduli are increased without change to the
fatigue relationship, the design method will tend to overestimate FBS fatigue life.
Whilst the existing fatigue relationship could be adjusted so that it predicts similar fatigue lives using higher
FBS design moduli more representative of in situ values, it is proposed to defer such changes until there is
an increased understanding on the FBS mix properties that influence fatigue performance.
As this information is not likely to be available for several years, it is proposed that an amendment be issued
to the interim method to improve guidance on binder contents to the interim method that now applies.

9. Summary and Conclusions

Improvements are required to the current Austroads interim mix design and structural design methods for
FBS pavements. This report details progress to address this under Austroads project TT1825 Mix Design
and Field Evaluation of Foamed Bitumen Stabilised Pavements.
9.1 Mix Design
9.1.1 Test Methods
A key objective of the project was to improve and harmonise national mix design procedures.
Under the guidance of a project Working Group, the mix design framework was agreed. It was agreed to
build on the past experience, principally that of the Queensland Department of Transport and Main
Roads (TMR), and use the indirect tensile modulus as the principal pavement performance measure in the
mix design process. The following test methods were developed:
• AGPT-T301 Determination of foaming properties of bitumen
• AGPT-T302 Mixing of foamed bitumen stabilised materials
• AGPT-T303 Compaction of test cylinders of foamed bitumen stabilised mixtures: Part 1 Dynamic
compaction using Marshall drop hammer
• AGPT-T305 Resilient modulus of foamed bitumen stabilised mixtures.
At the commencement of the research project there was a wide range of methods being used to compact
modulus test cylinders. Consequently, the measured moduli varied between laboratories. There was a need
to harmonise on the compaction method. The Working Group decided to adopt a dynamic compaction
method based on the use of the Marshall hammer as developed for testing asphalt specimens. Consistent
with the TMR test method, mixes were compacted into 152 mm diameter moulds rather than varying the
mould size and hammer weight depending on the material size. In addition, it was agreed that the Austroads
mix design method should be based on test specimens compacted using a 10.2 kg Marshall drop hammer.
9.1.2 Laboratory Characterisation
The findings in relation to laboratory characterisation were:

• The mix design moduli are well below the cured in situ values at the same density. A likely reason for this
is the very short period (3 days) of curing time used in the mix design. This is an issue warranting further
investigation.
• The dry density of laboratory test specimens has a major influence on the indirect tensile moduli. Pooling
the results of various projects, it is estimated that the modulus increases by 12% for a 1% increase in dry
density.
• The dry density of the mix design specimens can differ significantly from in service densities. Given the
high dependence of modulus on density, the process for determination of FBS modulus for use in
thickness design needs to include an adjustment of the modulus results from the density of the laboratory
test specimens to representative in situ values. As discussed in Section 9.3, the provision of this
adjustment has been deferred pending improved methods of fatigue characterisation.

• A complicating factor in providing a modulus adjustment process for density is the varying methods of
density measurement. Preliminary testing of laboratory-compacted specimens suggests the presaturation
test method, which excludes surface voids, results in dry densities that are 2–3% above the mensuration
method. Both these methods are currently being used across Australia. It is important that a consistent
approach to density measurement be used in the mix design and field compaction assessment. If it is
proposed to test field cores after construction to evaluate modulus and density compliance, the
presaturation method is recommended as the mensuration underestimates the density due to the
presence of surface voids.
• The moisture content at which materials are mixed with foamed bitumen and compacted for modulus
testing varies between laboratories and this may lead to inconsistencies in mix designs. For instance,
testing undertaken for the Newell Highway and Calder Freeway projects indicates differences in moisture
content of 0.5% can occur. Such a difference in moisture content has a significant influence on the test
specimen density and measured modulus. This aspect of the mix design process is yet to be harmonised.
Inconsistencies in mix acceptance due to variations in mixing moisture would be reduced if FBS mix
acceptance included an adjustment of the modulus results from the density of the laboratory test
specimens to representative in situ values.
• Differences in mix moduli of 100 mm diameter and 150 mm diameter test specimens were investigated.
The results suggest that when samples are compacted using gyratory compaction, the moduli of 150 mm
diameter specimens are significantly higher than those of 100 mm diameter specimens. This reinforces
the Working Group decision to use a single-size specimen (150 mm) in the Austroads mix design
process.
• Specimens compacted by gyratory compaction had significantly lower moduli than those compacted
using the heavy Marshall hammer. To provide more consistency in mix design assessments by
laboratories, the Working Group agreed to adopt the heavy Marshall hammer compaction method in the
Austroads mix design method. It was agreed that research should continue to relate the moduli of
gyratory and Marshall specimens as this may enable a future change to the gyratory method. However,
modification to the gyratory compaction method is required to achieve higher moduli in closer agreement
to in situ values.
9.1.3 Mix Design Moduli Compared to In Situ Values
Field cores from untrafficked areas of the trial sites extracted a year or more after construction had high
moduli (> 5000 MPa) when the FBS was adequately compacted. For some projects, field core indirect tensile
modulus values in excess of 10 000 MPa were measured. There was no indication from the mix design
results of such modulus values, possibly due to the very limited period of curing (3 days) of the laboratory
specimens.
It was beyond the scope of this project to investigate whether the mix design process results in optimum mix
performance, particularly in relation to load-induced and environmental cracking susceptibility. There may be
a need to extend the curing period in the mix design period and include a measure of mix brittleness. This
issue warrants detailed research.
As discussed in Section 3.2, the mix design process would be improved by testing the FBS mixes using a
range of moisture contents.
9.1.4 Compliance Testing of Field-mixed Materials
In some projects, loose field-mixed materials were sampled to assess compliance with mix design modulus
criteria.
From limited testing, it was concluded that time delays between field mixing and laboratory compaction can
result in lower test specimen densities, and hence lower moduli than laboratory-mixed and compacted
specimens. In addition, the time delay may reduce the moduli over and above that due to the reduction in
density. To address this issue, TMR require the compaction of the field-mixed material in the laboratory to
commence within three hours of field mixing.

In addition, the moisture content of the field-mixed material most likely will differ from that used in the mix
design: commonly the field-mixed materials are drier due to the higher compactive effort used in construction
compared to that used in the mix design process. As a result, the densities of the field-mixed specimens can
differ from that of the mix design specimens with an associated effect on modulus. In such cases, the
compliance assessment needs to consider adjustment of the modulus results for differences in specimen
densities.
9.2 Distress Types

One of the key objectives of the project was to gain an improved understanding of the distress modes of FBS
pavements. Of particular interest was whether fatigue cracking of the FBS is the predominant distress type
as assumed in the Austroads interim structural design method. Consequently, most of the trial pavements
were intentionally under-designed in terms of FBS thickness so that their fatigue cracking characteristics
could be observed within the four-year research project duration.
In terms of the observed performance of the six trial sections, cracking of the FBS layer was the primary
distress mode. It was concluded for the FBS mixes currently designed and constructed in Australia, the
Austroads structural thickness design method should consider load-induced fatigue cracking.
Rutting did not generally increase significantly after the first 24 hours of opening to traffic and there is no
indication that the trial pavements will need shape correction before treatment for cracking. In general, rutting
and roughness only increased after severe cracking developed. However, two short lengths of trial pavement
needed to be removed and replaced due to rutting:
• The Port Wakefield Road site included a 30 m long section where unknowingly an asphalt patch was FBS
treated and rutted quickly during the first summer (Appendix C.6). This reinforces the current construction
practice and removing asphalt patches prior to FBS, and prior to stabilisation possibly after spreading all
or part of the asphalt millings along the project.
• The Calder Freeway trial was opened to traffic within 2 hours of completion and prior to resurfacing. By
the next day, the surface had ravelled (Appendix D.7.2) and a 10 m length at the start of the project
needed to be removed and patched with asphalt. For moderate-to- heavily-trafficked roads, this suggests
that it is good practice to at least cure the FBS overnight and apply the surfacing before opening it to
traffic. Proof-rolling can also provide guidance on the timing of opening to traffic.
At the Calder Freeway and Western Freeway trial sites, flushing of the bitumen or embedment of the sprayed
sealing aggregate was observed. Advice was subsequently sought from experienced TMR staff on the
design of sprayed seal surfacing. Their advice was to avoid the use of cutting oil in the bitumen as this has
the potential to soften the bitumen within the FBS mix. In addition, TMR recommends the use of an armour
coat seal prior to the double application seal as a further measure to inhibit embedment.
9.3 Austroads Structural Thickness Design Method

The interim Austroads structural thickness design method for FBS pavements is described in the Guide to
Pavement Technology Part 5: Pavement Evaluation and Treatment Design (Austroads 2011a).
Another key objective of the project was to improve the interim Austroads procedures for the thickness
design of FBS materials for new pavements and structural rehabilitation treatments. This
mechanistic-empirical procedure includes prediction of FBS fatigue characteristics by adapting the existing
hot-mix asphalt fatigue relationship.

Based on the performance of the trial sections and the high modulus FBS mixes currently being used in
Australia, the Austroads structural thickness design method should continue to consider the following two
distress types:
• fatigue cracking of the FBS layer determined using the predicted horizontal tensile strain at the bottom of
the FBS layer, and
• rutting and shape loss structure determined using the vertical compressive strain on the top of the
subgrade.
The following three significant deficiencies were identified in the interim method:
• The FBS design moduli are currently determined from the measured cured soaked indirect tensile
modulus in the laboratory-mix design. Based on the cores extracted from the trial pavements, the design
moduli being used (1000–2500 MPa) are well below in situ values, except for the initial moduli soon after
opening to traffic. There is potential to increase the FBS design moduli to better reflect the values after
curing in the roadbed. Such an improved process would include a procedure to adjust for the mix design
moduli to in situ densities.
• The current method for FBS design modulus determination utilises an adjustment between indirect tensile
modulus and flexural modulus based on limited asphalt testing. That is, the indirect tensile modulus of
asphalt using a rise time of 40 milliseconds is similar to flexural modulus at 14.8 Hz. There is a need to
determine a procedure for FBS mixes.
• The fatigue relationship used for FBS is an adaptation of that used for asphalt fatigue. Given that FBS
mixes commonly contain lime or cement and that the foamed bitumen is distributed differently from the
more uniform coating of bitumen in asphalt, it seems unlikely that fatigue of FBS mixes varies with FBS
modulus and volume of bitumen as provided in the asphalt fatigue relationship. Consequently, FBS mixes
in use may not be optimum in terms of fatigue performance.
Despite these deficiencies in the interim design method, the fatigue life predictions appear reasonably
consistent with the observed performance based on the limited information available from field trials. The use
of very conservative FBS design moduli with the asphalt fatigue relationship, together with empirical
guidance on particle size distribution, plasticity and maximum binder contents, appears to result in suitable
pavement structures based on the limited performance information available.
The project has provided sufficient information to improve the method of determining FBS design moduli.
However, if the FBS design moduli were increased to reflect field-cured values without change to the fatigue
relationship, the design method will tend to overestimate the FBS fatigue life. Accordingly, it is proposed to
defer such changes to the design modulus until an improved fatigue relationship is developed. As this
relationship is not likely to be available for several years, it is recommended that an amendment be issued to
the interim design method to improve guidance on suitable binder contents.

References
Asphalt Academy 2009, Technical guidelines for bitumen stabilised materials. A guideline for the design and
construction of bitumen emulsion and foamed bitumen stabilised materials, TG2, 2nd edn, Asphalt
Academy Pretoria, South Africa.
Austroads 2005, Asphalt binder content (ignition oven method), test method AGPT-T234-05, Austroads,
Sydney, NSW.
Austroads 2006, Guide to pavement technology part 4D: stabilised materials, AGPT04D-06, Austroads,
Sydney, NSW.
Austroads 2008, A review of Austroads gyratory compaction research, AP-T94-08, Austroads, Sydney,
NSW.
Austroads 2009, Guide to pavement technology part 4K: seals, AGPT4K-09, Austroads, Sydney, NSW.
Austroads 2011a, Guide to pavement technology part 5: pavement evaluation and treatment design, 3rd edn,
AGPT05-11, Austroads, Sydney, NSW.
Austroads 2011b, Review of foamed bitumen stabilisation mix design methods, AP-T178-11, Austroads,
Sydney, NSW.
Austroads 2011c, Review of structural design procedures for foamed bitumen pavements, AP-T188-11,
Austroads, Sydney, NSW.
Austroads 2012, Guide to pavement technology part 2: pavement structural design, 3rd edn, AGPT02-12,
Austroads 2013, Improved design of foamed bituminous stabilised pavements, AP-T226-13, Austroads,
Sydney, NSW.
Austroads 2014, Design and performance of foamed bitumen stabilised pavements: progress report 2, AP-
T275-14, Austroads, Sydney, NSW.
Austroads 2015, Design and performance of foamed bitumen stabilised pavements: progress report 3, AP-
T303-15, Austroads, Sydney, NSW.
Bowering, RH 1973, ‘Soil stabilisation with foamed bitumen’, Australian Road Research Board (ARRB)
regional symposium, 6th, 1971, Swan Hill, Victoria, pp. 35-46.
Bowering, RH & Martin, CL 1976, ‘Foamed bitumen production and application of mixtures evaluation and
performance of pavements’, Journal of the Association of Asphalt Paving Technologists, vol. 45, pp. 453-
77.
Department of Planning, Transport and Infrastructure 2007, Construction of foamed bitumen stabilised
pavement, specification part R24, DPTI, Adelaide, SA.
Department of Planning, Transport and Infrastructure 2011, ‘Traffic counts for Port Augusta: Port
Wakefield/Port Wakefield Rd Site 6037 location 1.7 km south of RN 5000 (Angle Vale Rd)’, Spatial
Intelligence and Road Assets Section, DPTI, Adelaide, SA.
Fu, P & Harvey, J 2007, Temperature sensitivity of foamed bitumen stiffness: field and lab study. Internal
Journal of Pavement Engineering, Vol. 8, No. 2, pp 137-45.

Gipenz, CP 2014, ‘Performance of foamed bitumen stabilised pavements on New Zealand state highways’,
version 3.3, unpublished report, Fulton Hogan, NZ.
Jones, JD & Ramanujam, J 2004, ‘Rehabilitation of unbound pavements using foamed bitumen stabilisation’,
International symposium on unbound granular aggregates in roads, 6th Nottingham, United Kingdom,
Balkema, Leiden, Netherlands, pp. 301-10.
Jones, JD & Ramanujam, J 2008, ‘Design of foamed bitumen stabilised pavements’, internal report, Road
System and Engineering, Road and Delivery Performance Division, Department of Transport and Main
Roads, Brisbane, QLD.
Lancaster, J, Mc Arthur, L & Warwick, R 1994, ‘VicRoads experience with foamed bitumen stabilisation’,
Australian Road Research Board Ltd (ARRB) conference, 17th, 1994, Gold Coast, Queensland, Australia,
pp. 193-211.
Leek, C 2001, An investigation of the performance properties of in situ foamed bitumen stabilised
pavements, Institute of Public Works Engineering Australia, viewed 30 May 2017,
<http://www.auststab.com.au/foamedbitumen/FBPaperIPWEA2001.pdf>.
Leek, C 2010, Review of the performance of insitu foamed bitumen stabilised pavements in the City of
Canning, Austab, viewed 30 May 2017,
<http://www.auststab.com.au/pdf/design_&_specification/review_of_performance_of_insitu_foamed_bitu
men_pavements_in_canning.pdf>.
Maccarrone, S, Holleran, G, Leonard, DJ & Hey, S 1994, ‘Pavement recycling using foamed bitumen’,
Australian Road Research Board Ltd (ARRB) conference, 17th, 1994, Gold Coast, Queensland, Australia,
vol. 17, no. 3, pp. 349-65.
Main Roads Western Australia 2010, ‘Construction specification foamed bitumen stabilised trial Kwinana
freeway’, MRWA, Perth, WA.
Main Roads Western Australia 2013, Procedure for the design of road pavements, engineering road note no.
9, MRWA, Perth, WA.
Moffatt, MA & Jameson, GW 1998, ‘Asphalt design modulus analysis’, contract report RE7133, ARRB
Transport Research, Vermont South, Vic.
Molenaar, A & Pu, B 2008, ‘Prediction of fatigue cracking in cement treated base courses’ Rilem
International conference on cracking in pavements, 6th, 2008, Chicago, Illinois, USA, CRC Press, UK.
Queensland Department of Transport and Main Roads 2012, Pavement rehabilitation manual, TMR,
Brisbane, QLD.
Queensland Department of Transport and Main Roads 2014, Materials testing manual: part 4: bitumen
content of stabilised material, test method Q118, TMR, Brisbane, QLD.
Queensland Department of Transport and Main Roads 2017a, Insitu stabilised pavements using foamed
bitumen, technical specification MRTS07C, TMR, Brisbane, QLD.
Queensland Department of Transport and Main Roads 2017b, Testing of materials for foamed bitumen
stabilisation, technical note 150, TMR, Brisbane, QLD.
Queensland Department of Transport and Main Roads 2017c, Q138 Preparation and Testing of Foamed
Bitumen Materials, January, Brisbane, QLD.
Queensland Department of Transport and Main Roads 2017d, Q139 Resilient Modulus of Stabilised
Materials (indirect tensile method), January Brisbane, QLD.

Ramanujam, J & Jones, JD 2000, ‘Characterisation of foamed bitumen stabilisation’, Road System and
Engineering Technology Forum, Department of Transport and Main Roads, Brisbane, QLD.
Ramanujam, J & Jones, JD 2007, ‘Characterisation of foamed bitumen stabilisation’. Internal Journal of
Pavement Engineering, Volume 8 No., 111-122, published by Taylor and Francis Group.
Roads and Maritime Services 2012a, ‘Preliminary design for rehabilitation of the Newell Highway near
Tycannah Creek’, draft report P1376/02, RMS, Sydney, NSW.
Roads and Maritime Services 2012b, Dry density/moisture relationship of road construction materials, test
method T111, RMS, Sydney, NSW.
Roads and Maritime Services 2012c, Moisture content of road construction materials (standard method), test
Roads and Maritime Services 2012d, Field wet density of road construction materials (nuclear gauge in
direct transmission method), test method T173, RMS, Sydney, NSW.
Roads and Maritime Services 2012e, Compaction of test specimens of asphalt: Marshall procedure, test
VicRoads 2013, Selection and design of pavements and surfacings, code of practice RC 500.22, VicRoads,
Kew, Vic.
VicRoads 2014a, Hot mix asphalt, standard specification 407, VicRoads, Kew, Vic.
VicRoads 2014b, Registration of bituminous mixes, code of practice RC 500.01, VicRoads, Kew, Vic.
Zhang, J 2013, ‘Pavement investigation for foamed bitumen stabilisation of Bellata stage 2, (ch 46.363km –
45.922km)’, unpublished research project. RMS, Sydney, NSW.
Superseded Austroads publications
Austroads 1992, Pavement design: a guide to the structural design of road pavements, AP-17-92,
ASTM Standards
C25-11e1-2011, Standard test methods for chemical analysis of limestone, quicklime, and hydrated lime.
C642-13-2013, Standard test method for density, absorption and voids in hardened concrete.
D5581 07a 2013, Standard Test Method for Resistance to Plastic Flow of Bituminous Mixtures Using
Marshall Apparatus (6-inch-diameter specimen).
Australian Standards
AS 1289.3.1.2-2009, Methods of testing soils for engineering purposes: soil classification tests:
determination of the liquid limit of a soil: one point Casagrande method (subsidiary method).
AS 1289.3.2.1-2009, Methods of testing soils for engineering purposes: soil classification tests:
determination of the plastic limit of a soil: standard method.
AS 1289.3.3.1-2009, Methods of testing soils for engineering purposes: soil classification tests: calculation of
the plasticity index of a soil.
AS/NZS 2891.3.3:2013, Methods of sampling and testing asphalt: binder content and aggregate grading:
pressure filter method.

AS/NZS 2891.9.2:2014, Methods of sampling and testing asphalt: determination of bulk density of
compacted asphalt: presaturation method.
AS/NZS 2891.9.3:2014, Methods of sampling and testing asphalt: determination of bulk density of
compacted asphalt: mensuration method.
AS/NZS 2891.13.2:2013, Methods of sampling and testing asphalt: determination of the resilient modulus of
asphalt: indirect tensile method.

Appendix A Kwinana Freeway, Baldivis
A.1 Introduction
In February 2010, trial sections of FBS pavements were constructed for Main Roads Western
Australia (MRWA) on the Kwinana Freeway, Baldivis about 45 km south of Perth. Over the 7-year period
until February 2017, the annual average rainfall was about 640 mm, compared to the Baldivis long-term
annual average rainfall of about 745 mm.
Three pavement sections, each 100 m long and 3.8 m wide, were constructed with different FBS thicknesses
(namely 150, 240 and 290 mm) and resurfaced with 30 mm dense graded asphalt (DGA), a geofabric
reinforced seal and 30 mm open graded asphalt (OGA). Figure A 1 shows a view of the test site during
construction.
Only the performance of the section with the target FBS thickness of 150 mm was monitored in this project.
Project details are summarised in Table A 1.
Figure A 1: Kwinana Freeway FBS site

Table A 1: Description of Kwinana Freeway FBS pavement
Description
Location Kwinana Freeway (north of Mundijong Road between 37.72–37.82 SLK) northbound
Job size Length 100 m
Number of lanes 1 (northbound carriageway, outer lane)
Lane width 3.5 m lane width with 0.3 m wide shoulder
Stabilisation depth Nominal 150 mm FBS base
Foamed Mix design by MRWA (empirically-based method)
stabilisation
Host materials Mixture of:
specification
hydrated cement treated crushed rock (85%) and
crushed limestone subbase (15%)
Supplementary binder 0.8% quicklime
Bitumen 3% Class 170 bitumen
Foaming agent Unknown
Construction Work specifications and QA MRWA (2010)
method testing
Construction date February 2010
Surfacing 30 mm thickness open graded asphalt
geofabric reinforced seal
30 mm thickness of dense graded asphalt using polymer modified binder A15E
A.2 Site Investigation Prior to Stabilisation

Site investigation prior to stabilisation showed that the existing pavement had the following composition:
• 30 mm open graded asphalt (OGA)
• 30 mm dense graded asphalt (DGA)
• 10 mm sprayed seal
• 125 mm hydrated cement treated crushed rock base (HCTCRB)
• 225 mm crushed limestone subbase
• sand subgrade.
HCTCRB is crushed rock base that has been mixed with water and 2% cement, and then allowed to hydrate
in the stockpile for an extended period of time before wet mixing again, placing and compacting in the
roadbed.
A.3 Mix Design

A large bulk sample of each pavement layer was excavated using a 1 m wide profiler from the pavement at
the south end of the pavement section at SLK 38.015 prior to stabilisation. The OGA and DGA asphalt layers
were removed by the profiler. Laboratory particle size distribution test results for these materials are
summarised in Table A 2. The grading is slightly on the coarse side of the recommended envelope.

As explained in Section A.5, as the pavement was initially stabilised to a nominal depth of 210 mm, the host
material to be stabilised comprised:
• 30 mm open graded asphalt
• 30 mm dense graded asphalt
• 125 mm HCTCRB
• 25 mm crushed limestone subbase.
For this pavement section, MRWA specified that the host material be stabilised using 3% residual bitumen
and 0.8% quicklime. These binder contents were empirically based rather than determined from by mix
design.
Table A 2: Particle size distribution of the host material for Kwinana Freeway FBS pavement
Sieve size Percentage passing

(mm) Measured Recommended envelope(1)
75 100 100
26.5 99
19 99 87–100
13.2 94
9.5 85 67–88
4.75 62 50–65
2.36 44 38–50
1.18 34 26–38
0.600 26 18–30
0.425 21 16–26
0.300 17 14–24
0.150 11 11–19
0.075 7 8–16
1 Proposed for next edition of Austroads Guide to Pavement Technology Part 4D: Stabilised Materials.
A.4 FBS Thickness and Predicted Allowable Traffic Loading

A target 150 mm FBS thickness was selected to increase the likelihood of observing FBS fatigue cracking in
accordance with the Austroads project objective.
Table A 3 shows the final pavement structure and the selected design modulus values used in the interim
FBS thickness design procedure (Austroads 2011a) to calculate the allowable traffic loading of the
under-designed pavement. Mix design moduli are not available to determine FBS design modulus. An FBS
design modulus value of 2500 MPa was assumed, this being the maximum FBS design modulus in the
interim method.
Based on this data set, the predicted FBS fatigue life of 4.5 x 106 ESA at 50% design reliability was
calculated using:
• the interim Austroads method including the Equation 4 FBS fatigue relationship modified to 50% reliability
as discussed in Section 8.4.3
• SAR5/ESA ratio of 1.2 (MRWA 2013).

Table A 3: Data used to predict FBS life of the under-designed Kwinana Freeway pavement
Pavement
Design parameter Available data Technical basis for data selection
component
Subgrade Subgrade type Sand, design CBR 12% Presumptive design value for subgrade sand in WA
Subgrade design 120 MPa
modulus
Subbase Subbase thickness 200 mm Nominal thickness crushed rock base/subbase
remaining considering original design
(Appendix A.2)
Design modulus of 210 MPa (top sublayer) Table 6.5 of Austroads (2012)
subbase
FBS base FBS thickness 150 mm Based on measured values (Table A 4)
FBS design modulus 2500 MPa Nominal value corresponding to the maximum
design modulus in interim Austroads method
(Austroads 2011a)
Volume of bitumen 6% Presumptive value based on nominal 3% bitumen
by mass
Surfacing Thickness 30 mm open graded Design thicknesses
asphalt
30 mm dense graded
asphalt
Design modulus 800 MPa (OGA) Engineering Road Note 9 (MRWA 2013) for a thick
bound structure
1800 MPa Using Table 6.13 of Austroads (2012) a presumptive
(size 10 mm DGA with indirect tensile modulus of 3500 MPa for a size 10
A15E binder) mm asphalt with Class 170 bitumen was adopted.
Using Table 6.12 of Austroads (2012), the
equivalent value for mix with A15E polymer modified
binder was calculated to be 2625 MPa. This value
was then adjusted to a weighted mean annual
temperature of 29 ºC and design speed of 80 km/h
A.5 Pavement Construction

The pavement was constructed in February 2010 in accordance with Construction Specification Foamed
Bitumen Stabilised Trial Kwinana Freeway (MRWA 2010).
In planning the Kwinana Freeway trial sections, it was proposed to open the unsealed FBS base to traffic to
observe the early-life ravelling and rutting prior to reshaping and placement of the asphalt surfacing.
Consequently, the initial plan was to prepare the unsealed FBS to the level of the adjacent traffic lane, open
it to traffic, and several days later mill the top 60 mm of the FBS to provide for the 60 mm asphalt surfacing.
Although this early-life performance evaluation was not implemented, the sequence of the works program
aligned with the initial plan:
• The 30 mm thick open graded asphalt, 30 mm dense graded asphalt, 125 mm HCTCRB and 25 mm
crushed limestone subbase was pulverised, mixed, and compacted to provide 210 mm thickness of the
FBS base. Note that due to concerns about the breakdown of the crushed limestone subbase due to the
stabiliser, the lime and bitumen were incorporated into the pavement in the first pass of the stabiliser to
minimise the number of mixing passes. Nevertheless, a second mixing pass was required due to
inadequate depth of mixing in the first pass.
• Several days later, the top 60 mm of FBS base layer was removed by milling and brooming.
• A bitumen tack coat was placed, followed by a 30 mm thick, size 10 mm dense graded asphalt made
using A15 E polymer modified binder.

• A geotextile reinforced seal (GRS) was then placed on the asphalt to inhibit surface cracking and provide
a waterproof layer. Below the geotextile a bitumen bond coat was at 0.9 l/m2. On top of the geotextile,
Class 170 bitumen without cutter oil was sprayed at a rate of 1.6 l/m2. A size 10 mm sealing aggregate
was then spread on the bitumen.
• A 30 mm thick open graded asphalt layer was placed on the GRS.
Detailed specifications for the stabiliser, compaction equipment and construction tolerances were:
• A Wirtgen WR 2000 stabiliser was used and the construction tolerance for the application rate for the
binder and supplementary binder was ±10% of the specified values.
• Compaction equipment used comprised: (a) a vibrating padfoot roller of a minimum mass of 18 t; (b) a
vibrating smooth drum roller with a minimum mass of 12 t; and (c) a multi-tyre roller with a minimum mass
of 15 t. The stabilised pavement layers were compacted uniformly to the full depth and over the full width
to a characteristic dry density ratio of 98% (relative to modified Proctor MDD) or greater.
Quality assurance testing for layer thickness and density was carried out at three locations and the results of
layer thickness and density are given in Table A 4. The results confirmed that the minimum compaction
standard was achieved.
Table A 4: Pavement thicknesses and densities after construction
Nuclear density meter testing

Thickness (mm)
Test location SLK on the FBS base
(km) Wearing course Limestone Dry density ratio(1) Moisture ratio(1)
and FBS base subbase (%) (%)
37.750 210 210 105.7 55
37.775 240 170 103.3 53
37.800 220 220 105.3 51
Average 223 200 104.8 53
1 Based on modified Proctor MDD = 2.081 t/m3 and modified OMC = 8.2%.
A.6 Condition Monitoring
Visual Condition
Since construction, MRWA has regularly inspected the pavement and has reported no surface defects to
date. Figure A 2 shows the good condition of the pavement after six years of trafficking.
Figure A 2: Kwinana Freeway condition May 2015
Looking north at SLK 37.9 km Looking south at SLK 37.8 km

Source: Personal communication by email from Simon Kenworthy-Groen of MRWA on 25 May 2015.
Rutting
In December 2014, after almost five years of trafficking, the rutting was measured with a laser profiler. The
average rut depths over the 20 m length in each wheel path are shown in Figure A 3. The pavement is
performing well with very low rutting.
Figure A 3: Rutting after five years trafficking
10
8 OWP
IWP
7
Average rut depth

over 20 m 5
(mm)
4
0
37.72 37.74 37.76 37.78 37.8 37.82 37.84 37.86 37.88 37.9 37.92 37.94 37.96 37.98 38 38.02
Kwinina Freeway northbound outer lane SLK
Surface Deflections
MRWA conducted falling weigh deflectometer (FWD) surface deflection measurements at four months,
14 months, 26 months, 33 months, 39 months, 51 months, 58 months and 73 months after construction. The
deflection bowls were measured using a contact stress in the range 600–700 kPa, then normalised to
566 kPa. The surface temperatures during the FWD measurements varied significantly from 11 °C to 30 °C.
Using the surface temperatures, MRWA adjusted the deflections and curvatures from the measurement
temperature to the WMAPT for Perth (29 °C).
The temperature-corrected maximum deflections (D0) and curvatures (D0–D200) normalised to a stress of
566 kPa are plotted in Figure A 4. There is a no suggestion from the results that the deflections and
curvatures are increasing, consistent with fatigue damage to the foamed bitumen stabilised layer. Further
monitoring is required to assess whether this is correct.

Figure A 4: FWD testing results for Kwinana Freeway FBS site

A.7 Traffic Monitoring

Weekly traffic data on the northbound lane of Kwinana Freeway just south of Mundijong Rd Bridge for the
period 2006–13 are given in Table A 5.
Table A 5: Kwinana Freeway northbound carriageway daily heavy vehicle volumes
Total Austroads vehicle classification

heavy
Survey date
vehicles 3 4 5 6 7 8 9 10 11 12
per day
07 July – 13 July 2006 1119 414 113 59 25 183 40 200 27 38 20
19 April 2007 – 22 April 2007 1118 503 90 54 24 201 36 153 28 18 11
23 June 2010 – 2 July 2010 2013 1105 142 27 37 58 52 356 94 140 2
10 August 2011 – 17 August 2011 2134 1323 143 23 72 67 118 196 69 118 5
1 February 2013 – 4 February 2013 2705 1597 163 39 53 86 74 374 149 169 1
Source: Personal communications by emails from Simon Kenworthy-Groen of MRWA on 31 May 2013 and 25 May 2015.
Table 12 of Engineering Road Note 9 (MRWA 2013) provides the average ESA of pavement loading for each
Austroads vehicle class. Table A 6 lists the values recommended for use for this project.
Table A 6: Kwinana Freeway average ESA per heavy vehicle type
Austroads vehicle classification

3 4 5 6 7 8 9 10 11 12
0.49 2.63 2.8 0.68 1.49 3.73 4.69 6.79 8.88 11.54
Source: MRWA (2013, Table 12 Kwinana Freeway (H015) SLK56.84, Mandurah).
The annual ESA of loading were calculated based on:

• annual heavy vehicle volumes in the slow lane calculated from the average daily values (Table A 5) and
assuming 80% of heavy vehicles travel in the slow lane
• the ESA of damage per heavy vehicle type (Table A 6).
Using this data, the predicted cumulative traffic loading over the 7-year period from pavement construction in
February 2010 until February 2017 of 1.1 x 107 ESA was calculated (Figure A 5), assuming a heavy vehicle
growth rate of 4% after 2010.

Figure A 5: Predicted cumulative traffic loading

1.4E+07
1.3E+07
1.2E+07
1.1E+07
1.0E+07
9.0E+06
Cumulative 8.0E+06
traffic 7.0E+06
loading
(ESA) 6.0E+06
5.0E+06
4.0E+06
3.0E+06
2.0E+06
1.0E+06
0.0E+00
0 1 2 3 4 5 6 7 8
Years of trafficking
A.8 Performance Review

Using the cumulative traffic loadings given in Table A 5, it would take about 3–4 years of loading to reach the
predicted fatigue life of 4.5 x 106 ESA (Appendix A.4).
Currently, there is no surface distress after 1.1 x 107 ESA of traffic loading over 7 years: that is, no distress
after twice the predicted fatigue life.
Given the provision of a geotextile seal under the open graded asphalt surfacing, it is unlikely that cracking
will be observed by the completion of the Austroads research project TT2046 Improving the
Cost-effectiveness of Foamed Bitumen Stabilised Pavements (mid 2019). Therefore, reliance will be placed
on the change in measured surface deflections as an indicator of structural deterioration. As mentioned in
Appendix A.6, the most recent deflections are not consistent with the hypothesis that the FBS is being
damaged by the traffic. Further monitoring is required to confirm this hypothesis.
A.9 Kwinana Freeway Trial Summary

In February 2010, Main Roads Western Australia constructed a 150 mm thick FBS base on the Kwinana
Freeway in Perth. The material that was stabilised included a pulverised mixture of 30 mm thick open graded
asphalt, 30 mm thick dense graded asphalt, an old sprayed seal, 125 mm thick hydrated cement treated
crushed rock base (HCTCRB) and 25 mm crushed limestone subbase. This mixture was stabilised with a
nominal 3% bitumen and 0.8% hydrated lime.
The asphalt surfacing comprised a 30 mm thick open graded asphalt, a geotextile reinforced seal (GRS) and
a 30 mm thick dense graded asphalt. This surfacing will inhibit possible FBS fatigue cracking reflecting
through to the surface.

After 7 years of trafficking, the findings are:

• The pavement is performing well, with no surface distress after 7 years of traffic loading.
• There is no suggestion from the surface deflection testing taken over the last two years that the
deflections and curvatures are increasing, consistent with fatigue damage to the FBS layer.
• Other than compaction and thickness of the FBS layer, no information is available on the FBS mix
characteristics such as actual bitumen and lime contents and mix design moduli.
• Coring of the pavement may provide some additional information if undertaken by the end of the project.
Without this information, the usefulness of the performance data will be limited.

Appendix B Kewdale Road, Welshpool
B.1 Introduction
In November 2011, the City of Canning constructed FBS pavement sections on Kewdale Road, Welshpool, a
suburb of Perth in Western Australia. Three pavement sections were constructed:
• a main section (100 m long and 4 m wide) constructed with 320 mm FBS thickness
• under-designed Section 1: 50 m length with 150 mm nominal FBS thickness
• under-designed Section 2: 50 m length with 100 mm nominal FBS thickness.
Only Sections 1 and 2 were selected for performance review in this project as the 320 mm thick section is
unlikely to be distressed during the project period.
Details of locality and job statistics (job size, FBS specification and construction method) for the Kewdale
Road FBS pavement are given in Table B 1.
The annual average rainfall from November 2011 to November 2016 was about 660 mm compared to
Perth’s long-term annual average rainfall of about 730 mm.
Table B 1: Description of Kewdale Road FBS pavement
Description
Location Kewdale Road Welshpool, Western Australia (southbound north of the Dowd Street intersection)
Job size Length 2 sections, each 50 m in length
Number of lanes 1 (southbound slow lane)
Lane width 4m
Total area 400 m2
Stabilisation depth Section 1: nominal 150 mm FBS, measured 115 mm
Section 2: nominal 100 mm FBS, measured 115 mm
Subbase Cement treated crushed granite
Wearing course 30 mm dense graded asphalt
Foamed Mix design Empirically-based method
stabilisation
Host materials 30% recycled asphalt and 70% crushed granite base
specification
Supplementary binder 0.8% quicklime
Bitumen 3.5% Class 170 bitumen
Foaming agent Unknown
Construction Work specifications and QA testing Under supervision of City of Canning
method
Construction date 13 November 2011
Surfacing 30 mm thickness of size 10 mm dense graded asphalt
Figure B 1 shows the site locality plan: Section 1 is located 200–250 m from the intersection with Dowd
Street and Section 2 is located 150–200 m from the intersection. Note that the actual constructed
thicknesses of Sections 1 and 2 were similar contrary to the design intent. In addition, recent coring
(Appendix B.8) indicates there is a cement treated crushed granite subbase under the FBS layer. This is
unexpected and again contrary to the design intent of inducing high tensile strains to the base of the FBS
layer.

Figure B 1: Location of Kewdale Road trial sites

B.2 Site Investigation Prior to Stabilisation

In October 2011, coring and sampling of the pavement materials were undertaken at five locations and
dynamic cone penetrometer (DCP) testing of the subgrade was also performed prior to stabilisation. The
results (Table B 2) indicated that the existing pavement structure prior to rehabilitation comprised:
• 50–60 mm dense graded asphalt
• 150–275 mm thickness of size 20 mm crushed granite base
• 130–230 mm thickness of size 40 mm gravel or crushed limestone subbase
• subgrade sand with in situ CBRs in the range of 15–50%.
Surprisingly, this pavement investigation did not find a cement treated crushed granite subbase later
detected in coring (Appendix B.8).
Table B 2: Results of coring and DCP testing on Kewdale Road FBS pavement prior to stabilisation
Section 1 Section 2
Chainage (m)
20 20 50 70 80
Offset (m) 1.5 2.0 1.5 1.5 2.0
Asphalt thickness (mm) 50 50 60 50 50
Crushed rock base thickness(1) (mm) 220 275 260 220 150
Crushed limestone/gravel subbase thickness (mm) 165 185 130 165 220
Total pavement depth (mm) 435 510 450 435 420
Subgrade CBR (%) 30 30 50 15 NR(2)
1 Crushed rock base thickness after 50 mm granular re-sheet.

2 Not recorded.
There were also some areas of asphalt patching in both sections.
Prior to stabilisation, a 45 mm granular resheet of size 20 mm crushed granite base was placed on Section 2
to raise its pavement profile (Figure B 2). This allowed the construction of the 100 mm thick FBS layer by
firstly stabilising the overlaid pavement to a depth of 150 mm, then trimming the surface by 45 mm to the
height of the adjacent traffic lane.
Figure B 2: Crushed granite base re-sheet of Section 2 before stabilisation

B.3 Mix Design

Samples of materials in Sections 1 and 2 were collected from the roadbed after pulverising the top 150 mm
layer with the stabiliser. Results of particle size distribution and plasticity index (PI) of these materials are
summarised in Table B 3 and Figure B 3. The host material lacked fines compared to the recommended
particle size distribution for moderate-to-heavily-trafficked roads.
Table B 3: Results of particle size distribution and PI for the untreated materials in Kewdale Road
pavement (prior to stabilisation)
Percentage passing PI
Size (mm)
19 16 13.2 9.5 6.7 4.75 2.36 1.18 0.600 0.425 0.300 0.150 0.075 (%)
Section 1 100 99 97 90 80 71 56 41 28 21 15 7 3 8
Section 2 100 98 91 75 62 52 38 28 20 16 12 6 2 4
Recommended(1) 87–100 67–88 50–65 38–50 26–38 18–30 16–26 14–24 11–19 8–16 ≤ 10
1 Proposed for next edition of Austroads Guide to Pavement Technology Part 4D: Stabilised Materials.
A mix design was not undertaken. Based on previous experience in the City of Canning, the host material
was stabilised with a target 3.5% residual bitumen and 0.8% quicklime.
Figure B 3: Particle size distribution of untreated materials in Kewdale Road FBS pavement
100
Section 1
90 Section 2
Proposed Lower limit
80 Proposed Upper Limit
70
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)

B.4 Predicted Allowable Traffic Loading

When the pavement was cored late in the project (Appendix B.8), a cement treated crushed granite subbase
was found at least for part of the trial site. The thickness of this layer could not be determined in the coring,
but is likely to be in excess of 100 mm.
Due to the support provided by this cement treated subbase, the tensile strains at the base of the FBS layer
are low and hence the pavement sections will be able to provide feedback on the Austroads thickness
design procedures in relation to fatigue of FBS mixes.
For these reasons, predicted allowable loadings using the interim Austroads method have not been
calculated for the Kewdale Road trial sections.
B.5 Pavement Construction

The sections were stabilised on 13 November 2011 and opened to traffic the same day. The preliminary
construction report (Austroads 2013) indicated that:
• A Wirtgen WR 2500 SK stabiliser was used and the construction tolerance for the application rates for the
bitumen and quicklime were ±10% of the specified values. The minimum stabilisation depth of 150 mm
was initially constructed for both Section 1 and Section 2. The 150 mm layer in Section 2 was then cut
back to the design 100 mm thickness during the compaction and shaping operation.
• Compaction equipment included: (a) an 18 tonne vibrating padfoot roller, (b) a 12 tonne vibrating smooth
drum roller; and (c) a 20 tonne multi-tyre roller. The specified minimum field compaction was
98% modified Proctor maximum dry density.
Figure B 4: Kewdale Road FBS site during stabilisation
The sections were opened to traffic prior to the placement of the 30 mm thick dense graded asphalt surfacing
several days later. The unsealed foamed bitumen stabilised surface ravelled under initial trafficking
(Figure B 5), as observed during the Calder Freeway construction (Appendix D.7.2).

Figure B 5: Ravelling of Section 1 due to trafficking before the asphalt surfacing was placed
B.6 Pavement Condition Monitoring

Since construction in November 2011, the City of Canning and Main Roads WA have monitored the
pavement condition.
In February 2013, about 15 months after construction, the test pavements were inspected, and rutting was
measured on the 30 mm thick asphalt surfacing every 5 m along the two sections. There was minimal rutting,
with all measurements less than 5 mm depth. The measurements were repeated in May 2014, April 2015,
December 2015 and November 2016 about 2.5 years, 3.5 years, 4 years and 5 years after construction.
There were no significant changes to the rut depths.
There was also no fatigue cracking in either section and they were in sound condition (Figure B 6). During
the coring in April 2017 (Appendix B.8), two very fine transverse shrinkage cracks were observed in each
section. As discussed in Appendix B.8, this cracking may relate to a cementitious subbase rather than the
FBS base.

Figure B 6: Kewdale Road, April 2015
Section 1 near chainage 50 m Section 2 near chainage 100 m
B.7 Measured Surface Deflections

Surface deflections were measured using an FWD at the following times:
• 14 November 2011, the day after construction and prior to placement of the asphalt surfacing
• 5 January 2012, about 7 weeks after stabilisation and after placement of the asphalt surfacing
• 6 December 2012, after 13 months trafficking
• 29 May 2014, after 2.5 years trafficking
• 1 April 2015, after 3.5 years of trafficking.
• 9 December 2015, after 4 years trafficking
• 9 November 2016, after 5 years trafficking.
The deflection bowls were measured using a contact stress in the range 600–700 kPa, then normalised to
566 kPa. The surface temperatures during the FWD measurements in 2011, 2012, late 2015 and 2016 were
similar and within 4 °C of the WMAPT of 29 °C, whereas the surface temperature during the May 2014
measurement was significantly lower, at 14 °C. During the April 2015 measurement, the surface temperature
was 22 °C. As the pavement temperature at mid-depth was not measured, the deflections and curvatures
cannot be reliably temperature-corrected.
The maximum deflections (D0) and curvatures (D0–D200) normalised to a stress of 566 kPa but not corrected
for measurement temperature are plotted in Figure B 7. Note that the deflections measured on
14 November 2011 (the day after stabilisation) were measured without the 30 mm thick asphalt surfacing.
Had the surfacing been in place, it is estimated that the maximum deflections would have been about
10% lower and the curvatures about 20% lower. Note that, in part, the lower deflections and curvatures
measured in 2014 may have been due to the lower pavement temperature.

Figure B 7: Kewdale Road measured surface deflections

0.55
Section 1 Section 2
150 mm 100 mm
0.45
0.35
Maximum
deflection 0.25
at 566 kPa
(mm)
14 Nov 2011, 1 day
0.15 5 Jan 2012, 7 weeks
6 Dec 2012, 13 months
29 May 2014, 2.5 years
1 April 2015, 3.5 years
0.05
9 Dec 2015, 4 years
9 Nov 2016, 5 years
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125
-0.05
Chainage
0.24
Section 1 Section 2
0.22 150 mm 100 mm
14 Nov 2011, 1 day
0.20
5 Jan 2012, 7 weeks
0.18 6 Dec 2012, 13 months
29 May 2014, 2.5 years
0.16 1 April 2015, 3.5 years
9 Dec 2015, 4 years
0.14 9 Nov 2016, 5 years
Curvatures
at 566 kPa 0.12
(mm)
0.10
0.08
0.06
0.04
0.02
0.00
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 105 110 115 120 125
Chainage
The results to date suggest that 7 weeks after stabilisation the pavements had cured. After 5 years of
trafficking, there is no evidence from the testing that the FBS is being damaged by fatigue.

B.8 Field Cores

Due to the unexpectedly good performance after 5 years of trafficking, in April 2017 cores were excavated
from the pavements. For each section, eight cores were taken with four in the outer wheel path and four
between wheel paths (Figure B 8).
Figure B 8: Location of coring and record of visual inspection
Unexpectedly it was observed that:

• For three of the four cores taken, the outer wheel path of Section 1 and one of the Section 2 cores, a
bound cement treated crushed granite subbase was found under the FBS (e.g. Figure B 9). In addition,
cement fines covered the bottom of most of the other cores.
• Sections 1 and 2 had similar FBS thicknesses (mean = 115 mm) contrary to the design intent.
• The moduli of the field cores were low compared to other sites monitored (Table B 4, Table B 5).

Figure B 9: Examples of cemented subbase under FBS at Kewdale Road
Section 1 chainage 30 m OWP Section 1 chainage 40 m OWP
Table B 4: Modulus and density of field cores for Section 1 taken 5–6 years after construction

Asphalt FBS Dry As
Chainage Ratio
Offset thickness thickness density(3) received Cured dry Curedwet
(m) wet/dry
(mm) (mm) (t/m3) from modulus(1) modulus(2)
modulus
roadbed
10 Outer 35 105 mm FBS 1.90 1390 1820 1390 0.77
wheel path 35 mm CTCR
Between 30 130 mm FBS 2.00 2410 2450 2240 0.91

wheel 0 mm CTCR
paths
20 Outer 30 110 mm FBS 1.99 2000 2690 1740 0.65
Between 35 110 mm FBS 2.05 3750 3150 3330 1.06
wheel 0 mm CTCR
paths
30 Outer 30 120 mm FBS 1.82 2370 3250 2480 0.76
Between 30 105 mm FBS 2.29 5500 5660 5990 1.06
wheel 0 mm CTCR
paths
40 Outer 35 125 mm FBS - - - - -
Mean 30 mm 115 mm FBS 2.01 2900 3170 2860 0.87
1 Dried for three days at 40 °C before modulus testing at 25 °C.
2 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under vacuum of 95 kPa then tested
for modulus at 25 °C.

Table B 5: Modulus and density of field cores for Section 2 taken 5–6 years after construction

Asphalt Dry As
Chainage FBS thickness Ratio
Offset thicknes density(3) received Cured dry Cured wet
(m) (mm) wet/dry
s (mm) (t/m3) from modulus(1) modulus(2)
modulus
roadbed
60 Outer wheel 30 100 mm FBS 1.87 4850 6280 4940 0.79
path 15 mm CTCR
Between 30 115 mm FBS 1.88 3950 2880 3490 1.21

wheel paths 0 mm CTCR
path 0 mm CTCR
Between 30 110 mm FBS 1.89 4440 5910 3810 0.65
path 0 mm CTCR
Between 30 125 mm FBS 1.96 6020 6700 4180 0.62
path 0 mm CTCR
Mean 30 115 mm FBS 1.91 4760 5080 4370 0.95
2 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under vacuum then tested for
modulus at 25 °C.
B.9 Traffic Monitoring

Annual average daily traffic (AADT) and the percentage heavy vehicles data are given in Table B 6.
Engineering Road Note 9 (MRWA 2013) provides a presumptive ESA per heavy vehicle value of 1.5. Using
this value, the annual ESA of loadings in Table B 6 were estimated.
Table B 6: Traffic data for Kewdale Road
Percentage
Annual daily Annual
Lane Date AADT heavy vehicles
heavy vehicles ESA
(slow lane)
Southbound slow May 2010(1) 6000 26 1590 8.7 x 105
lane
July 2012(1) 7300 28 2050 1.1 x 106
August 2015(2) 4290 22 960 5.2 x 105
May 2017(3) 5260 27 1440 7.9 x 105
1 Source: Austroads (2013).

2 Source: Personal communication by email from Colin Leek, City of Canning, 12 August 2015.
3 Source: Personal communication by email from Colin Leek, City of Canning, 19 June 2017.
Assuming the annual traffic of 8 x 105 ESA, the cumulative traffic over the 5 years to November 2016 is
about 4 x 106 ESA.

B.10 Kewdale Road Trial Summary

In November 2011, the City of Canning constructed two FBS pavement sections on Kewdale Road,
Welshpool, WA, to provide performance information for the project:
• Section 1: 50 m length with nominal 150 mm FBS thickness, actual thickness 115 mm
• Section 2: 50 m length with nominal 100 mm FBS thickness, actual thickness 115 mm.
Both sections have a 30 mm thick dense graded asphalt surfacing.
The host material stabilised was a mixture of crushed granite base and recycled asphalt, with a small
quantity of crushed limestone subbase. This mixture was stabilised with a nominal 3.5% bitumen and
0.8% quicklime.
After 5 years of trafficking, the findings are:

• Both sections are performing well with no apparent surface distress with only very fine transverse
shrinkage cracking.
• Recent pavement coring indicates the presence of a cement treated crushed granite subbase. This
subbase provides support to the FBS layer and reduces the traffic-induced tensile strains. This is the
most likely reason for the lack of observed fatigue cracking despite the FBS only being about 115 mm
thick and a cumulative traffic loading of about 4 x 106 ESA.
• Due to the presence of the cement treated granite subbase it is unlikely that FBS fatigue will occur at this
site. Hence performance monitoring was discontinued.

Appendix C Port Wakefield Road, Virginia
C.1 Introduction
About 1 km length of the northbound carriageway slow lane of Port Wakefield Road, Virginia, South Australia
was foamed bitumen stabilised in April 2011. The project comprised the following three pavement sections:
• a control asphalt section (75 m long between chainages 0 and 75 m) comprising 100 mm thickness of
dense graded asphalt on granular base and subbase
• an under-designed sub-section (75 m long between chainages 75 and 150 m) comprising 150 mm thick
FBS base with a double/double seal surfacing
• a main section (900 m long between chainages 150 and 1067 m) constructed with 200 mm thick FBS
base with a double/double seal surfacing.
Details of locality and job statistics (job size, FBS specification and construction method) for the Port
Wakefield Road FBS pavement sections are given in Table C 1. The annual average rainfall in the nearby
township of Virginia is about 400 mm. The total measured rainfall between construction in April 2011 and
April 2013 was 790 mm, close to that expected from the average rainfall.
Table C 1: Description of the Port Wakefield Road FBS pavements
Description
Location Port Wakefield Road, Virginia, South Australia, between Angle Vale Road and Park Road
Job size Length Nominal 150 mm thickness FBS: chainages 75–150 m
Nominal 200 mm thickness FBS: chainages 150–1067 m
Number of lanes 1 (northbound outer lane)
Lane width 3.5 m (stabilisation width 3.9 m)
Foamed Stabilisation depth Section 1: nominal 150 mm
stabilisation Section 2: nominal 200 mm
specification
Host materials Calcrete limestone crushed rock base PM1/PM2
Mix design Empirically-based method
Foamed bitumen 3.0% Class 170 bitumen
Supplementary binder 1% hydrated lime
Foaming agent Not recorded
Construction Work specifications and QA DPTI Specification: Part 224 Foamed Bitumen Pavement
method testing (January 2007)
Construction date April 2011
Surfacing Sprayed bituminous double seal: size 16 mm/7 mm
Figure C 1 shows the test site prior to stabilisation. Only the outer lane was stabilised with hydrated lime and
foamed bitumen.

Figure C 1: Port Wakefield Road FBS site before stabilisation
C.2 Site Investigation Prior to Stabilisation

Prior to stabilisation, coring and sampling of the pavement and subgrade materials was undertaken at two
locations (chainages 100 m and 125 m, 1.75 m from the outer edge line) and dynamic cone
penetrometer (DCP) testing of the subgrade was also performed. The results indicated that the unbound
granular pavement was uniform and had a total pavement thickness in the range of 480–500 mm, which
included:
• 14 mm/7 mm sprayed seal
• 160 mm calcrete limestone crushed rock base
• 190 mm size 60 mm quartzite sandstone rubble subbase
• 150 mm size 75 mm calcrete limestone rubble subbase
• subgrade clay, estimated in situ CBR in the range 9–20% (Table C 4).
C.3 Mix Design

The base material sampled from the roadbed was identified as crushed calcrete limestone, with a particle
size distribution complying with the Department of Planning, Transport and Infrastructure (DPTI)
requirements for crushed rock base, and a PI less than 6%. Accordingly, DPTI considered the material was
suitable for foamed bitumen stabilisation.
For this job, DPTI specified a 3.0% residual bitumen and 1.0% hydrated lime in the FBS process, based on
previous experience.

C.4 FBS Thickness Design and Predicted Allowable Traffic Loading

The nominal 200 mm thickness of FBS was used for the main section (900 m long between chainages
150 and 1067 m). This FBS thickness was selected by DPTI.
The works also included a nominal 150 mm thickness of FBS used for the 75 m long under-designed section
to increase the probability of pavement distress.
As discussed in Appendix C.7.2, excessive bitumen was used in the construction of the 150 mm thickness
section. As this mix was not representative of FBS, the calculations below are confined to the 200 mm
thickness section.
Table C 2 shows the pavement of the section with a nominal 200 mm thickness of FBS and the selected
design modulus values used to predict the allowable loading using the interim FBS thickness design
procedure (Austroads 2011a).
Table C 2: Data used to predict the FBS fatigue life of Port Wakefield Road trial
Pavement Design
Available data Technical basis for data selection
component parameter
Subgrade Type Clay, in situ CBR 9% In situ DCP testing Table C 4
Design modulus 90 MPa
Subbase Thickness 300 mm Normal granular base/subbase as per original design
Design modulus 245 MPa Austroads Guide to Pavement Technology Part 2,
of top sublayer Table 6.4 (Austroads 2012)
FBS base Thickness 170 mm Field cores (Table C 8)
Design modulus 950 MPa Estimated using the moduli of 18-month-old field
cores from Port Wakefield Road and Western
Freeway Ballan field core and mix design results
Volume of binder 7% Based on the measured bitumen content of cores
(Table C 9)
Wearing course Thickness N/A Double/double spray seal
Design modulus N/A
The interim thickness design method (Austroads 2011a) requires the FBS design modulus to be determined
from indirect tensile moduli measured on laboratory-manufactured specimens cured for 3 days at 40 °C and
then soaked in water under vacuum of 95 kPa for 10 minutes.
As a mix design was not undertaken, the FBS design modulus was estimated as follows using the moduli of
18-month-old field cores together with the test results from the Western Freeway Ballan trial (Appendix F):
• For the 200 mm thick FBS section, the mean modulus of the cores taken from the untrafficked areas of
pavement after 18 months was 5500 MPa (Table C 8).
• By comparison, the mean modulus of untrafficked cores of 3% bitumen mix of the Western Freeway,
Ballan trial after 1 year was about 11 900 MPa (Table F 19). Hence the moduli of the Port Wakefield core
were about 45% of the Western Freeway values.
• For the Ballan 3% bitumen mix, the mix design cured wet mix design was 2310 MPa (Table F 5).
• Based on the ratio of the field core moduli, the Port Wakefield cured wet mix design modulus was
estimated to be 45% of 2310 MPa, that is 1100 MPa.
Using the Austroads (2011a) speed and temperature adjustments at a WMAPT of 27 °C and a design traffic
speed of 80 km/h, the FBS design modulus of 950 MPa was calculated.

Using the interim design method, the design input data in Table C 2 and Equation 4 adjusted to 50% design
reliability (Section 8.4.3), the nominal 200 mm thick section was predicted to have an FBS fatigue life of
2.9 x 106 ESA at 50% design reliability. Note however that the estimated cured wet modulus of 950 MPa is
below the minimum (1500 MPa and 2000 MPa) commonly accepted in mix design (Section 2.4).
C.5 Pavement Construction
FBS Sections
The pavements were constructed between 11 to 15 April 2011, in accordance with the DPTI specification
(Department of Planning, Transport and Infrastructure 2007). The details of stabilisation and compaction
process are described below.
Stabilisation of the nominal 150 mm thick FBS section (chainage 75–150 m) commenced on 11 April 2011.
The sprayed seal was mixed through the crushed rock base using a Wirtgen WR2000 stabiliser, and the
hydrated lime was added at a measured spread rate of 2.6 kg/m2 compared to the design value of 3.0 kg/m2.
The section was foamed bitumen stabilised (Figure C 2) and compacted. The pavement was subsequently
sealed four days later, on 15 April 2011.
Figure C 2: Stabilisation equipment used at the Port Wakefield Road FBS site
In terms of the 200 mm thick FBS section between chainage 150 and 1067 m on 11 April 2011, the stabiliser
could only pulverise, wet mix and lime treat a 425 m long section on this day due to the presence of a
pavement patch between chainage 509 m and 573 m which caused excessive wear to the milling teeth.
During construction, this patch was thought to be a previous cement stabilisation treatment. It was
subsequently found to be an asphalt patch (Appendix C.6).
The following day, 12 April 2011, the remaining length (490 m) was pulverised, and another asphalt patch
was encountered between chainage 780 and 845 m, which again hindered progress.

Note that current best practice is to remove patches by milling and either spread the millings along the
project or dispose of the material. This practice was not adopted on this project: the patches were stabilised
with the same treatment as applied to the crushed rock.
Following pulverisation between chainages 575–1067 m on 12 April 2011, the lime was spread at a
measured rate of 4.0 kg/m2 close to the design value of 4.1 kg/m2.
On 13 April 2011, the entire length of nominal 200 mm thick FBS (chainages 150 m and 1067 m) was
foamed bitumen stabilised and compacted. Note that between chainages 150–575 m, there was a 2-day
delay between lime stabilisation and foamed bitumen stabilisation, whereas for chainages 575–1067 m the
delay was 1 day. In the event of more than a 24-hour delay TMR’s specification (Queensland Department of
Transport and Main Roads 2017a) states that treatment with an additional 0.5% lime should be considered.
This was requirement was not included in the DPTI specification.
After mixing with foamed bitumen, the 200 mm thick FBS section was compacted (Figure C 3) using:
• a vibrating padfoot roller of a mass of 12 t
• a vibrating smooth drum roller with a mass of 12 t
• a multi-tyre roller with a mass of 15 t.
Figure C 3: Compaction equipment used at the Port Wakefield Road FBS site

After the FBS was compacted, the in situ densities were measured to a depth of 150 mm and material
excavated at selected test sites to determine the modified Proctor maximum dry density (MDD) and optimum
moisture content (OMC) (Table C 3). The time delay between field mixing and laboratory compaction is
unknown. As discussed in Section 6, time delays can reduce MDD values. The density ratio at each site was
calculated by dividing the in situ dry density by the modified Proctor MDD. The results confirmed the FBS
layer met the specified minimum compaction level of 98% modified Proctor MDD. However, this compaction
level may have been overestimated as no allowance has been made for the possible effect on the laboratory
MDD values of the time delay between field mixing and laboratory compaction (Section 6.3).
Table C 3: Field dry densities and laboratory maximum dry densities for the nominal 200 mm thick section
Field dry density Modified Proctor Modified Proctor

Chainage (m) Density ratio (%)
(t/m3) MDD (t/m3) OMC (%)
196 1.98 1.98 9.0 100.0
352 1.98 1.95 9.0 102.0
431 2.01 1.93 9.0 104.0
487 2.06 2.01 6.0 102.5
591 2.02 2.01 8.0 100.5
711 2.02 1.99 8.0 101.0
838 1.98 2.02 7.0 98.5
1007 2.05 2.02 7.5 101.0
Mean 2.01 1.99 7.9 101.2
Due to rain on 14 April 2011, the application of the two-coat sprayed seal surface was delayed until
15 April 2011.
Asphalt Section
To benchmark the FBS performance, it was decided to construct a 100 mm thick asphalt inlay over a short
length of Port Wakefield Road adjacent to the 150 mm thick FBS section.
The 100 mm thick asphalt section was constructed by milling the existing pavement to a depth of 100 mm,
spraying a bitumen tack coat and placing the asphalt in two layers (40 mm wearing course, 60 mm thick
intermediate course) (Figure C 4). Following common local practice, a tack coat was not placed between the
two asphalt layers, as both layers were placed on a single day.
Figure C 4: 100 mm thick dense graded asphalt inlay

C.6 Visual Condition Monitoring

DPTI inspected the pavement about every 2 months from opening to traffic in April 2011 until reconstruction
in early 2014.
In January 2012, the first summer after construction, flushing was observed over a 30 m length of the
200 mm FBS section length, chainage 510 to 575 m. By March 2013, after the second summer and about
two years of trafficking, the rutting was up to about 50 mm deep in some areas (Figure C 5). Coring of the
pavement showed that this area of the pavement was an asphalt patch prior to stabilisation. As discussed in
Appendix C.5.1, during construction, this patch was stabilised rather than removed by milling and replaced
with crushed rock. Foamed bitumen stabilisation of this asphalt patch resulted in a mix with inadequate rut
resistance.
Figure C 5: Rutting due to foamed bitumen stabilisation of an old asphalt patch
In August 2012 after about 16 months trafficking, fatigue cracking was observed for the first time. Two
months later, in October 2012, a detailed inspection of the asphalt section, the nominal 150 mm FBS section,
and the 200 mm thick FBS section was undertaken. Fatigue cracking was observed in all three sections.
The cracking of the FBS sections started as fine transverse cracks in the wheel path and developed into
crocodile cracking (Figure C 6). This mode of crack development was similar to that later observed on the
Western Freeway Ballan trial (Appendix F.9).
In relation to the cracking of the 150 mm thick FBS section, as discussed in Appendix C.7.2, the recovered
bitumen contents of extracted cores were in the range of 4.4% to 5.7%, considerably in excess of the design
content of 3%. The high bitumen content of the FBS in this section casts doubt on whether the fatigue
cracking observed in this section is characteristic of FBS materials.
As discussed in Appendix C.7.3, for the 200 mm thick FBS section, the recovered bitumen contents were
variable (2.8% to 6.6%, excluding an area in which an asphalt patch was stabilised). It was noted that in
August 2012 no fatigue cracking was observed at the coring site (BH12) which had the lowest bitumen contents
(3.1%) and was near the design value. Only one other coring site (BH10) had a bitumen content (3.7%) close
to the design value. Fatigue cracking was observed (Figure C 6) nearby in August 2012.
Had Port Wakefield Road been more uniform in mix properties with bitumen contents close to design values,
there would have been greater confidence in concluding that the observed fatigue cracking was
characteristic of FBS mixes. Nevertheless, the nature of the cracking was consistent with that subsequently
observed at the Western Freeway Ballan trial (Appendix F.9).

Figure C 6: Example of cracking observed in 200 mm FBS section, after about 18 months of trafficking
Fine transverse cracking near chainage 925 m
Block or crocodile cracking

Flushing/embedment of the two-coat sprayed seal surface occurred at some locations (Figure C 7). It is
unknown whether the seal binder contained cutter oil which may have softened the bitumen in the FBS layer
and contributed to embedment.
Figure C 7: Example of flushing/embedment of the sprayed seal after about 2½ years of trafficking
C.7 Pavement Investigation

As a result of the premature pavement distress observed in August 2012, in October 2012 cores were
extracted from the pavement and dynamic cone penetrometer (DCP) testing of the subgrade (Table C 4) was
undertaken to estimate in situ CBR using the Austroads (2012) method. DCP testing was also undertaken
during construction in April 2011.
Asphalt Section
The asphalt section was cored at chainage 62.6 m in the outer wheel path and between wheel paths in an
area with fine fatigue cracking (Figure C 8).
While the core from the unloaded area between wheel paths was extracted with the upper and lower asphalt
layers bonded together, in the outer wheel path, the 40 mm thick wearing course was debonded from the
underlying asphalt (Figure C 8). As the cracking appeared to be confined to the wearing course, the cracking
may have been due to inadequate bonding between asphalt layers. The two asphalt layers were placed on
the same day and without trafficking of the intermediate course. Following normal local practice, a bitumen
tack coat was not applied to the surface of the intermediate course prior to the placement of the overlying
wearing course.

Table C 5 shows that the cracked wearing course asphalt from the outer wheel path was significantly lower
in modulus than the sound asphalt from the unloaded area between wheel paths.
Note that the subgrade strength was high strength with an estimated in situ subgrade CBR exceeding
10% (Table C 4).
Figure C 8: Port Wakefield Road asphalt coring
Core between wheel paths, bonding between asphalt

Lower portion of delaminated outer wheel path core
layers

Table C 4: Estimated in situ subgrade CBR
Estimated CBR (%) at depths below top

Chainage of subgrade
Section Description Offset Date
(m) 0–100 100–200 200–300 300–400
mm mm mm mm
Asphalt BH (during construction) 37.5 – April 2011 6 5 5 8
BH1 62.6 BWP October 2012 > 30 11 13 15
FBS BH (during construction) 100 – April 2011 6 11 14 14
nominal BH2 101 BWP October 2012 11 18 28 30
150 mm
thick BH (during construction) 125 – April 2011 18 25 6 7
BH4 122 BWP October 2012 > 30 > 30 15 11
BH6 144.5 OWP October 2012 13 30 > 30 N/A
FBS BH8 420 BWP October 2012 11 11 15 N/A
nominal BH11 820 BWP October 2012 11 11 15 N/A
200 mm
thick BH12 950 BWP October 2012 7 12 15 N/A
Table C 5: Laboratory test results of Port Wakefield asphalt cores after 18 months in service
Chainage Total asphalt Bulk density IT modulus

Description Offset Layer
(m) thickness (mm) (t/m3) (MPa)
BH1 62.6 OWP 90 Wearing course 2.23 1980

Intermediate course 2.24 7750
BH1A 62.6 BWP 90 Wearing course 2.21 9000
Intermediate course 2.25 9160
Mean 90 2.23 6970
Nominal 150 mm thick FBS Section
Four FBS cores were taken from the untrafficked area between wheel paths, and one core was taken in the
outer wheel path. The average FBS thickness was 130 mm, which was less than the 150 mm design thickness.
The cores were transported to ARRB’s laboratory in Melbourne for testing. About 30 mm was sawn from the
top and bottom of each core and the remaining 70–78 mm length was tested for indirect tensile modulus and
strength. As shown in Table C 6, the moduli were high; the average of the indirect tensile moduli was
6700 MPa. Figure C 9 shows three of the five cores obtained and the pavement condition at the time of coring.
Following indirect tensile modulus and strength testing at a temperature of 25 °C, the particle size distribution
and the recovered bitumen content of the cores were measured. The bitumen contents were measured using
the pressure filter method (AS/NZS 2891.3.3:2013) with Solvex MT solvent. The results are given in Table C 7.
Table C 6: Test results of FBS cores from nominal 150 mm section after 18 months in service
FBS Underlying Bulk

Chainage IT strength IT modulus
Description Offset thickness asphalt density
(m) (kPa) (MPa)
(mm) patch (t/m3)
BH2, top layer 101 BWP 140 No 2.10 770 7550
BH3 112 BWP 130 No 2.06 760 6890
BH4 122 BWP 140 No 2.09 920 7520
BH5 131 BWP 130 No 2.07 770 6250
BH6 144.5 OWP 110 No 2.09 720 5290
Mean 2.08 790 6700

Figure C 9: Cores obtained from the nominal 150 mm thick FBS section
BH3 chainage 111 m BH4 chainage 121 m BH6 chainage 144.5 m
Table C 7: Particle size distribution and bitumen content of field cores obtained in nominal 150 mm thick
FBS section
% passing sieve
Sieve size
Recommended(1) BH2 BH3 BH4 BH5 BH6
19.0 mm 87–100 100 100 97 98 100
13.2 mm – 97 96 95 95 89
9.5 mm 67–88 87 85 84 83 74
6.7 mm – 74 75 72 73 61
4.75 mm 50–65 62 62 60 60 49
2.36 mm 38–50 46 46 45 45 35
1.18 mm 26–38 35 35 34 34 28
600 µm 18–30 28 27 26 27 23
300 µm 14–24 22 21 20 22 19
150 µm 11–19 16 15 15 17 15
75 µm 8–16 11 10 9 12 11
Recovered bitumen content (%) – 5.1 4.4 5.7 4.8 5.7
Stabilisation depth (mm) – 140 130 140 130 110
Chainage (m) – 101 112 122 131 144.5
1 Proposed for Austroads.

The design bitumen content of 3.0% equates to application rates of 9.1 kg/m 2 and 12.2 kg/m2 for stabilisation
depths of 150 mm and 200 mm, respectively (assuming an in situ dry density of 2.03 t/m3). The average
measured bitumen content was 5.1%, which is considerably higher than the design content of 3.0%. The
high bitumen contents may have been due to:
• the mistaken use of the bitumen application calculated for the nominal 200 mm FBS depth on the shorter
section where the depth was reduced to a nominal 150 mm
• the stabilisation depths being below the 150 mm design thickness
• the inclusion of the old double seal in the recycled pavement.
The variation in measured bitumen content with stabilisation depth is illustrated in Figure C 10. Also shown
are the bitumen contents for the design bitumen application rate of 9.1 kg/m2 and for 12.2 kg/m2. Considering
the old double seal added an additional 1.5 to 2.0 kg/m2 of bitumen, the measured bitumen contents are
consistent with the use of an application rate of 12.2 kg/m2 for the nominal 150 mm stabilisation depth. That
is, the results suggest the mistaken use of the 200 mm thick FBS bitumen application rate on the 150 mm
thick FBS section.
The high bitumen content may have contributed to the higher densities for the nominal 150 mm FBS section,
and hence the higher moduli as shown in Figure C 11. Even though these bitumen contents are high,
generally the rutting is not yet excessive (Figure C 20).
Figure C 10: Measured bitumen contents of nominal 150 mm FBS cores compared to design application
rates
6.0
5.5
Application rate 9.1 kg/m2
Application rate of 12.2 kg/m2
5.0
Measured
4.5
Bitumen
content 4.0
(% by mass)
3.5
3.0
2.5
2.0
100 110 120 130 140 150 160 170 180 190 200
Depth of FBS stabilisation (mm)

Figure C 11: Variation in field core moduli at 18 months with density
12000
10000
150 mm FBS
200 mm FBS bottom
8000 200 mm FBS top
Indirect tensile
modulus 6000
(MPa)
4000
2000
0
1.86 1.88 1.90 1.92 1.94 1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10
Bulk density (t/m3)
Nominal 200 mm thick FBS
Six FBS cores were cut from the untrafficked area between wheel paths. The average thickness was
172 mm (Table C 8), which is below the 200 mm design thickness.
The cores were transported to the ARRB laboratory in Melbourne for testing. For cores less than 160 mm in
length (BH7, BH9), about 30 mm was sawn from the top and bottom of each core and the remaining 73–
84 mm length was tested for indirect tensile modulus and strength. Cores from the other sites were cut into
top and bottom halves prior to testing for density, modulus and strength. The results are listed in Table C 8.
Table C 8: Test results of FBS cores from the nominal 200 mm thick section after 18 months in service
IT
FBS Underlying Bulk IT modulus
Chainage strength
Description Offset thickness asphalt density at 25 ºC
(m) at 25ºC
(mm) patch (t/m3) (MPa)
(kPa)
BH7 200 BWP 150 No 2.02 430 2760
BH8, bottom layer 420 BWP 1.88 350 2220
BH9 660 BWP 160 No 2.04 980 7880
BH10, top layer 750 BWP 180 No 2.04 930 11 430
BH11, top layer 820 BWP 175 Yes 2.11 810 6430
Mean 172 2.02 570 5550

Figure C 11 shows that the moduli tended to increase with density and the FBS layer tended to have lower
density and moduli in the bottom portion compared with the top portion, consistent with the Newell Highway
(Figure E 29) and the Calder Freeway (Figure D 24) findings.
The very high modulus of cores from BH11 (not shown in Figure C 11, but listed in Table D 8) appears to be
due to the host material being an old asphalt patch. Note that at BH11, the FBS material was well-supported
by the underlying old asphalt patching material (Figure C 12). Due to this support, the density and modulus
of the bottom half of the layer was higher than other locations.
Figure C 12: BH11 core, FBS of an asphalt patch together with underlying asphalt patching material
Following modulus and strength testing, the particle size distribution and the recovered bitumen content of
the cores were measured. The results are given in Table C 9.
Table C 9: Particle size distribution and bitumen content of field cores obtained in nominal 200 mm thick
FBS section
% passing sieve
Sieve size BH8 BH8 BH10 BH10 BH11 BH11 BH12 BH12
BH7 BH9
top bottom top bottom top bottom top bottom
19.0 mm 95 95 100 99 99 96 100 89 93 100
13.2 mm 86 93 93 93 93 92 99 78 84 91
9.5 mm 67 84 80 81 81 81 93 68 71 78
6.7 mm 57 73 69 67 67 69 84 62 61 67
4.75 mm 48 62 56 55 54 60 72 49 50 57
2.36 mm 36 46 46 40 37 45 55 38 37 44
1.18 mm 29 36 38 30 26 34 43 30 29 32
600 µm 24 29 32 24 20 27 34 24 23 24
300 µm 20 24 27 18 14 21 25 16 19 17
150 µm 16 18 21 13 10 15 16 8 14 12
75 µm 12 12 14 8 5 9 10 5 9 7
Recovered bitumen content (%) 4.4 5.5 6.6 5.3 3.1 4.3 8.1 3.8 3.4 2.8
Stabilisation depth (mm) 150 180 160 180 175 190
Chainage (m) 200 420 660 750 820 950

The high bitumen content of core BH11 was due to the stabilisation of an old asphalt patch (Figure C 12)
between chainages 780 to 845 m (Appendix C.5.1).
Excluding the results from BH11, the average measured binder content was 4.4%, which is well above the
3% design content.
The recovered binder contents were correlated to some extent with the depth of stabilisation (Figure C 13).
The data suggests that the high binder content for BH7 was in part due to the depth of stabilisation being
150 mm rather than 200 mm.
As the sprayed seal surfacing of the original unbound granular pavements was included in the FBS material,
this would also have contributed about an additional 0.5% to the measured bitumen contents.
However, the very high bitumen contents at BH8 and BH9 cannot be explained, other than these again being
FBS stabilisation of old asphalt patches. The high bitumen content at this site was consistent with their
appearance (Figure C 14). It was concluded that the FBS mix at these sites is not considered to be
consistent with the design intent.
Figure C 13: Measured bitumen contents of nominal 200 mm FBS cores compared to design application rate
6.5
6.0 BH8
BH11 - asphalt patch
5.5
BH9
5.0
Bitumen
content 4.5
(% by mass) BH7
4.0
BH10
3.5 Application rate of 12.2 kg/m2
Application rate of 12.2 kg/m2 plus old seal (1.8 kg/m2)
Measured
3.0
BH12
2.5
100 110 120 130 140 150 160 170 180 190 200
Depth of FBS stabilisation (mm)

Figure C 14: Core taken at chainage 420 m, BH8
It was concluded that only the mixes at BH10 and BH12 had characteristics consistent with the design intent.
As shown in Figure C 15, fatigue cracking was not observed near the BH12 core site after 18 months
trafficking. However, severe fatigue cracking was observed about 30 m to 50 m away from the core
(chainages 925 and 990 m, Figure C 16). Severe fatigue cracking was observed near BH10 after 18 months
trafficking (Figure C 15).
Had Port Wakefield Road been more uniform in mix properties with measured binder contents consistently
close to design values, there would have been greater confidence in concluding that the observed fatigue
cracking was characteristic of FBS mixes.

Figure C 15: Cores extracted from the nominal 200 mm thick FBS section
BH10 chainage 750 m after 18 months trafficking BH12 chainage 950 m after 18 months trafficking

Figure C 16: Severe cracking between chainages 990–1000 m in October 2012 after 18 months trafficking
C.8 Relationship between Indirect Tensile Modulus and Strength

For the FBS material placed at Port Wakefield Road, the measured relationship between indirect tensile
modulus of field cores and indirect tensile strength listed in Table C 6 and Table C 8 is plotted in Figure C 17.
The strength and modulus results were highly correlated.
Figure C 17: Relationship between indirect tensile strength and modulus

12000
y = 8.59x
R² = 0.80
10000
8000
Indirect
tensile 6000
modulus
(MPa)
4000
2000
0
0 200 400 600 800 1,000 1,200 1,400
Indirect tensile strength (kPa)

C.9 Pavement Deflections

Pavement surface deflections were measured in the outer wheel path using a falling weight
deflectometer (FWD) at the following times:
• 1 March 2012, after about 11 months of trafficking
• 27 September 2012, after about 17 months of trafficking.
The deflection bowls were measured using a contact stress in the range 550–580 kPa, then normalised to a
stress of 566 kPa.
The maximum deflections (D0) and curvatures (D0–D200) normalised to a stress of 566 kPa are plotted in
Figure C 18.
Figure C 18: Port Wakefield Road measured surface deflections

1.0
0.9
0.8
0.7
0.6
Maximum
deflection 0.5
at 566 kPa
(mm) 0.4
0.3 Asphalt control March 12

150 mm FBS March 12
0.2 200 mm FBS March 12
Asphalt control Sept 12
Previous
0.1 150 mm FBS Sept 12 asphalt patch Previous
asphalt
200 mm FBS Sept 12
patch
0.0
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 100010501100
Chainage
0.28
Previous Previous
0.26 asphalt asphalt
patch patch
0.24
0.22
0.20
0.18
Curvature 0.16
at 0.14
566 kPa
(mm) 0.12
0.10
0.08
0.06 Asphalt March 12

150 mm FBS March 12
0.04 200 mm FBS March 12
Asphalt control Sept 12
0.02 150 mm FBS Sept 12
200 mm FBS Sept 12
0.00
0 50 100 150 200 250 300 350 400 450 500 550 600 650 700 750 800 850 900 950 100010501100
Chainage

Although the pavement surface temperatures were measured during the deflection measurements
(Figure C 19), the temperatures at depth are required to correct the deflections measured for FBS from the
testing temperature to a representative in-service temperature. Note that the surface temperature increase
with chainage measured in September 2012 reflected the surface heating during the morning of testing.
Given this variability, it was not possible to accurately estimate the mid-layer pavement temperatures from
surface temperatures. Accordingly, the process developed in Appendix D.15 to correct deflections to the
WMAPT was not used.
Figure C 19: Pavement surface temperatures during FWD measurements

45
40
35
30
Surface Mar-12
temperature Sep-12
(°C)
25
20
15
10
0 100 200 300 400 500 600 700 800 900 1000 1100
Chainage (m)
In March 2012, no cracking was observed by DPTI staff driving over the site, but six months later distress
was apparent in all sections. Figure C 18 shows that the deflections and curvatures of the nominal 200 mm
thick FBS section generally increased during this six-month period, consistent with the onset of cracking.
Despite the high-strength subgrade and thickness of underlying granular subbase, the curvatures 11 months
after construction were high (> 0.1 mm) compared to other sites being monitored. A possible reason for this
is fatigue damage to the FBS layers in the first 12 months of trafficking, despite surface cracking not being
observed. However, not known is the extent to which the increase in surface temperature between the March
and September 2012 deflection measurements resulted in the observed increases in deflections and
curvatures.
C.10 Rutting and Roughness

In May 2013, the rutting and the roughness of the project were measured.
The 1.2 m straight edge rut depths are given in Figure C 20. The high rutting in the vicinity of chainage
550 m was due to the stabilisation of an old asphalt patch (Figure C 5).
The lane IRI roughness values are plotted in Figure C 21. The roughness is high compared to most other
trial sections.

Figure C 20: Port Wakefield Road rut depth measurements, May 2013
35
30 Outer wheel path
Inner wheel path
25
Rut 20
depth
(mm)
15
10 asphalt 150 mm FBS
0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1
Chainage (km)
Figure C 21: Port Wakefield Road roughness measurements, May 2013
4.2
4.0
3.8
3.6
3.4
3.2
Lane 3.0
IRI
(m/km) 2.8
2.6
2.4
2.2
2.0
1.8
1.6
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
Chainage (km)

C.11 Traffic Monitoring

DPTI provided 7-day vehicle classification count data (two-way) for Port Wakefield Road (at a traffic count
site located 1.7 km south of Angle Vale Road) for two periods in July 2006 and August 2011 (DPTI 2011).
Table C 10 provides a summary of the daily number of heavy vehicles.
Table C 10: Traffic data for Port Wakefield Road, 1.7 km south of Angle Vale Road
Year Northbound average daily heavy vehicles Annual ESA

2006 745 9.8 x 105
2011 1240 1.6 x 106
Weigh-in-motion (WIM) axle load data is not available at this site to estimate the number of ESA per heavy
vehicle and hence the cumulative traffic loading. However, DPTI suggested an estimated value of the
average number of ESA per heavy vehicle of 3.6 for the outer lane of Port Wakefield Road based on WIM
data collected on similar road sites (Austroads 2013). Hence, the estimated annual traffic loading in 2011
was 1.6 x 106 ESA, assuming all heavy vehicles travel in the slow lane. Over the 16-month period between
construction in April 2011 and the first observation of fatigue cracking (August 2012), the cumulative traffic
loading was estimated to be 2.1 x 106 ESA.
C.12 Comparison of Observed and Predicted Performance

As detailed in Appendix C.4, using the interim thickness design method the predicted FBS fatigue life for the
nominal 200 mm thickness is 2.9 x 106 ESA at 50% design reliability based on an estimated rather than
measured mix design cured wet modulus.
Using the estimated annual loading of 1.6 x 106 ESA, the interim design process predicts cracking of the
FBS sections after about 22 months of trafficking, which is reasonably consistent with the initial cracking after
16 months (August 2012). However, as discussed in Appendix C.7.2, the bitumen content of the FBS mix
was variable, with a most coring having bitumen contents above 4%. The lack of a mix design for this project
and the non-uniformity of bitumen contents limits conclusions that can be drawn about the interim thickness
design method.
The pavement was rehabilitated with thick asphalt in early 2014 at which time monitoring of the Port
Wakefield Road trial ceased.
C.13 Port Wakefield Road Trial Summary

In April 2011, two trial sections comprising nominal 150 mm and 200 mm thicknesses of 3% FBS mix were
constructed on Port Wakefield Road, Virginia, South Australia. The average bitumen content of the nominal
150 mm thick section was very high, about 5.1%. As such, this short 75 m section of the project was not
considered representative of commonly constructed FBS mixes.
In relation to the nominal 200 mm thick section the trial findings were:
• As a mix design was not undertaken for this project, based on modulus testing of field cores there is
doubt as to whether the mixes used would comply with current mix design modulus requirements.
• The trial section was constructed to a reasonable standard of compaction, about a mean 101% modified
Proctor MDD. However, this compaction level may have been overestimated as no allowance has been
made for the possible effect on the laboratory MDD values of the time delay between field mixing and
laboratory compaction.
• Generally, there was a significant reduction of density with depth, about 5% difference in compaction
between the top and bottom of the FBS layer. This may be related to the comparatively light rollers used.

• The average depth of stabilisation was about 170 mm for the nominal 200 mm deep section.
• About 18 months after opening to traffic, transverse fatigue cracking commenced which subsequently
developed to crocodile cracking, the extent of which led to pavement reconstruction of the entire length
after three years of trafficking.
• Testing of six cores indicated variable recovered bitumen contents, and all but two of the cores had
contents above 4%. This casts doubts about whether the foamed bitumen application rate was excessive
and, if so, the usefulness of the field performance information.
• At the two sites that had recovered bitumen contents of 3.1% and 3.7%, the performance was different
from each other: after 18 months trafficking the higher bitumen content site fatigue cracked and the lower
content did not. Nevertheless, at the time the pavement was reconstructed crocodile fatigue cracking was
observed at least for one site where there was confirmation of FBS mix quality.
• In March 2013, after two summers of trafficking, rutting up to 50 mm deep occurred over a 65 m length of
the inner wheel path. Coring of this area showed that an asphalt patch had been stabilised. Foamed
bitumen stabilisation of this patch resulted in a mix with inadequate rut resistance.

Appendix D Calder Freeway, Woodend
D.1 Introduction
In March 2013, VicRoads and the stabilisation industry agreed to support the Austroads project by
constructing an FBS pavement on the Calder Freeway, Woodend, Victoria. The trial section was designed
with a reasonable likelihood of distress within the period of the research project.
Details of locality and job statistics (job size, FBS specification and construction method) of the Calder
Freeway FBS pavement are given in Table D 1.
The annual average rainfall in the nearby township of Woodend is 840 mm. The total rainfall from
March 2013 until substantial cracking in March 2016 was 1980 mm, which equates to an average annual
rainfall of 660 mm, well below the long-term average.
Table D 1: Description of the Calder Freeway FBS pavement
Description
Location Calder Freeway bypass of Woodend, northbound carriageway 69.3 to 69.5 km
Number of lanes 1 (outer lane, northbound)
Lane width 4.8 m (3.5 m slow lane and 1.3 m shoulder)
Total area 480 m2
Stabilisation depth Design 150 mm FBS base, constructed mean thickness about 200 mm
Foamed stabilisation Mix design Undertaken by Downer Infrastructure and TMR
specification
Host materials VicRoads Class 1 crushed rock (size 20 mm), plasticity index 2 to 6
Supplementary binder Design 1.5% quicklime, constructed mean 1.1% quicklime
Bitumen Design 3.5% Class 170 bitumen, constructed 2.8% bitumen
Foaming agent 0.6% Teric 311
Construction method Work specifications VicRoads specifications
and QA testing
Construction date Foamed bitumen stabilisation, 4 March 2013
Two-coat primerseal of FBS, 5 March 2013
Surfacing Size 14/7 primerseal, includes 8 parts cutter oil and 10 parts rubber
Figure D 1 shows a view of the trial site. Prior to stabilisation, the pavement was a granular pavement with a
sprayed seal surface, originally constructed in 2001 as part of the bypass of the town of Woodend. After
about 10 years in service, pothole patching of the top 100 mm of the base was required in the outer wheel
path. This asphalt patching material and the existing sprayed bituminous seal were mixed with the crushed
rock base during the foamed bitumen stabilisation.
As part of the trial, an asphalt inlay was placed (Appendix D.8) over a 100 m length of adjacent pavement to
enable comparison of FBS fatigue performance with that of dense graded asphalt.

Figure D 1: Calder Freeway FBS site during construction
Source: Photograph taken by VicRoads staff.
D.2 Site Investigation Prior to Stabilisation

The Calder Freeway bypass of Woodend was originally opened to traffic in December 2001. It was designed
and constructed using a design traffic loading of 1.1 x 107 ESA over a 30-year period. The pavement
structure prior to stabilisation comprised:
• a prime and double/double size 14/7 seal
• 200 mm Class 1 (size 20 mm crushed rock) base
• 150 mm Class 3 (size 20 mm crushed rock) upper subbase
• 150 mm Class 4 (size 40 mm crushed rock) lower subbase
• 300 mm Type A capping layer
• subgrade clay (laboratory soaked design CBR of 2%).
D.3 Properties of the Untreated Crushed Rock Base

The crushed rock base was a nominal VicRoads Class 1 base of low plasticity and a particle size distribution
as given in Figure D 2. The samples were taken from the roadbed in March 2013, after the pulverisation pass
(Appendix D.7.1) and before addition of the lime and bitumen. The grading indicates the crushed rock base
was suitable for stabilisation.

Figure D 2: Particle size distribution of the untreated crushed rock base

100
90 Proposed Austroads lower limit

Proposed Austroads upper limit
80
Feb 2012 sample from road bed, tested by Downer
70 Oct 2012 sample from road bed, tested by TMR
March 2013 sample from road bed during construction
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)
D.4 Mix Design Prior to Construction

Prior to construction, samples of the crushed rock base were obtained from the pavement for the mix design.
The designs were undertaken by two experienced laboratories: Downer Infrastructure (contractor for the
Calder Freeway trial) and Queensland Department of Transport and Main Roads (TMR).
Downer Infrastructure Testing
In early 2012, mix design testing was undertaken by Downer Infrastructure at the request of ARRB. The
untreated crushed rock base was non-plastic (unexpected as it was specified to have a plasticity index in the
range 2–4), with a particle size distribution as shown in Figure D 2. The modified Proctor maximum dry
density (MDD) of the untreated material was 2.25 t/m3 and the optimum moisture content (OMC) was 4.8%.
The modified Proctor MDD of lime and foamed bitumen treated material (3.5% bitumen, 1% quicklime) was
2.18 t/m3 and the OMC was about 6%.
By comparison, in subsequent testing of material retrieved during construction, the FBS material MDD was
2.17 t/m3 and the OMC was 5.5% (Appendix D.7.8). However, this field-mixed material was compacted in the
laboratory about 24 hours after sampling from the roadbed. As discussed in Section 6.3, a 24-hour delay
between mixing and compaction results in about a 2% reduction in density. Had the field-mixed material
been compacted soon after sampling, it is expected the MDD would have been about 2.21 t/m3.
To prepare the modulus test specimens, the untreated base was mixed at the bitumen, quicklime and moisture
contents shown in Table D 2. The target mixing moisture content was 85% of modified Proctor OMC of the
untreated material, that is about 4.1%. The actual mixing moisture contents used varied from 3.9 to 4.3%,
slightly wetter than the mean field moisture content measured during construction (3.9%, Table D 8).

Modulus test specimens (150 mm diameter) were then compacted using a Servopac gyratory compactor
using a stress of 235 kPa and 3° angle of gyration. The specimens were then tested for indirect tensile
modulus at a temperature of 25 °C and 40 milliseconds rise time. The results are given in Table D 2. Note
that for the 3.5% bitumen specimens, the mean dry density of 2.19 t/m3 is well above the in situ mean
densities subsequently constructed (Table D 8).
At a bitumen content of 3.5%, the cured wet modulus of 1030 MPa was below the Downer mix acceptance
criteria (minimum 1500 MPa), despite the test specimen densities being higher than the mean field densities
(2.13 t/m3, Table D 8). Note that only 1% quicklime was used in the testing. Higher moduli may have been
achieved using 1.5% quicklime, which was the target in construction. Note also that the untreated material
used in the mix design was sampled from a pavement area several hundred metres outside the trial section.
Even though the grading of the material was similar (Figure D 2), other material properties may have
differed.
Table D 2: Downer Infrastructure mix design results
Mean indirect tensile modulus (MPa) at

Bitumen Quicklime Moisture Mean dry 25 °C Ratio
content content content density(4) wet/dry
(%) (%) (%) (t/m3) Initial Cured dry Cured wet modulus
3.0 1.0 3.9 2.14 463 2220 1180 0.53
3.5 1.0 4.2 2.19 411 2680 1030 0.37
4.0 1.0 4.3 2.19 522 4530 1970 0.44
Downer’s minimum requirements – – 1500 –
Queensland Department of Transport and Main Roads Testing
In December 2012 prior to construction, another sample of the crushed rock base was excavated from the
pavement to enable mix design testing by TMR. Following TMR practice, the modulus test specimens were
compacted into a 150 mm diameter mould using a heavy (9.92 kg) Marshall hammer by applying 50 blows to
each end of the specimen.
The particle size distribution of this sample had significantly less fines than the material used in the Downer
testing (Figure D 2). In addition, a plasticity index of 4–5 was measured by TMR whereas the earlier Downer
material was measured to be non-plastic. Downer suggested a possible reason for the difference in plasticity
index as reported in Appendix E.3.3.
For the TMR mix, the standard Proctor dry density-moisture compaction curve of the untreated material was
measured: the standard Proctor MDD was 2.22 t /m 3 and OMC was 6.6%.
The first step in the mix design was to wet the crushed rock to a target moisture content of 70% standard
Proctor OMC of the untreated material – that is 4.6%. This moisture content is significantly wetter than the
mean field moisture content measured during construction (3.9%, Table D 8), but as discussed in
Appendix D.7.4, these measured field values may be lower than the actual values at commencement of field
compaction due to the high temperatures during construction.
The bitumen, hydrated lime, moisture contents and modulus values are shown in Table D 3. Note that the
densities of the test specimens were not measured.
The FBS mixes at bitumen contents of 2%, 3% and 4% all meet the TMR acceptance criteria shown in
Table D 3.

Table D 3: TMR mix design testing
Mean indirect tensile modulus (MPa) at 25 °C

Bitumen Hydrated lime Moisture Ratio of
content content content Initial Cured dry Cured wet wet/dry
(%) (%) (%) modulus(1) modulus(2) modulus(3) modulus
2 2.0 4.6 1 400 12 980 5 970 0.46

3 2.0 4.5 1 460 10 050 5 510 0.55
4 2.0 4.6 1 380 7 600 4 780 0.63
TMR minimum requirements for base with
700 4 000 2 000 0.50
> 1000 ESA/day
Note that the TMR moduli are significantly higher than the Downer values (Table D 2). Later testing of
materials retrieved from the roadbed (Appendix D.7.9) suggested that this is in part due to the different
methods used to compact the test specimens: TMR used the Marshall impact hammer and Downer the
gyratory compaction. This issue is discussed in Section 4.3. The higher moisture content used by TMR is
likely to be another factor in the higher moduli, as discussed in Section 3.
Adopted Mix Design
As shown in Table D 3, TMR testing indicated that the crushed rock base stabilised with 3% bitumen and
2% hydrated lime met all TMR minimum mix design requirements for a basecourse on heavily trafficked
roads. To allow for construction tolerances, TMR practice at the time (TMR 2012) was to add 0.5% bitumen
content to that obtained in the mix design. Accordingly, it was decided to adopt the following binder contents:
• 3.5% bitumen
• 1.5% quicklime (equivalent to about 2% hydrated lime).
D.5 Mix Design after Construction

During construction, samples of the untreated crushed rock base were obtained (Appendix D.7.5) to repeat
the mix design testing. The testing was undertaken by the Downer laboratory in 2015.
The particle size distribution of the untreated crushed rock was measured, and the results are plotted in
Figure D 2. The material varied from the TMR mix design material but was similar to the original Downer mix
design material.
The untreated crushed rock was compacted into 100 mm diameter moulds to determine the standard and
modified Proctor dry density-moisture compaction curve. The modified Proctor MDD was 2.24 t/m3 and the
OMC was 5.6%, as shown in Figure D 3. Although the MDD value was similar to that measured by Downer
during the mix design (Appendix D.4.1), the OMC was 0.8% higher.
The standard Proctor MDD was 2.19 t/m3 and the OMC was 8.2%, as shown in Figure D 3. Again, the MDD
value was similar to that measured by TMR during the mix design (Appendix D.4.2). However, again there
was a significant difference in the OMC values. The OMC measured by TMR was 6.6%, that is 1.6% less
than the Downer value measured on the field-mixed material.

Figure D 3: Standard and modified Proctor MDD and OMC of untreated material sampled from roadbed
during construction
Mix designs are commonly undertaken at either:

• 80–85% of the modified Proctor OMC of the untreated crushed rock, which for this crushed rock equates
to a target moisture content of 4.5–4.8%
• 70% of the standard Proctor OMC of the untreated crushed rock, which for this crushed rock equates to a
target moisture content of 5.7%.
Considering these values and the 4.6% moisture content used in the TMR mix design (Appendix D.4.2), the
mix design was undertaken using a moisture content of 4.8%. The bitumen content (3%) and hydrated lime
content (1.5%) are reasonably consistent with the amounts of binders added during construction
(Appendix D.7.6, Appendix D.7.7).
Following TMR practice, the modulus test specimens were compacted into a 152 mm diameter mould using
a heavy (10.2 kg) Marshall hammer by applying 50 blows to each end of the specimen. The measured mean
dry density and moduli are given in Table D 4. It is apparent that the mix complies with the TMR minimum
requirements for base. However, the results are very different from TMR’s mix design testing using
2% hydrated lime prior to construction (Table D 3).
The mean dry density of these Marshall specimens of 2.12 t/m3 equates to about 100% modified MDD of the
treated material.
The moduli were also measured at lower moisture contents as discussed in Section 3.4.

Table D 4: Mix design testing post-construction
Hydrated Mean indirect tensile modulus (MPa) at

Bitumen Moisture Mean dry 25 °C Ratio of
lime
content Initial Cured dry Cured wet
3 1.5 4.8 2.124 700 5200 2720 0.52
TMR minimum requirements for base with
700 4000 2000 0.50
> 1000 ESA/day
D.6 Predicted Allowable Traffic Loading
FBS Crushed Rock Base
A key objective of the project was to obtain pavement performance data and feedback on the interim
thickness design method (Austroads 2011a). Accordingly, the Calder Freeway pavement was designed with
a reasonable likelihood of distress occurring within the project period. A nominal FBS thickness of 150 mm
was agreed, this being the minimum practical thickness based on stabilisation industry advice. However, the
mean constructed FBS thickness was close to 200 mm (Appendix D.11.1).
To evaluate the interim design method, the allowable traffic loading was predicted and subsequently
compared to the observed performance (Appendix D.17).
The allowable traffic loadings were predicted for the pavement composition shown in Figure D 4. Note that
the FBS thickness used was 185 mm based on the mean constructed thickness less a 10 mm tolerance.
Figure D 4: Calder Freeway pavement composition used in allowable traffic loading calculations
The interim thickness design method (Austroads 2011a) states that the FBS design modulus is determined
from indirect tensile moduli measured on laboratory-manufactured specimens cured for three days at 40 °C
and then soaked in water under partial vacuum of 13 kPa for 10 minutes. Hence thickness design
calculations were undertaken using FBS moduli derived from cured wet moduli of laboratory specimens. The
mean cured wet modulus and mean dry density of the test specimens are listed in Table D 4.

Using the Austroads (2011a) speed and temperature adjustments for a weighted mean annual pavement
temperature (WMAPT) of 24 °C and a design traffic speed of 80 km/h, an FBS design modulus of 2590 MPa
was calculated from the measured cured wet modulus (2720 MPa). As the interim design method includes a
maximum FBS design modulus of 2500 MPa, this value was used in the predictions.
Table D 5: Calder Freeway calculation of FBS design modulus
Post-construction mix design
IDT modulus at 25 °C Mean dry density(1) WMAPT (°C)

Heavy vehicle FBS design
40 ms rise time (t/m3) design speed (km/h) modulus (MPa)
(MPa)
2720 (cured wet) 2.12 24 80 2500 (max)
The fatigue life of the FBS layer was predicted using the Austroads interim design method, the input data
given in Table D 6 and Equation 4 adjusted to 50% design reliability (Section 8.4.3). The predicted allowable
traffic to FBS fatigue was 1.1 x 106 ESA at 50% design reliability.
Table D 6: Data used to predict FBS fatigue life of the Calder Freeway pavement
Pavement
component
Subgrade Subgrade type Clay Presumptive laboratory soaked CBR of 3% for
basaltic clay
Capping layer Fill thickness 300 mm Figure D 4
Design modulus 50 MPa Based on back-calculated moduli (Table G 2
(top sublayer) toTable G 23)
Subbase Subbase thickness 310 mm Assume high standard base material from Table 6.5
of GPT Part 2 (Austroads 2012)
(top sublayer)
FBS base FBS thickness 185 mm 10 mm less than mean constructed thickness
Appendix D.11.1
Design modulus 2500 MPa Table D 5
FBS volume of binder 6% Using 2.8% bitumen by mass and dry density of
FBS = 2.14 t/m3
Wearing course Thickness N/A Size 14/7 mm double/double primerseal
Design modulus N/A
As discussed in Appendix D.16, the estimated annual traffic loading is about 7 x 105 ESA. Consequently, the
interim design method predicts there is a 50% chance of observing FBS fatigue cracking about 1–2 years
after opening (that is late 2014). This allowable traffic loading was subsequently compared to the observed
performance (Appendix D.17).
Based on the findings of this project (Appendix D.11.1), the cured wet modulus is clearly well below the
modulus of early-life cores and back-calculated moduli estimated from early-life surface deflections. It was
concluded (Appendix D.11.1) that the cured dry moduli are more closely related to moduli of untrafficked
early-life field cores. Nevertheless, had the design modulus been estimated from the cured dry modulus, the
same design modulus would have been calculated if the method included a maximum design value of
2500 MPa.

Asphalt Section
The asphalt section was designed to have similar predicted fatigue life to the FBS stabilised section. A
size 20 Type SI dense graded asphalt (VicRoads 2014a) was selected and placed as a single layer to avoid
bonding issues between asphalt layers. VicRoads Code of Practice 500.22 (VicRoads 2013) lists a typical
design modulus of 3900 MPa for a WMAPT of 24 °C and heavy vehicle design speed of 80 km/h.
Using the GPT Part 2 (Austroads 2012) design method and Table D 7 inputs, the predicted asphalt fatigue
life is 2.9 x 106 ESA at 50% design reliability. As discussed in Appendix D.16, the estimated annual traffic
loading is about 7 x 105 ESA. Consequently, the Austroads design method (Austroads 2012) predicts there is
a 50% chance of observing FBS fatigue cracking within four years of opening to traffic. As discussed in
Appendix D.12, fatigue cracking was observed after only 12 months of trafficking.
Table D 7: Data used to predict asphalt fatigue life of the Calder Freeway pavement
Pavement
component
Subgrade Subgrade type Clay Presumptive laboratory soaked CBR of 3% for
basaltic clay
Capping layer Fill thickness 300 mm Figure D 4
Design modulus 50 MPa Based on back-calculated moduli (Table G 2 to
(top sublayer) Table G 23)
Granular base Subbase thickness 400 mm Assume high standard base material from Table 6.5
and subbase of GPT Part 2 (Austroads 2012)
(top layer)
Asphalt Asphalt thickness 100 mm Constructed mean thickness (Table D 14)
Design modulus 3900 MPa VicRoads Code of Practice 500.22
Design volume of 10.3% Adjusted 10.4% in VicRoads Code of Practice
binder 500.22 for the in situ air voids
Design air voids 8.2% Based on measured air voids of lower halves of the
field cores (Table D 15)
D.7 Construction and Testing of the FBS Trial Section
Construction of the FBS Pavement
The FBS pavement was constructed on 4 March 2013 and the unsealed surface was opened to traffic the
same day. The following day, the surface was sealed using a size 14/7 primerseal and opened to traffic
again.
The project was undertaken by VicRoads, Downer Infrastructure and Stabilised Pavements Australia Pty Ltd
as a contribution to this national research project.
The pavement section was 100 m in length and 4.8 m in width, covering the slow lane and part of the sealed
shoulder. The fast lane was not treated.
The construction sequence was as follows (Figure D 5):

• First mixing pass with Wirtgen Stabiliser WR2400 in two 2.4 m widths; this mixed the sprayed seal and
asphalt pothole patches through the crushed rock base to a nominal depth of 150 mm.
• The mixed material was then reshaped with the grader.
• Quicklime was then spread on the surface and slaked.

• The slaked lime was mixed through the crushed rock base to a nominal depth of 150 mm.
• The surface was then shaped and lightly compacted with a 12 t vibrating smooth drum roller.
• The lime reacted with the crushed rock base for about two hours while the bitumen heated to 170 °C in
the tanker.
• The foaming agent (0.6% Teric 311, about 70 litres) was added to the bitumen in the tanker and then
connected to the stabiliser.
• With the tanker and the stabiliser connected, the foamed bitumen was then mixed through the
lime-treated crushed rock base. The average bitumen application rate was 11.65 kg/m2. This equates to
about 3.6% bitumen content for the 150 mm design thickness, close to the target of 3.5%. However, the
mean constructed thickness was about 195 mm (Appendix D.11.1). For this thickness, an average
application rate of 11.65 kg/m2 equates to 2.8% bitumen content.
• The FBS crushed rock base was then shaped with a grader and compacted using:
(a) a 12 t vibrating smooth drum roller, and (b) a 16 t multi-tyred roller (Figure D 5).
Note that the surface was shaped and lightly compacted prior to the addition of bitumen. At that stage the
surface level was at or below the adjacent fast lane (Figure D 6: ). After the addition of the foamed
bitumen and heavy compaction, the surface level was still slightly below the adjacent fast lane. Hence the
grader reshaped the surface to move stabilised material from the shoulder to the traffic lane and then
compaction was completed (Figure D 7).
As mentioned in Appendix D.12.2, in some areas delamination of the top 30–40 mm of the FBS layer was
observed when slabs were later saw cut from the pavement in November 2015. The reworking of the surface
to correct levels may have been the cause of this delamination.
Figure D 5: Foamed bitumen stabilisation of the Calder Freeway, Woodend
Source: Photographs taken by ARRB and VicRoads staff.

Figure D 6: Surface level at or below the design level prior to addition of foamed bitumen and heavy
compaction
Figure D 7: Reworking the surface to achieve design levels

Figure D 8 shows the compacted unsealed surface prior to opening to traffic the same day.
Figure D 8: Finished unsealed surface prior to overnight trafficking
Surface Prior to Sealing
Prior to construction, there was discussion between VicRoads and the contractor as to whether the stabilised
pavement should be cured overnight and sealed the next day before opening to highway traffic. For this heavily
trafficked road, VicRoads required the road to be opened to traffic immediately following construction.
Consequently, the unsealed pavement was opened to traffic about two hours after compaction was completed.
When the unsealed pavement was inspected after the first 18 hours of trafficking, the unsealed surface had
ravelled under the heavy traffic loading, as shown in Figure D 9. In addition, there was a significant amount
of deformation of the FBS layer near the start of the job, also shown in Figure D 9. As this rutting was
excessive, the first 10 m of the job was subsequently milled and patched with a 100 mm thick asphalt layer
during the construction of the asphalt control section (Appendix D.8).
Figure D 9: Condition after trafficking overnight
Widespread ravelling overnight of the unsealed surface Deformation of the FBS material near the start of the trial
section

Double/Double Primerseal
A size 14/7 primerseal was placed on 5 March 2013, the day following the FBS construction.
The total binder application rate was 1.8 l/m2 (1.2 l/m2 and 0.6 l/m2). The sprayed binder included 8 parts
cutter oil and 10 parts rubber; the addition of the rubber was not planned but occurred as the sealing
contractor was using this binder for another VicRoads job in the area that day.
Subsequent advice from TMR recommended against the use of cutter oil in the seal as it may soften the
binder in the FBS layer and increase embedment of sealing aggregate.
Figure D 10: Primersealing
Field Density Testing
The field densities were measured at six sites in the slow lane after the completion of initial rolling
(Figure D 11). At each test site, a sample of FBS material was excavated for determination of oven-dried
moisture content. The wet density, moisture contents and dry densities are given in Table D 8.
Figure D 11: Nuclear gauge direct transmission field density measurements

Table D 8: FBS field wet densities, moisture contents and dry densities
Oven-dried
Wet density Dry density Density ratio(1)
Project chainage moisture content
(t/m3) (t/m3) (%)
(%)
10 2.18 3.8 2.10 95.0
20 2.22 3.9 2.13 96.4
30 2.22 3.6 2.14 96.8
40 2.22 4.1 2.13 96.4
50 2.21 3.8 2.13 96.4
60 2.24 4.2 2.15 97.3
Mean 2.22 3.9 2.13 96.4
1 Determined using a modified Proctor maximum dry density of 2.21 t/m3.
The mean field moisture content was 3.9% and the mean dry density 2.13 t/m3.
As described in Appendix D.7.8, the modified Proctor MDD of the foamed stabilised base was 2.17 t/m3 and
the modified OMC was about 4.4%. However, this field-mixed material was compacted in the laboratory
about 24 hours after sampling from the roadbed. As discussed in Section 6.3, a 24-hour delay between
mixing and compaction results in about a 2% reduction in density. Had the field-mixed material been
compacted soon after sampling, it is expected the MDD would have been 2.21 t/m3.
Using this adjusted MDD, the mean field density equated to 96–97% modified Proctor MDD. Note that this
level of compaction is about 6–7% below that subsequently achieved in the Western Freeway Ballan trial
(Section F.7.1).
The measured mean field moisture content of 3.9% was about 80% of the modified Proctor OMC determined
in the mix design (Appendix D.4.1).
It should be noted that field cores were taken after two years trafficking along the edge line and between the
wheel paths (Appendix D.11.1). The mean in situ dry densities were 2.09 t/m3 using the mensuration method
and 2.17 t/m3 using the presaturation method (Table D 13). The nuclear gauge mean dry density (2.13 t/m3)
was within the range of the field core results.
Following these field density measurements, the surface was reworked to provide a tighter surface by lightly
watering and re-rolling. A repeat series of measurements indicated no significant change in the mean wet
densities following this rework.
Properties of Untreated Material Sampled from the Roadbed
Samples of the untreated material were obtained during construction after the first mixing pass to pulverise
the existing base, pothole patches and the sprayed seal surfacing (Appendix D.7.1).
The untreated crushed rock was tested for particle size distribution, and the results are plotted in Figure D 2.
The material varied from the TMR mix design material but was similar to the Downer mix design material.
The material was non-plastic, again consistent with the Downer mix design material but different from the
plasticity index of 4–5 measured by TMR in its mix design (Appendix D.4.2).
The untreated crushed rock was compacted in a 150 mm mould to determine the modified Proctor dry
density-moisture compaction curve. The modified Proctor MDD was 2.25 t/m3 and the OMC was 5.3% as
shown in Figure D 12. These values are similar to those measured by Downer during the mix design
(Appendix D.4.1).

Figure D 12: Modified Proctor MDD and OMC of untreated and FBS materials sampled from roadbed during
construction
2.28
2.26 Untreated materials, 150 mm mould
2.24 FBS field-mixed material, 150 mm mould
2.22
2.20
2.18
2.16
2.14
Dry 2.12
density
2.10
(t/m3)
2.08
2.06
2.04
2.02
2.00
1.98
1.96
2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5
Lime Content
During construction, the quicklime was sampled and sent to Downer Infrastructure for testing. The active lime
content, measured in accordance with the test method ASTM C25-11e1, was 66% of CaO which was
considerably less than the lime supplied to the Newell Highway project (Appendix E.5.2). The quicklime lime
used in the roadbed had less available lime than used in the mix designs.
Samples of the untreated and lime-treated crushed rock were obtained from the roadbed at two chainages
before the foamed bitumen was mixed through the material. The samples were tested for lime content (as
CaO) in accordance with the test method ASTM C25-11e1.
The quicklime contents measured were 1.8% and 1.2% at chainages of 30 m and 60 m respectively,
compared to the design value of 1.5%.
In addition, using the quantity of lime spread, the contractor advised that the average quicklime spread rate
was 4.49 kg/m2. As the mean depth of stabilisation was about 195 mm (Appendix D.11.1), rather than the
design value of 150 mm, this spread rate equates to a quicklime content of 1.1% or about 1.5% hydrated
lime.
Bitumen Content
The 3.5% design bitumen content equated to an application rate of 11.2 kg/m2 for a 150 mm stabilisation
depth. The contractor advised that based on the litres of bitumen used (estimated from dipping the bitumen
tanker), the average bitumen application rate was greater – 11.65 kg/m2. As the mean depth of stabilisation
was about 195 mm rather than the design value 150 mm, an application rate of 11.65 kg/m2 equates to an
average 2.8% bitumen content.

Loose foamed stabilised crushed rock was sampled from the roadbed prior to compaction and tested for
bitumen content using three methods:
• the ignition oven method (Austroads 2005)
• pressure filter solvent washing method (AS/NZS 2891.3.3:2013)
• TMR solvent extraction method (Queensland Department of Transport and Main Roads 2014).
The measured binder contents were 3.4%, 3.7% and 2.9% respectively.
In addition, a bitumen content of 3.4% was obtained by ignition oven testing of a field core taken eight days
after construction (Appendix D.11.1).
Note that all these measured bitumen contents included asphalt pothole patching material and the pulverised
bituminous sprayed seal surfacing of the original pavement. It is also noted that single application
double/double sprayed seal would have added about 1.8 kg/m2 of bitumen, equivalent to increasing the
bitumen content by 0.4% for 195 mm depth of stabilisation.
Deducting 0.4% for the old sprayed seal, the mean measured bitumen content applied during foaming was
about 3.0%, reasonably consistent with the value of 2.8% as estimated from the amount of bitumen used.
Note that the design bitumen content was higher (3.5%).
Maximum Dry Density and Optimum Moisture Content
Uncompacted lime-treated and FBS-treated material was sampled from the roadbed during construction to
determine the modified Proctor dry density-moisture compaction curve. The modified Proctor MDD was
2.17 t/m3 and the OMC was 4.4%, as shown in Figure D 12.
However, this field-mixed material was compacted in the laboratory about 24 hours after sampling from the
roadbed. As discussed in Section 6.3, a 24-hour delay between mixing and compaction results in about a
2% reduction in density. Had the field-mixed material been compacted soon after sampling, it is expected the
MDD would have been about 2.21 t/m3.
Modulus Testing of Laboratory-compacted Field-mixed Material
Loose foamed bitumen stabilised crushed rock was sampled from the roadbed prior to compaction. The
samples were transported to Downer’s laboratory. The loose field-mixed samples were compacted in the
laboratory the following day. That is, there was about a 24-hour delay between field mixing of the lime and
foamed bitumen and the compaction of modulus test specimens, as follows:
• four specimens compacted in a 100 mm diameter mould using 50 blows of a 4.54 kg Marshall hammer at
each end
• four specimens compacted in a 100 mm diameter mould using gyratory (Servopac) compaction to a
target wet density of 2.26 t/m3
• six specimens compacted in a 150 mm diameter mould using gyratory (Servopac) compaction to a target
wet density of 2.26 t/m3.
The measured moisture content of this field-mixed material was 4%, closer to the value measured as part of
the field density measurement (Appendix D.7.4).
The measured dry densities using the mensuration method, initial, cured dry and cured wet moduli are given
in Table D 9. The cured dry moduli are plotted in Figure D 13. It is apparent that the specimens compacted
with the Marshall hammer had cured dry moduli about twice that of specimens compacted using the gyratory
method. The differences in moduli due to compaction methods and specimen size are discussed in
Section 4.

Figure D 13: Modulus variation with compaction method

8000
100 mm mould, Marshall hammer
100 mm mould, Servopac

7000
150 mm mould, Servopac
6000
5000
Indirect
tensile modulus
(MPa)
4000
3000
2000
1000
2.14 2.15 2.16 2.17 2.18 2.19 2.2 2.21 2.22 2.23 2.24 2.25 2.26 2.27
Dry density (t/m3)
Table D 9: Effect of compaction method on modulus

Dry Moisture
Compaction Sample Cured Cured Ratio
density(4) content Initial
method number dry wet wet/dry
(t/m3) (%) modulus(1)
modulus(2) modulus(3) modulus
100 mm M1 2.15 4.0 1090 6410 2740 0.43
diameter
M2 2.16 4.0 1070 6800 2780 0.41
mould,
Marshall M3 2.18 4.0 – 6980 3160 0.45
hammer M4 2.15 4.0 – 6900 3280 0.48
Mean 2.16 4.0 1080 6770 2990 0.44
100 mm S1 2.17 4.0 700 3260 1660 0.51
diameter
S2 2.19 4.0 510 2360 1230 0.52
mould,
gyratory S3 2.19 4.0 – 2660 1670 0.63
compaction to S4 2.20 4.0 – 2770 1730 0.62
a target wet
density Mean 2.19 4.0 605 2760 1570 0.57
150 mm 1 2.26 4.0 700 5900 3240 0.55
diameter
2 2.23 4.0 640 5740 3040 0.53
mould,
gyratory 3 2.23 4.0 610 5710 2930 0.51
compaction to 4 2.23 4.0 – 5620 3130 0.56
a target wet
density 5 2.23 4.0 – 5380 2920 0.54
6 2.23 4.0 – 5290 2910 0.55
Mean 2.24 4.0 650 5610 3030 0.54

The Table D 9 moduli measured on the field-mixed material compacted with gyratory compaction into
150 mm diameter moulds are considerably higher than those obtained by the same laboratory in the mix
design (Table D 2). For example, the mix design cured dry modulus was 2200 MPa at a bitumen content of
3%, whereas the mean value for the field-mixed material was 5620 MPa. In part, the field-mixed moduli are
higher due to their higher densities: the field-mixed specimens were about 2.5% higher in density.
Given the 24-hour delay in compacting the field-mixed samples, it would be expected that the modulus
specimens would be both lower in density and modulus than obtained in the mix design. The fact that the
reverse was observed casts doubt on the moduli obtained in the early mix design testing of specimens
compacted using the gyratory method (Table D 2).
D.8 Construction of Asphalt Section

On 21 March 2013, a dense graded asphalt inlay using a size 20 mm Class 320 bitumen (VicRoads mix
type SI) was placed over a 100 m section abutting the northern end of the FBS section to compare
performance with the FBS section. The design thickness was 100 mm and the mean constructed
thicknesses were about 100 mm and 110 mm, in the outer wheel path and between wheel paths
respectively.
The existing seal and top 100 mm of crushed rock base was milled from the pavement, the milled surface
broomed, a bitumen emulsion tack coat applied to the edges, and then the asphalt placed and compacted in
a single layer (Figure D 14). The underlying crushed rock base was very firm and tight and did not require
compaction prior to asphalt placement.
Figure D 14: Asphalt section of the Calder Freeway, Woodend
As discussed in Appendix D.7.2, a 10 m long section at the beginning of the FBS pavement was milled and
inlaid with asphalt as part of the works as it rutted in the first 24 hours of trafficking.

D.9 Surface Deflections
Foamed Bitumen Stabilised Sections
Falling weight deflectometer (FWD) deflection testing of the FBS crushed rock base was undertaken 8 days,
1 month, 1, 2 and 3 years after construction. Measurements were taken:
• on the edge line
• in the outer wheel path
• between the wheel paths.
The measured maximum deflections (D0) and curvatures (D0–D200) were normalised to a plate contact stress
of 566 kPa. The FBS mid-depth temperatures were measured during the deflection testing. The average
pavement temperatures were 29 °C, 12 °C, 21 °C, 18 °C and 16 °C at 8 days, 1 month, 1, 2 and 3 years
after opening to traffic, respectively. The maximum deflections and curvature were corrected from the
pavement temperature at the time of measurement to the WMAPT of 24 °C using the interim deflection-
temperature method developed in Appendix D.15.
The maximum deflections and curvatures are shown in Figure D 15 and Figure D 16. In the unloaded areas
(between wheel paths and along edge line), the deflections and curvatures reduced during the first year,
consistent with the FBS material increasing in modulus as it cures. After the first 12 months the deflections
and curvatures tended to stabilise.
In the outer wheel path, the deflections and curvatures decreased significantly during the first month
following construction, again consistent with curing of FBS. However, one year after construction, the trend in
the deflections and curvatures differed from that observed in the unloaded areas. Compared to the values
after one month, the deflections and curvatures in the outer wheel path did not decrease to the same extent
as the unloaded areas. After one year, the mean curvature in the outer wheel path was about twice the
between wheel path value. After two and three years of trafficking, the deflections and curvatures in the
wheel path increased significantly, particularly in areas of cracking.
Dense Graded Asphalt made using Class 320 Bitumen
Surface deflections were also measured on the 100 mm thick asphalt section two weeks, and one and two
years after construction. The maximum deflections and curvatures normalised to 566 kPa and adjusted from
the pavement temperature during measurement to the WMAPT of 24 °C are shown Figure D 17.
As discussed in Appendix D.12, asphalt fatigue cracking was observed when the pavement was inspected in
March 2014 and March 2015, that is, after about one and two years of trafficking. Figure D 17 is consistent
with the hypothesis that fatigue cracking leads to increases in deflections and curvatures. It was also
observed that the fatigue cracking appears to have led to an increase in deflections and curvatures between
wheel paths, albeit to a lesser extent than in the outer wheel path.
By January 2016, the cracking in the asphalt section had progressed in extent and severity such that
VicRoads removed the entire asphalt thickness and replaced it with a polymer modified binder asphalt
(Appendix D.9.3).

Figure D 15: Calder Freeway, Woodend FBS maximum deflections

Calder Freeway Woodend
foamed bitumen stabilised material outer wheel path
0.6
cracking OWP
cracking IWP
cracking IWP
0.5
0.4
Maximum
deflection
0.3
at 566 kPa Outer wheelpath 8 days
(mm)
Outer wheelpath 1 month
0.2 Outer wheelpath 1 year

Outer wheelpath 2 years
0.1
0.0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

foamed bitumen stabilised material between wheel paths
0.5
0.4
0.3
Maximum
deflection
at 566 kPa
(mm)
0.2
Between wheelpaths 8 days
Between wheelpaths 1 month
Between wheelpaths 1 year
0.1 Between wheelpaths 2 years
Between wheelpaths 3 years
0.0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

foamed bitumen stabilised material edge line
0.5
0.4
0.3
Maximum
deflection
at 566 kPa
(mm)
0.2 Edgeline 8 days
Edgeline 1 month
Edgeline 1 year
Edgeline 2 years
0.1
Edgeline 3 years
0.0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

Figure D 16: Calder Freeway, Woodend FBS curvatures

foamed bitumen stabilised materials outer wheelpath
0.14
cracking OWP
cracking IWP
cracking IWP
0.12 Outer wheelpath 8 days

Outer wheelpath 1 month
Outer wheelpath 1 year
0.10 Outer wheelpath 2 years
Curvatures 0.08
(D0-D200)
at 566 kPa
(mm) 0.06
0.04
0.02
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

foamed bitumen stabilised materials between wheelpaths
0.12
Between wheelpaths 8 days

Between wheelpaths 1 month
0.10 Between wheelpaths 1 year
0.08
Curvatures
(D0-D200)
0.06
at 566 kPa
(mm)
0.04
0.02
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

foamed bitumen stabilised materials edgeline
0.12
Edgeline 8 days
Edgeline 1 month
0.10 Edgeline 1 year
Edgeline 2 years
Edgeline 3 years
0.08
Curvatures
(D0-D200)
0.06
at 566 kPa
(mm)
0.04
0.02
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

Figure D 17: Calder Freeway, Woodend asphalt section

dense graded asphalt 2 weeks, 1 and 2 years post-construction
cracking
1.0
0.9
cracking cracking cracking
0.8
0.7
0.6
Maximum
deflection
0.5
at 566 kPa
(mm)
0.4
0.3 Outer wheelpath 2 weeks

Outer wheelpath 1 year
0.2 Outer wheelpath 2 years
Between wheelpaths 2 weeks
0.1 Between wheelpaths 1 year
0.0
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)

dense graded asphalt 2 weeks, 1 and 2 years post-construction
0.22 cracking
0.20
cracking cracking cracking
0.18
0.16
0.14
Curvatures
0.12
(D0-D200)
at 566 kPa 0.10
(mm)
0.08
0.06
0.04
Outer wheelpath 2 weeks Outer wheelpath 1 year
0.02 Outer wheelpath 2 years Between wheelpaths 2 weeks
Between wheelpaths 1 year Between wheelpaths 2 years
0.00
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)
Polymer Modified Binder Asphalt
In January 2016, the entire length of original asphalt section (Appendix D.9.2) was removed due to the
extent and severity of cracking. The replacement size 20 mm asphalt mix was manufactured using an
A10E polymer modified binder (PMB). This mix was chosen to enable an assessment of the difference in
fatigue life due to the use the A10E (SBS) polymer.

About 1 month after this PMB asphalt was placed, the deflections and curvatures were measured and
compared to those taken in April 2013 on the initially placed asphalt mix made using conventional Class 320
binder. The results are plotted in Figure D 18.
From Austroads (2012) it is expected that the modulus of an A10E mix is about 70% of that of a Class 320
bitumen mix. This is most likely the reason why the deflections and curvatures tend to be higher for the PMB
asphalt than the Class 320 bitumen mix.
Figure D 18: Comparison of asphalt moduli of patching materials using Class 320 bitumen and A10E binder
asphalt (Class 320 binder) at 2 weeks (April 2013) and PMB asphalt at 1 month (Feb 2016)
1.0
0.9
0.8
0.7
0.6
Maximum
deflection
0.5
at 566 kPa
(mm)
0.4
0.3 Outer wheelpath 2 weeks
Between wheelpaths 2 weeks

0.2
Outer wheel path PMB asphalt 1 month
0.1
Between wheel paths PMB asphalt 1 month
0.0
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)

asphalt (Class 320 binder) at 2 weeks (April 2013) and PMB asphalt at 1 month (Feb 2016)
0.22
0.20
Outer wheelpath 2 weeks
0.18 Between wheelpaths 2 weeks

Outer wheel PMB asphalt 1 month
0.16 Between wheel paths PMB asphalt 1 month
0.14
Curvatures
0.12
(D0-D200)
at 566 kPa 0.10
(mm)
0.08
0.06
0.04
0.02
0.00
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)

D.10 Moduli Back-calculated from Measured Surface Deflections
FBS Section
Using the measured surface deflection bowls at eight days, one month, one, two and three years after
construction, as well as the mean constructed FBS thickness of 200 mm from the coring (Appendix D.11.1),
the pavement and subgrade layer moduli were back-calculated. The results are listed in Appendix G.1.1.
Of particular interest are the FBS moduli variations with time and also the differences between values in the
unloaded pavement areas (edge line and between wheel paths) and loaded pavement area (outer wheel
path). Accordingly, it was necessary to adjust the back-calculated moduli from the measurement temperature
during FWD testing to WMAPT for Woodend (24 °C) using the Austroads (2011a) modulus-temperature
dependency.
The temperature-corrected FBS moduli are plotted in Figure D 19. It was noted that the mean FBS moduli in
the unloaded areas (outer edge line and between wheel paths) increased about 60% with curing between
measurements taken at eight days and one month, whereas the increase in the loaded area (outer wheel
path) was significantly less, at 40%. One month after construction, the moduli in the outer wheel path were
generally lower than in the unloaded areas, despite the anticipated additional compaction due to trafficking.
The lower moduli may be due to the FBS being damaged by trafficking.
Figure D 19: Calder Freeway, Woodend back-calculated FBS moduli

7 days after construction
5000
4500 Outer edge line
Outer wheel path

4000
Between wheel paths
3500
FBS
modulus
at 24°C
3000
(MPa)
2500
2000
1500
1000
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

1 month after construction

5000
Outer edge line

4500
Outer wheel path
4000 Between wheel paths
3500
FBS
modulus
3000
at 24°C
(MPa)
2500
2000
1500
1000
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

12000
Outer edge line
11000
Outer wheel path
10000
Between wheel paths
9000
8000
FBS
7000
modulus
at 24°C 6000
(MPa)
5000
4000
3000
2000
1000
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

2 years after construction

12000
Outer edge line
11000 cracking OWP
Outer wheel path cracking IWP cracking IWP
10000
Between wheel paths
9000
8000
FBS
7000
modulus
at 24°C 6000
(MPa)
5000
4000
3000
2000
1000
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
3 years after construction

12000
Outer edge line
11000 Outer wheel path
Between wheel paths

10000
9000
8000
7000
FBS
modulus
6000
at 24°C
(MPa) 5000
4000
3000
2000
1000 cracking IWP cracking IWP

cracking OWP
0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
After one year in the roadbed, the untrafficked areas significantly increased in modulus to a mean value of
about 7100 MPa, which is more than twice the value at one month (3300 MPa). In contrast, the mean FBS
modulus in the outer wheel path did not change significantly. To investigate this further, Figure D 20 shows
the variation in the mean FBS moduli with cumulative traffic loading. The data was analysed separately
according to whether cracking was apparent or not after two years of trafficking.
The following was noted:

• In terms of the wheel path moduli, as expected, the FBS moduli values after one, two and three years
trafficking tend to be lower in the cracked areas.

• The moduli of untrafficked areas increased markedly in the first year, which is consistent with curing.
However, the data suggests that the moduli tend to be lower in the second and third years, particularly for
chainages adjacent to wheel path cracking.
Figure D 20: Variation of FBS modulus with loading
8000
Edge line
1 year
Between wheel paths 3 years
6000
2 years
5000
FBS mean
modulus
at 24°C 4000
(MPa) 3 years
1 year
3000 1 month
2 years
2000
1000
30,000 300,000 3,000,000
Cumulative traffic loading (ESA)
Areas (chainage 15–45 m) without visible cracking in outer wheel path after two years or more
9000
Edge line 1 year

8000
Outer wheel path
Between wheel paths
7000
3 years
6000
2 years
FBS mean
modulus 5000
at 24°C
(MPa)
4000
1 month 1 year
3000
3 years
2000 2 years
1000
30,000 300,000 3,000,000
Cumulative traffic loading (ESA)
Areas (chainage 65–95 m) with visible cracking in outer wheel path after two years or more

Asphalt
Using the measured surface deflection bowls, asphalt thicknesses of 100 mm and 110 mm for the outer
wheel path and between wheel paths, respectively (Appendix D.11.2), and pavement and subgrade layer
moduli were back-calculated for eight days, one month, one, two and three years after construction. The
results are listed in Appendix G.1.2.
Of particular interest are the asphalt moduli variations with time and also the differences between values in
the unloaded pavement areas (edge line and between wheel paths) and loaded pavement area (outer wheel
path). To undertake this analysis, the back-calculated moduli need to be adjusted from the pavement
temperature during FWD testing to the WMAPT for Woodend (24 °C) using the Austroads (2011a)
temperature dependency. The temperature-corrected asphalt moduli are plotted in Figure D 21.
The mean back-calculated moduli are listed in Table D 10, grouped according to whether fatigue cracking
was observed in the outer wheel path at two years (March 2015).
Table D 10: Mean asphalt back-calculated asphalt moduli
Mean back-calculated modulus corrected to a temperature of 24 °C (MPa)

Outer wheel path Between wheel paths
Age since construction
Sound Cracked Sound Cracked
chainages(1) chainages(1) chainages(1) chainages(1)
2 weeks 6750 5740 7140 4970
1 year 7910 3440 7870 5040
2 years 8920 3690 7790 4110
1 Based on cracking assessment in March 2015 (Figure D 38).
Figure D 21: Calder Freeway, Woodend back-calculated asphalt moduli

Deflections measured 2 weeks after construction
10000
Outer wheel path

9000
Between wheel paths
8000
Asphalt 7000
modulus
at 24°C 6000
(MPa)
5000
4000
3000
2000
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)

Deflections measured 1 year after construction

12000
Outer wheel path
11000
Between wheel paths
10000
9000
8000
Asphalt
modulus 7000
at 24°C
(MPa) 6000
5000
4000
3000
2000
cracking cracking cracking cracking

1000
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)
Deflections measured 2 years after construction

13000
Outer wheel path
12000
Between wheel paths
11000
10000
9000
Asphalt
8000
modulus
at 24°C 7000
(MPa)
6000
5000
4000
3000
2000
cracking cracking cracking cracking
1000
100 110 120 130 140 150 160 170 180 190 200
Chainage (m)
The following was observed:

• As expected, the moduli in the outer wheel paths tended to be lower at chainages where surface cracking
was observed after two years. These moduli reduced with trafficking.
• The cracking in the wheel paths appears to have led to a reduction in modulus in the area between the
wheel paths, despite this pavement area not being directly loaded.
• For chainages where surface cracking was not observed after two years, the moduli tended to increase.

Note that when tested one year after construction, the mean back-calculated modulus of the untrafficked
areas at chainages 130 m, 160 m and 190 m was 5160 MPa at a WMAPT of 24 °C. At a temperature of
25 °C, this reduces to about 4760 MPa. This mean modulus is similar to the mean of indirect tensile modulus
(4500 MPa) of field cores extracted after one year between wheel paths (Figure D 15).
D.11 Field Core Moduli
Foamed Bitumen Stabilised Material
Cores of the FBS layer were taken along the edge line at eight days, one month and two years after
construction. Surface deflections were measured at these locations prior to coring.
At eight days, some of the cores could not be fully extracted as the upper and lower portions sheared apart
in the coring process. However, all four cores were extracted intact one month after construction.
Figure D 22 shows an example of a core taken on the edge line at chainage 40 m one month after
construction, while Figure D 23 shows examples of cores taken between the wheel paths two years after
opening to traffic.
Figure D 22: FWD deflection testing and coring at Calder Freeway site eight days after construction
200 mm

Figure D 23: Examples of cores extracted two years after opening to traffic
The cores were transported to Downer Infrastructure’s laboratory in Somerton, Victoria for testing. Each core
was sawn into upper and lower portions and the dry density and indirect tensile modulus measured. The
results are summarised in Table D 11 and Table D 12.
Table D 11: Modulus and density of field cores taken at eight days

Upper or
Layer Dry As Cured Cured Ratio
Chainage lower
Offset thickness density(3) received dry wet wet/dry
(m) portion of
(mm) (t/m3) from modulus(1) modulus(2) modulus
field core
roadbed
Upper 2.13 2540 4920 3470 0.70
15 Edge line 164
Lower 2.01 1910 4060 2540 0.63
Upper 2.04 3690 6510 4560 0.70
35 Edge line 147
Lower 1.96 1700 3670 2130 0.58
55 Edge line Not retrieved Upper 2.12 3170 6930 1960 0.28
Upper 2.13 4470 7540 3940 0.52
75 Edge line 218
Lower 2.00 1750 4060 1900 0.47
95 Edge line Not retrieved Upper 2.19 4190 9360 3490 0.37
Upper 2.12 3610 7050 3480 0.51
Mean 183
Lower 1.99 1780 5090 2870 0.58
2 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under partial vacuum then tested for
modulus at 25 °C.

Table D 12: Modulus and density of field cores taken at one month along the untrafficked edge line
Upper or Indirect tensile modulus (MPa) at 25 °C

Layer Dry
Chainage lower As received Cured Cured Ratio
Offset thickness density(3)
(m) portion of from dry wet wet/dry
(mm) (t/m3)
field core roadbed modulus(1) modulus(2) modulus
Upper 2.13 4438 5980 4300 0.72
20 Edge line 197
Lower 1.99 1940 4130 2820 0.68
Upper 2.11 4140 8520 4310 0.51
40 Edge line 204
Lower 2.01 1950 2920 2620 0.90
Upper 2.13 3740 5300 2770 0.52
60 Edge line 205
Lower 2.02 1020 1830 1890 1.03
Upper 2.06 1660 3530 2590 0.73
80 Edge line 196
Lower 1.98 1750 1790 1700 0.95
Upper 2.11 3470 5830 3490 0.62
Mean 201
Lower 2.00 1660 2670 2260 0.89
2 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under partial vacuum then tested for
modulus at 25 °C.
Figure D 24 shows that the moduli of cores obtained at eight days and one month were reasonably similar
for a given density. The results plotted are the cured dry moduli values and are compared to the mean cured
dry and wet moduli determined in the mix design (Table D 4). For this project, it is apparent that the mix
design cured dry modulus was similar to that in the first month after construction. The mix design mean
cured wet modulus was well below these early-life in situ values.
Figure D 24: Comparison of mix design cured dry modulus with early-life field core moduli
10000
Field cores 8 days
field core 1 month
9000
8000 Mix design cured wet modulus
7000
6000
Indirect
tensile 5000
modulus
(MPa)
4000
3000
2000
1000
0
1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12 2.14 2.16 2.18 2.20 2.22 2.24
Dry density (t/m3)

Two years after construction (March 2015), cores were taken along the outer edge line and between wheel
paths. The core thicknesses, dry densities and moduli are given in Table D 13.
Note that the dry densities were measured using both the mensuration and saturated surface dry (SSD)
methods.
Table D 13: Modulus and density of field cores taken two years after construction from untrafficked areas
Upper Indirect tensile modulus (MPa) at 25 °C

Layer or lower Dry Dry As
Chainage Cured Cured
Offset thickness portion density(3) density(4) received
(m) dry wet
(mm) of field (t/m3) (t/m3) from
core modulus(1) modulus(2)
roadbed
20 Between 192 Upper 2.155 2.253 13 670 15 570 9 860
wheel paths
Lower 2.050 2.148 8 040 9 680 4 350
30 Edge line 198 Upper 2.147 2.209 12 560 15 930 8 170
Lower 2.034 2.115 9 420 10 030 4 720
Between 203 Upper 2.146 2.214 11 260 13 850 8 440
wheel paths
Lower 2.028 2.133 5 560 6 430 2 830
40 Between 192 Upper 2.095 2.183 12 420 12 760 6 840
wheel paths
Lower 2.052 2.153 7 620 8 140 3 980
50 Edge line 195 Upper 2.133 2.197 12 560 12 880 6 360
Lower 2.056 2.150 7 760 8 080 4 420
Between 160(5) Upper 2.147 2.229 18 160 19 130 11 610
wheel paths
Lower Insufficient length for modulus testing
60 Between 153(5) Upper 2.146 2.216 13 220 12 820 10 530
wheel paths
Lower Insufficient length for modulus testing
70 Edge line 191 Upper 2.129 2.194 11 910 11 950 6 200
Lower 2.029 2.134 7 490 8 540 3 640
Between 200 Upper 2.177 2.239 12 880 17 290 9 600
wheel paths
Lower 2.079 2.170 10 280 10 340 6 070
80 Between 190 Upper 2.181 2.246 18 800 18 940 13 070
wheel paths
Lower 2.086 2.172 11 040 11 280 5 490
90 Edge line 190 Upper 2.067 2.145 8 670 9 040 4 940
Lower 1.953 2.060 6 500 7 840 3 860
Between 198 Upper 2.182 2.242 14 500 15 820 11 440
wheel paths
Lower 2.089 2.169 10 580 12 300 6 490
Upper 2.14 2.21 13 380 14 660 8 920
Mean 195
Lower 2.05 2.14 8 430 9 270 4 590
2 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under vacuum then tested for
modulus at 25 °C.
4 AS/NZS 2891.9.2:2014 Presaturation Method.
5 Suspect bottom of core not retrieved.
Figure D 25 shows the variation in the indirect tensile modulus of between wheel path cores as received from
the roadbed with density and age with the 2 year data labelled with the sampling chainage.

Figure D 25: Between wheel path field core moduli variation with density and time, sampling chainages
marked
20000
80
18000 50
16000 1 month cores-upper portion

1 month core-lower portion
2 year cores - upper portion 90
14000
2 year cores - lower portion 60 20
50 30 70
40
12000 70
Indirect 80 30
tensile 70 90
10000
modulus 30
(MPa) 20 90
8000
70 50
40
90
6000
30
4000
2000
0
1.95 1.97 1.99 2.01 2.03 2.05 2.07 2.09 2.11 2.13 2.15 2.17 2.19 2.21
Dry density by mensuration (t/m3)
The following conclusions were drawn from the analysis to date:

• The mean in situ moduli in the untrafficked areas have increased by about a factor of three between the
values one month and two years after construction.
• Unlike the back-calculated moduli trends (Figure D 19), the cores taken adjacent to cracked chainages
were not significantly different from the cores adjacent to sound areas. Presumably the back-calculated
moduli were affected by the wheel path cracking, as the deflection bowl is influenced by the material
properties over a large volume of FBS material compared to the 100 mm diameter core.
• Using all the measured layer thicknesses given in Table D 13 and excepting the outliers at chainages
50 and 60 m between wheel paths, the mean FBS thickness was 193 mm, which was well above the
target thickness of 150 mm. This reduces the bitumen and lime contents as discussed above.
• The mean in situ dry density of the lower portion of the FBS layer was 3–4% lower than the upper portion.
This reduction in density with depth exceeded that of the Western Freeway trial. Had a heavy padfoot
roller been used, as it was for the Western Freeway Ballan trial (Appendix F.6.4), it is anticipated that the
achieved compaction levels would have been higher with a lower reduction in compaction with depth.
• Using the two-year core data (Table D 13), the mean modulus of the lower portion of the cores was about
65% of the mean modulus of the upper portions. This difference in modulus was largely attributed to the
reduction in density with depth.
Asphalt
In March 2014, six cores were taken from the asphalt section: three in the outer wheel paths and three
between wheel paths. In addition, a further six cores were taken in the outer wheel path in March 2015
(Figure D 26). The cores were tested for layer thickness, air voids and indirect tensile modulus.

Figure D 26: Examples of asphalt cores extracted in March 2015
The measured asphalt thicknesses are listed in Table D 14.
Table D 14: Asphalt thicknesses
Total asphalt thickness (mm)

Chainage (m)
Outer wheel path Between wheel paths
110 95 –
120 100 –
130 101 112
140 92 –
150 94 –
160 103 111
170 103 –
180 100 –
190 118 113
Mean 101 112
The layer thicknesses varied from 92 to 118 mm, with a mean value of 101 mm in the outer wheel path,
which was very close to the design thickness (100 mm).
The air voids and indirect tensile modulus are given in Table D 15 and plotted in Figure D 27. The air voids in
the outer wheel path were about 3% below the values in the untrafficked area between wheel paths,
indicating the asphalt had compacted under trafficking. As expected, this additional densification was higher
for the top halves of the asphalt layer than the bottom halves.
Figure D 27 shows that, as expected, for a given air void level, the moduli in the trafficked area were lower
than the untrafficked area, which is consistent with fatigue damage to the asphalt layer.
Table D 16 lists the indirect tensile modulus of cores taken in the outer wheel path after two years in service.
For these cores, rather than cut the cores into top and bottom halves, about 10–20 mm was cut from the top
and bottom of the core, resulting in a test specimen generally 60–70 mm in length. As shown in Figure D 27,
these moduli were similar to those obtained after one year in service.

Table D 15: Asphalt field core results after one year in service
Top half of core Bottom half of core

Project chainage (m)
Air voids (%) Modulus(1) (MPa) Air voids (%) Modulus(1) (MPa)
Outer wheel path
130 2.5 7460 3.9 5750
160 3.7 – 5.2 –
190 4.7 5150 5.9 5340
Mean 3.6 6300 5.0 5550
Between wheel paths
130 5.5 6260 7.7 4910
160 7.0 5260 7.4 3690
190 8.2 5960 9.5 2890
Mean 6.9 5830 8.2 3830
1 Indirect tensile modulus at 25 °C and rise time of 40 milliseconds.
Table D 16: Asphalt field core in outer wheel path after two years in service
Air voids Modulus(1) Surface cracking at time of coring (from

(%) (MPa) Figure D 38)
110 3.9 7140 Nil
120 2.9 7230 Nil
140 3.7 6590 Nil
150 3.5 6640 Nil
170 2.8 7280 Nil
180 4.4 5340 Cracked
Mean 3.5 6700
1 Indirect tensile modulus at 25 °C and rise time of 40 milliseconds.
Figure D 27: Variation of asphalt indirect tensile modulus with air voids
8000 Cores, outer wheel path after 1 year in service
Cores, between wheel paths after 1 year in service
Cores, outer wheel path after 2 years in service

7000
Lab mixes 1998
y = 10007-686x
6000 R² = 0.48
Modulus
5000
(MPa)
E = 8665 -644.7AV
R² = 0.76
4000
3000
2000
2 3 4 5 6 7 8 9 10
Air voids (%)

The mean indirect tensile modulus at a temperature of 25 °C of the untrafficked asphalt cores, after
12 months curing in the roadbed, was about 4800 MPa at about 7.0% mean air voids. These results are
reasonably consistent with a mean modulus of 4430 MPa at a mean 8.6% air voids measured on
laboratory-mixed and compacted asphalt from five Melbourne metropolitan region asphalt plants (Moffatt &
Jameson 1998).
In comparison with the untrafficked asphalt mean modulus of 4800 MPa after 12 months curing, the mean
modulus of untrafficked FBS cores after two years curing in the roadbed was 11 000 MPa, which is
substantially higher.
The materials from the three March 2013 outer wheel path asphalt cores were combined to enable
measurement of bitumen content and particle size distribution. This process was also undertaken using the
between wheel path cores. The results are given in Table D 17 and are consistent with VicRoads (2014b)
grading and bitumen content requirements for size 20 mm dense graded asphalt. The mix included
10% recycled asphalt pavement.
Table D 17: Particle size distribution and bitumen content of asphalt field cores
VicRoads (2014b) Code of Practice % passing sieve

Sieve size RC 500.01 Between wheel
Size 20 mix limits Outer wheel path
paths
19.0 mm 90–100 100 100
13.2 mm 75–88 85 89
9.5 mm 61–75 71 72
6.7 mm 49–64 57 59
4.75 mm 41–55 48 49
2.36 mm 27–41 35 36
1.18 mm 18–33 26 27
600 µm 12–25 20 20
300 µm 8–19 14 14
150 µm 5–13 9.3 9.4
75 µm 3–6 6.7 6.8
Bitumen content 4.0–6.5 4.7 4.8
D.12 Surface Cracking of Foamed Bitumen Stabilised Section
Visual Inspection
In March 2014 after 12 months trafficking, the FBS section was similar in appearance to that observed a
month after construction: there was no cracking. However, when inspected in March 2015, February 2016,
February 2017 and March 2018 cracking had developed as recorded in Figure D 28.

Figure D 28: Surface cracking of the FBS in March 2015, February 2016, February 2017 and March 2018
Note that in May 2014 after about 14 months trafficking, a fine longitudinal crack was observed in the outer
wheel path near chainages 15 m and 30 m (Figure D 29). Note that this cracking was not apparent in
February 2016 or February 2017or March 2018, possibly being hidden by flushing/embedment of the seal
(Figure D 29).
In May 2014 there was also a different type of crack near chainage 98 m, which subsequently developed into
small block cracking (Figure D 30).
Figure D 29: Longitudinal cracking observed in the outer wheel path near chainage 30 m
May 2014, fine longitudinal cracking February 2016, cracking not apparent

Figure D 30: Cracking observed in the inner wheel path near chainage 98 m
May 2014 Close up February 2016
Figure D 31 is an example of the cracking observed in the outer wheel path at 80–85 m chainage in
November 2014, March 2015, February 2016, February 2017 and March 2018. The surface cracking
commences as longitudinal followed by transverse cracking, which may in the future lead to small block or
crocodile cracking. However, the extent and severity of this cracking has changed gradually between
November 2014 and March 2018. It is noted that the construction joint between the two 2.4 m wide
stabilisation passes is within 300–400 mm of the centre of the outer wheel path. This may have contributed
to the formation of the longitudinal cracking in the outer wheel path, particularly if the lime and bitumen
contents were higher or lower in this area of overlap.
Figure D 31: Example of the longitudinal and transverse cracking in outer wheel path between chainages
80 to 85 m
November 2014 March 2015

February 2016 February 2017
March 2018
During the February 2016 visual inspection, the cracking in the inner wheel path had extended to cover
about a third of the trial length (Figure D 28). Even though this wheel path was not in the vicinity of a
construction joint, again the cracking commenced as longitudinal and progressed to crocodile/block cracking
in some limited areas (Figure D 32). Note that the extent of cracking was similar in February 2017 after an
additional 12 months of trafficking, however in the vicinity of 60–65 m several small blocks had developed.
Note this type of wheel path longitudinal cracking is in contrast with the cracking observed on Western
Freeway Ballan (Appendix F.9.5) and Port Wakefield Road (Appendix C.6):
• At Western Freeway trial, the fatigue cracking commenced as transverse cracks, then longitudinal cracks
formed between the transverse cracks which developed in block/crocodile cracking. This is a similar
mechanism to that observed for asphalt, except when severely cracked the blocks tended to be larger
than asphalt crocodile cracking.
• At Port Wakefield Road, the fatigue cracking commenced as transverse cracks which developed into
crocodile cracking. However, the Port Wakefield Road trial had high bitumen contents and hence the
cracking characteristics may not be representative of FBS pavements.
The fine longitudinal cracking now developing in the Newell Highway trial (Figure E 36) has similar attributes
to the Calder Freeway, albeit there is some doubt about the influence of reactive subgrades at the Newell
Highway site. It is noted that Newell Highway projects had higher lime contents than the Western Freeway
and Port Wakefield Road projects (Table 7.1). This is discussed further in Section 7.2.

Figure D 32: Example of the cracking observed in the inner wheel path after 3 years trafficking
At chainage 70 m looking south Crocodile cracking at chainage 64 m
Investigation of Cause of Cracking
A total of 28 slabs were saw cut between the wheel paths to enable future flexural modulus and fatigue
testing under Austroads project TT2046 Improved Cost-effectiveness of Foamed Bitumen Stabilised
Pavements. During the excavation, the top 20–40 mm of the FBS layer delaminated in three of the 28 slabs
(Figure D 33). This delamination was unexpected as it did not occur in any of the many cores taken during
the course of the project.
As shown in Figure D 34, in addition to the slabs with a delaminated top layer, for another seven slabs fine
horizontal cracks at a 20–40 mm depth below the surface were apparent. These cracks may eventually lead
to delamination. In total about a third of the slabs were either delaminated or had fine horizontal cracks just
below the surface. During construction (Appendix D.7.1), the top of FBS was reworked to achieve the design
surface level. This process may have resulted in the upper 20–40 mm being susceptible to delamination in
some areas.
Figure D 33: Examples of delamination of top 20–40 mm of FBS

Figure D 34: Extent of delamination and horizontal cracking
Inner wheel path
1 5 9 13 17 21 25
Delaminated
2 6 10 14 18 22 26
3 7 11 15 19 23 27
Crack visible
4 8 12 16 20 24 28
Chainage 26 25 24
Outer wheel path
As one of the prime objectives was to determine whether fatigue cracking was a distress type for FBS
pavements, it was important to determine whether the observed surface cracking in the wheel paths
(Appendix D.12.1) was due to delamination of the top 20–40 mm. Hence in February 2016, cracked areas in
the wheel paths were cored to establish the depth to which the cracking occurred.
The locations of the eight core holes are shown in Figure D 28: five cores (labelled A, B, C, D and H) were
located on cracks in the inner wheel path, the remaining three cores (labelled E, F and G) were taken in the
outer wheel path. The coring was undertaken using compressed air and dry ice so that the cracks were not
filled with wet plastic fines.
Tracing the surface cracking below the surface generally proved difficult as seen from Figure D 35 and
Figure D 36, in part due the fine surface cracking at some sites and also the presence of fines which tended
to fill the fine cracking below the wider surface cracking. Water applied to the core surface assisted in tracing
the cracking in some cases.

Figure D 35: Cores taken in the inner wheel path (H is included in the next figure)

Figure D 36: Outer wheel path cores E, F and G and inner wheel path core H
The cracking below the surface was most apparent for core D (Figure D 35) and clearly indicated the
cracking extended well below the FBS surface. In none of the cores was the cracking obviously associated
with delamination of the top 20–40 mm of FBS.

The outer wheel path core taken at chainage 84 m (core F) fell apart during coring. As mentioned previously,
the longitudinal construction between the two passes of the stabiliser was near this wheel path. However,
delamination of the top 20–40 mm of FBS was not apparent.
In summary, the FBS cracking was not observed to be confined to the top 20–40 mm of FBS. For the cores
where surface cracking could be traced below the surface, the wider surface cracking appeared to extend to
the bottom of the FBS layer.
Although the outer wheel path cracking may have been influenced by construction issues near the
longitudinal construction, this was not the case for the widespread cracking in the inner wheel path. It was
concluded that the most likely cause of this cracking was full-depth load-induced fatigue.
D.13 Surface Cracking of Asphalt Section

When the asphalt section was inspected in March 2014, about one year after construction, fatigue cracking
of the 100 mm thick asphalt section was observed. Figure D 37 shows an example of cracking observed
after 1 and 2 years trafficking at the same chainage.
Figure D 37: Fatigue cracking observed on outer wheel path between chainage 155 to 160 m
March 2014 (1 year) March 2015 (2 years)
When the pavement was inspected in March 2015 after two years trafficking, fatigue cracking was observed
for about 30% of the length (Figure D 38). VicRoads removed and replaced the asphalt with a polymer
modified binder mixture in early 2016.

Figure D 38: Cracking of the asphalt section in March 2015
D.14 Rutting and Roughness

At five weeks, one, two, three and four years after construction of the FBS crushed rock base, the rutting and
roughness of the two sections were measured with a laser profiler.
The rutting results are shown in Figure D 39. As mentioned in Appendix D.7.1, the pavement ravelled and
rutted when trafficked overnight prior to sealing. Note particularly the high rutting of the inner wheel path.
During construction, considerable effort was applied to the inner wheel path to achieve a tight FBS base
surface suitable for bituminous sealing. This reworking of the base may have led to an excess of fines on the
surface or higher moisture contents resulting in early-life ravelling and rutting. By comparison, the area in the
vicinity of the outer wheel path did not require the same effort to prepare and showed significantly lower
ravelling and rutting.

Figure D 39: Rutting of the FBS and asphalt pavements
18 Foamed bitumen stabilised

Dense graded asphalt
crushed rock base
16
Outer wheel path 2013

14
12
10 Outer wheel path 2017
Rut depth
(mm)
8
0
0 20 40 60 80 100 120 140 160 180 200
Chainage (m)
24 Foamed bitumen stabilised

Dense graded asphalt
crushed rock base
22
20
18
Inner wheel path 2013
16 Inner wheel path 2014
14 Inner wheel path 2015
Rut depth Inner wheel path 2016
(mm) 12
Inner wheel path 2017
10
0
0 20 40 60 80 100 120 140 160 180 200
Chainage (m)

Note that rutting occurred despite the mix design indicating that the initial modulus of the Marshall
compacted specimens (Appendix D.4.2) had achieved the minimum 700 MPa required by the TMR mix
design method (TMR 2017b). This may have been due to the field moisture contents in localised areas being
in excess of the moisture contents used in the mix design.
In terms of the FBS section, the rutting remained relatively constant during the first two years of trafficking.
However, the results in years 3 and 4 did show increases in rutting, but on average the rut depth is
increasing only about 1 mm every 2 years. As such it will be cracking that is limiting the pavement service life
rather than rutting.
In terms of the asphalt section, the rutting did not change markedly before this trial section was removed and
replaced in January 2016 except for the areas which were fatigue cracked: moisture ingress is the likely
cause of the rutting increase.
In April 2013, five weeks after construction, the roughness results were as follows:
• FBS section: lane IRI of 2.4 m/km
• asphalt section: lane IRI of 3.6 m/km.
Note, however, the roughness results only apply to short (< 100 m) sections and are influenced by
construction joints and as such are not a reliable indicator of performance. There was no significant increase
in roughness after 4 years of trafficking.
D.15 Effect of Temperature on Deflections and Moduli

Surface deflections were monitored on various projects to evaluate the extent to which the modulus changes
with time and traffic loading. As the FBS material can differ in temperature between deflection
measurements, a procedure was required to adjust the measured deflections from the pavement
temperature at the time of deflection measurement to the WMAPT. Accordingly, a procedure similar to the
adjustments currently in use for asphalt (Austroads 2011a) was developed.
To assist in this regard, after three days oven drying at a temperature of 40 °C, the 1-month-old field cores
were tested for indirect tensile modulus over a range of temperatures. The results are plotted in Figure D 40.
The modulus variation is consistent with the FBS modulus variation given in Austroads (2011a).
The variation of modulus with temperature appears to be about one-third the asphalt modulus variation
provided in Austroads (2012).
As an interim measure, pending further research, it is proposed that the asphalt deflection and curvature
adjustment factor in Austroads (2011a) be used for FBS pavements, but with the ratio of WMAPT to the
measured temperature (Tmeas) adjusted to reflect the lower temperature dependency. The adjusted
WMAPT/Tmeas ratios are shown in Figure D 41.

Figure D 40: FBS modulus variation with temperature

12000
20m top portion
20 m bottom portion
40 m top portion
10000
40 m bottom portion
60 m top portion
60 m bottom portion
8000 80 m top portion
80 m bottom portion
Indirect Mean
tensile
6000
modulus
(MPa)
4000
2000
0
10 12 14 16 18 20 22 24 26 28 30
Temperature (°C)
Figure D 41: Correction of WMAPT to measured temperature ratios
1.8
1.7 FBS with sprayed seal surface

FBS thickness 90% of total thickness
1.6
FBS thickness 80% of total thickness
1.5 FBS thickness 70% of total thickness
1.4 FBS thickness 60% of total thickness
1.3
WMAPT/Tmeas
1.2
for use in
deflection/curvature 1.1
adjustment
1.0
0.9
0.8
0.7
0.6
0.5
0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6
WMAPT/Tmeas at time of deflection measurements
In the case of pavements with asphalt surfacing on the FBS layer, the adjustment varies according to the
ratio of the FBS layer thickness to the total thickness of the asphalt and FBS layers. In using the charts in
Austroads (2011a), the total thickness of asphalt and FBS layers is used with the adjusted WMAPT/Tmeas
ratio.

D.16 Traffic Loading

VicRoads provided the 2011, 2012, 2013, 2014, 2015 and 2016 heavy vehicle volumes for Calder Freeway
at Gisborne (14 km from the Woodend trial section), as given in Table D 18.
Table D 18: Traffic volume for the Calder Freeway northbound carriageway
Estimated annual
Annual ESA in slow
Year Total heavy vehicles(1) Number of survey days number of heavy
lane
vehicles
2012 3.8 x 105 343 4.0 x 105 9.3 x 105
2013 3.0 x 105 311 3.6 x 105 8.8 x 105
2014 3.1 x 105 296 3.8 x 105 7.1 x 105
2015 3.8 x 105 346 4.0 x 105 7.5 x 105
2016 1.9 x 105 181 3.8 x 105 7.1 x 105
1 ‘Total heavy vehicles’ assumes all ‘rejected’ and ‘unknown’ vehicles are heavy vehicles.
Source: Calder Freeway Gisborne weigh-in-motion site data supplied in VicRoads emails in 2013, 2015 and 2017.
Table D 19 lists the average equivalent standard axles (ESA) per heavy vehicle derived from the
weigh-in-motion (WIM). As the 2015 data appears low compared to previous years, it was assumed the
average ESA per heavy vehicle was the same as in 2014, that is 1.95 ESA/HV. This value was also
assumed to apply in 2016.
Based on this data and assuming 95% of the heavy vehicles travel in the slow lane, the estimated annual
ESA of traffic loading was calculated.
Table D 19: Results of average ESA per heavy vehicle on Calder Freeway
Survey period Average ESA per heavy vehicle

January 2012 – June 2012 2.38
July 2012 – December 2012 2.52
January 2013 – June 2013 2.01
January 2014 – June 2014 1.93
January 2015 – June 2015 1.23
Source: Calder Freeway Gisborne weigh-in-motion site data supplied in VicRoads emails in 2013 and 2015.
The cumulative ESA of loading since opening to traffic in March 2013 is plotted in Figure D 42.

Figure D 42: Predicted cumulative traffic loading for trial section
3.0E+06
2.8E+06
2.6E+06
2.4E+06
2.2E+06
2.0E+06
1.8E+06
Cumulative 1.6E+06
ESA 1.4E+06
1.2E+06
1.0E+06
8.0E+05
6.0E+05
4.0E+05
2.0E+05
0.0E+00
0 0.5 1 1.5 2 2.5 3 3.5 4
D.17 Comparison of Observed and Predicted Performance

A key objective of the project was to improve the Austroads interim thickness design procedure for FBS
pavements (Austroads 2011a).
As stated in Appendix D.6, the allowable traffic loading was predicted using the interim method: the predicted
FBS fatigue life was 1.1 x 106 ESA at 50% design reliability. As shown in Figure D 42, this traffic loading is
reached after about 1½ years. Consequently, the design model predicts there is a 50% chance of observing
FBS fatigue cracking after 1½ years trafficking (early 2016).
When the pavement was inspected in March 2014, one year after opening to traffic, fatigue cracking was not
observed. However, as mentioned in Appendix D.12, during the second year of trafficking, a significant
extent of fine longitudinal cracking and transverse (Figure D 28) in the outer wheel path was observed. As
discussed in Appendix D.12.1, a longitudinal construction joint may have influenced the development of
cracking. In contrast, this joint does not affect the performance of the inner wheel path. After two years,
longitudinal cracking was observed along about 5–10% of the length of the inner wheel path. After a further
year of loading, the cracking covered more than 30% of the length.
Based on the performance of the inner wheel path, it was concluded that for locations removed from the
influence of construction joints the predicted allowable traffic loadings using the interim method are in
reasonable agreement with the observed performance.

D.18 Calder Freeway Trial Summary

In March 2013, a trial section was constructed on the Calder Freeway, Woodend, Victoria. The findings to
date are:
• The trial section was constructed to a lower standard of compaction compared to the Western Freeway
and Newell Highway trials, with about a mean 96% modified Proctor MDD. There was a significant
reduction of density with depth, with about a 5% difference in compaction between the top and bottom of
the FBS layer. This deficiency in compaction may be related to the compaction equipment used,
specifically the lack of a heavy padfoot roller.
• The average depth of stabilisation was about 195 mm, well in excess of the 150 mm design thickness.
• As a result of the greater depth of stabilisation, the measured bitumen contents of materials sampled
during construction indicated the amount of bitumen added was about 2.8%, significantly below the
3.5% design value. Similarly, the quicklime content was 1.1%, which was below the design value of 1.5%.
The active lime content was low, 66% of CaO.
• The Marshall mix design moduli measured prior to construction were well above the TMR minimum
requirements. The Marshall mix design was repeated after construction using untreated crushed rock
sampled during construction. Again, the mix conformed to the TMR minimum modulus requirements.
• Early-life FBS cores taken along the edge line and between wheel paths one month after construction
had mean indirect tensile moduli considerably lower than the Western Freeway and Newell Highway
trials. The mean untrafficked moduli were about 3500 MPa and 1700 MPa for the top and bottom of the
FBS layer. These moduli more than tripled two years after construction. Based on FWD surface deflection
measurements, this modulus increase did not occur in the wheel paths, presumably due to the
load-induced fatigue damage.
• When the pavement was opened to traffic the day of construction, the surface ravelled with deep rutting
at the start of the section. As a result, the ride quality was well below the high standard subsequently
achieved on the Newell Highway and Western Freeway trials.
• In November 2015, slabs were cut from between the wheel paths for laboratory testing. Unexpectedly, the
top 20–40 mm of FBS delaminated from the underlying FBS in some areas. This delamination has not
been observed during pavement coring. A possible reason for the delamination was the need to rework
the surface during construction to bring the surface up to the design level.
• About a year after opening to traffic, fine longitudinal cracking was observed in the outer wheel path. After
three years (March 2016) of trafficking the extent of this longitudinal cracking developed to cover about a
third of the section length. Fine transverse cracking has now developed, yet not to the extent that it has
led to significant areas of block or crocodile cracking. The longitudinal crack in the outer wheel path is
near the construction joint between the two 2.4 m wide stabilisation passes. This may have contributed to
the formation of the longitudinal cracking, however this joint has not influenced the inner wheel path
cracking.
• After a further 12 months trafficking (February 2017), the extent of cracking did not increase markedly.
• The nature of the cracking differed from the initial transverse fatigue cracking observed at the Western
Freeway and Port Wakefield Road trials. In some isolated areas, the cracking has progressed to small
block cracking.
• Due to the observed delamination of the top 20–40 mm in some areas, cracked areas were cored to
determine whether the surface cracking was due the delamination. It was concluded that the cracking
extended well below the surface, consistent with classical full-depth fatigue cracking.

Appendix E Newell Highway, Bellata
E.1 Introduction
In October 2013, Roads and Maritime Services (Roads and Maritime) New South Wales agreed to support
the Austroads project by constructing a trial section on the Newell Highway, approximately 1 km south of the
town of Bellata, between Moree and Narrabri in north-western New South Wales. The trial section was
constructed as part of more than 8 km of foamed bitumen stabilisation of the Newell Highway in this region.
The northbound trial section starts at chainage 46.053 km and finishes at 46.153 km, where zero chainage is
the Tibbereena Street (Newell Highway)-Doyle Street intersection in Narrabri. Chainage 46.153 km is about
190 m south of the southern end of the guardrail of the Myall Hollow Creek bridge.
A total width of 5.5 m was stabilised comprising a 3.5 m wide traffic lane and 2.0 m wide outer shoulder.
Details of the Newell Highway FBS pavement are given in Table E 1. The annual average rainfall in nearby
Bellata is about 640 mm. The total rainfall over the 41-month period from construction in October 2013 until
March 2017 was 1935 mm which equates to an annual rainfall of 570 mm, indicating a drier than average
period during the performance monitoring.
According to the Northcote Soil Classification map, the subgrade in this area is a reactive clay which is
locally known as ‘black soil’. It is characterised by high plasticity, low CBR, and high potential to swell when
wet and crack when dry. As discussed in Appendix E.6.5, during the prolonged dry period, longitudinal
cracking due to this reactive soil has developed in the FBS pavement adjacent to the trial site.
Figure E 1 shows a view of the trial site.
Table E 1: Description of the Newell Highway FBS pavement
Description
Location Chainage 46.053 to 46.153 km on Newell Highway northbound
(where chainage zero starts at Tibbereena St – Doyle St junction, Narrabri)
Chainage 46.153 km is 190 m south of the end of the guardrail of Myall Hollow Creek bridge
Number of lanes 1 (northbound)
Lane width 5.5 m (3.5 m traffic lane and 2.0 m shoulder)
Total area 550 m2
Stabilisation depth Nominal 175 mm FBS base: mean constructed thickness 165 mm
Foamed Mix design See discussion in Appendix E.3
stabilisation
Host materials 50 mm of imported dense granular base (size 20 mm)
specification
125 mm of existing natural gravel base
Bitumen Specified content: 3.5% Class 170 bitumen
Actual content: measured 2.7% (Appendix E.5.3), from bitumen used about 4%
Supplementary binder Specified content: 1.5% quicklime
Actual contents:
2.6% for 2.4 m width closest to centre
1.7% for the outer 1.1 m of the traffic lane (Appendix E.5.2)
Construction Work specifications Roads and Maritime specifications
method and QA testing

Description
Number of strips 3 (2.4 m, 2.4 m and 0.9 m with 100 mm overlap)
stabilised
Construction date Foamed bitumen stabilisation, 4 and 5 October 2013
Initial sprayed seal placed five days after FBS, 10 October 2013
Final seal 11 March 2015
Surfacing Initial treatment: size 14 sprayed seal with scatter coat
Final treatment: double/double (14 mm/7 mm) with a total bitumen application
rate of 1.7 l/m2 .using an S45R (rubber) polymer modified binder
Figure E 1: Newell Highway site being prepared for FBS
Source: Zhang (2013).
E.2 Site Investigation Prior to Stabilisation
Traffic Loading
The Newell Highway is a heavily-trafficked inland route with truck traffic carrying freight from Victoria and
South Australia to Queensland.

A Roads and Maritime pavement design report (Roads and Maritime 2012a) provides design traffic
information about this section of the Newell Highway between Narrabri and Moree. From the report data, the
annual traffic loading in 2012 was calculated to be 8.7 x 105 ESA. From historical traffic count data, the
report calculated the annual heavy vehicle growth rate was a very low 0.8%. Using this growth rate,
Figure E 2 shows the predicted increase in cumulative traffic loading.
Figure E 2: Predicted cumulative traffic loading for trial section
5.0E+06
4.5E+06
4.0E+06
3.5E+06
3.0E+06
Cumulative
2.5E+06
ESA
2.0E+06
1.5E+06
1.0E+06
5.0E+05
0.0E+00
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Pavement Visual Condition Prior to Works
In August 2013, prior to the stabilisation, the pavement was inspected and photographed by Roads and
Maritime staff. Figure E 3 shows the pavement prior to treatment including areas of rutting and shoving of the
granular base.
Figure E 3: Pavement visual condition prior to treatment

Pavement Investigation Prior to Works
To investigate the material types and layer thicknesses of the pavement to be treated, in August 2013 a pit
was dug in the traffic lane near the edge line of the northbound carriageway at chainage 46.225 km, about
70 m north of the section subsequently identified for the trial section. The granular base at this location had
shoved (Figure E 4).
Figure E 4: Pavement condition at chainage 46.225 km, test pit
Figure E 5 shows the pit which was dug to a depth of 1650 mm below surface level. Based on this
excavation, Figure E 6 shows the pavement composition prior to treatment.

Figure E 5: Test pit at chainage 46.225 km
Figure E 6: Newell Highway pavement composition prior to treatment

As it was proposed to stabilise the pavement to a depth of 175 mm, after removing the 25 mm thickness of
sprayed bituminous seals, granular material was excavated 25–200 mm below the surface for subsequent
laboratory testing (Appendix E.2.5). This material was described on site as a sub-rounded to rounded sandy
natural gravel. The remaining sandy gravel material was then excavated and sampled down to the top of a
select fill layer, a depth of 550 mm below the surface. The gravelly clay fill below the sandy gravel was
sampled at two depths, 550–950 mm and 950–1420 mm. The top of the silty clay subgrade was found at a
depth of 1420 mm below the surface. Samples of the silty clay were taken at depths of 1420–1650 mm
below the surface.
Estimated In Situ CBR of the Natural Subgrade
Dynamic cone penetrometer testing was undertaken on the silty clay subgrade. The estimated in situ
Californian Bearing Ratio (CBR) of the top 400 mm of subgrade (depth 1420–1820 mm below the surface)
was 4%. The laboratory soaked CBR of this material was very low, CBR = 1% (Table E 2).
Laboratory Testing of Excavated Materials
The excavated materials were tested for:

• particle size distribution
• liquid limit, plastic limit and plasticity index
• MDD and OMC using standard Proctor compaction effort
• in situ moisture content
• four-day soaked laboratory Californian Bearing Ratio (CBR).
The test results are given in Table E 2.
Table E 2: Particle size distributions of pit samples prior to treatment
Silty clay
Sandy gravel Gravelly clay fill
subgrade
Test method Portion Description Depth Depth Depth Depth Depth
25–200 200–550 550–950 950–1420 1420–1650
mm mm mm mm mm
Roads and Whole Passing 37.5 mm sieve (%) 100 100 100 99 99
Maritime test sample
Passing 26.5 mm sieve (%) 98 98 98 91 98
method T106
Passing 19.0 mm sieve (%) 93 94 89 77 97
Passing 13.2 mm sieve (%) 79 87 77 62 92
Passing 9.5 mm sieve (%) 68 80 71 53 90
Passing 6.7 mm sieve (%) 59 73 65 45 89
Passing 4.75 mm sieve (%) 52 67 61 41 87
Passing 2.36 mm sieve (%) 40 54 53 34 85
Roads and Whole Passing 425 µm sieve (%) 19 25 43 26 78
Passing 75 µm sieve (%) 7.0 7.5 26 22 61
method T107
Passing 13.5 µm sieve (%) 4.5 4.0 18 14 51
AS 1289.3.1.2 Fraction Liquid limit (%) – – 32 44 65
passing
AS 1289.3.2.1 Plastic limit (%) – – 13 14 16
425 µm
AS 1289.3.3.1 sieve Plasticity index (%) – – 19 30 49
Roads and Fraction Maximum dry density (t/m3) 1.820 1.656 1.977 1.885 1.483
Maritime test passing
Optimum moisture content 14.9 19.1 11.8 13.4 25.5
method T111 19 mm
(%)
sieve

Silty clay
Sandy gravel Gravelly clay fill
subgrade
Test method Portion Description Depth Depth Depth Depth Depth
25–200 200–550 550–950 950–1420 1420–1650
mm mm mm mm mm
Roads and Whole California Bearing Ratio (%) 50 25 6 6 1
method T117
Road and Whole In situ moisture content (%) 8.9 13.7 10.8 11.0 23.4
Percentage of OMC (%) 60 72 92 82 92
method T120
The particle size distribution of the top 175 mm of sandy natural gravel is shown in Figure E 7 compared to
the Roads and Maritime R76 specification limits for material to be FBS. The material was natural gravel with
rounded coarse aggregate particles, which was the cause of the shoving and rutting of the pavement
(Figure E 3). Consistent with the treatment for the remainder of the 8 km of FBS works along the Newell
Highway, it was decided to re-sheet the pavement with 50 mm thickness of dense graded crushed rock base
prior to foamed bitumen stabilisation.
Figure E 8 shows the grading of the combined material sampled during construction, as described in
Appendix E.5.1. This material was a mixture of the imported dense granular base, natural gravel and existing
sprayed seal surfacing.
Figure E 7: Particle size distribution of top 175 mm sandy gravel sampled in August 2013 prior to
construction
100
Sandy gravel base
90 R76 lower limit
R76 upper limit
80
70
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)

Figure E 8: Particle size distribution of the material stabilised and sampled during construction
100
Oct 2013 sample from road bed, tested by Downer
90
Oct 2013 sample from road bed, tested by TMR
80
R76 lower limit
70
R76 upper limit
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)
Note the very high plasticity (PI = 49) of the soil deep (> 1400 mm) below the pavement. This is the reactive
black soil that is the likely cause of the longitudinal cracking in the vicinity of the trial (Appendix E.6.5).
E.3 FBS Mix Design
Roads and Maritime
Prior to construction, Roads and Maritime undertook mix design testing of material sampled from the
roadbed about 500 m away from the trial section. Roads and Maritime advised that the material was similar
to material in the trial section, except that it included 20% by mass of imported dense graded base rather
than about 28% used in the trial section.
Some aspects of the Roads and Maritime mix design process varied from that of Downer Infrastructure
(Appendix E.3.2) and TMR (Appendix E.3.3), as follows:
• Before and after mixing, the material to be compacted was held at a temperature of 35–40 °C, whereas
TMR and Downer prepared specimens at laboratory temperature of about 23 °C.
• Two per cent hydrated lime was used in the laboratory testing which is consistent with TMR mix design
and similar to the 1.5% quicklime used by Downer.
• The mixture was compacted into 100 mm internal diameter moulds using a standard 4.54 kg Marshall
hammer. This method is not included in the proposed Austroads mix design method. It is proposed that
the TMR method be used, namely compaction into 152 mm diameter mould and a 10.2 kg Marshall
hammer.
• To obtain the cured wet modulus, the specimens were soaked for 24 hours in water, whereas both TMR
and Downer soaked specimens for 10 minutes under partial vacuum.

Table E 3 lists the results. Roads and Maritime concluded that the material of 3.5% bitumen and
2% hydrated lime met the minimum mix design requirements and decided to adopt the following design:
• 3.5% bitumen
• 1.5% quicklime (equivalent to about 2% hydrated lime).
These binder contents were consistent with mix designs used elsewhere on the Newell Highway and also
similar to designs used in other projects in Australia at the time.
Table E 3: Roads and Maritime mix design for Newell Highway
Mean indirect tensile modulus (MPa) at 25 °C

Bitumen Hydrated Mean dry Ratio
content lime content density(4) Cured dry Cured wet wet/dry
(%) (%) (t/m3) Initial modulus(1) modulus
3.0 2.0 1.98 1410 3450 1140 0.33
3.5 2.0 2.14 1270 6090 3580 0.59
4.0 2.0 2.04 1110 6590 4440 0.67
Roads and Maritime – 700 4000 2000 0.50
minimum requirements
3 Soaked in water for 24 hours at 25 °C, no vacuum.
Downer Infrastructure Testing
During construction in October 2013, untreated material was sampled from the pulverised pavement for mix
design testing by Downer Infrastructure and by TMR (Appendix E.3.3).
The unbound base material (Figure E 9) consisted of coarse-to-fine aggregates (a mixture of crushed rock
and sandy gravel), bitumen coated sprayed sealing aggregates, coarse-to-fine sands, and clays.
Figure E 9: Untreated material used in Downer mix design testing

From this testing, Downer concluded that the untreated granular base was non-plastic, with a particle size
distribution as shown in Figure E 8. Using a 150 mm diameter mould, the standard Proctor MDD of the
untreated material was 2.01 t/m3 and the OMC was 10.0% (Figure E 10).
Figure E 10: Untreated and treated material standard Proctor compaction curves
2.04
2.02
2.00 Untreated materials, 150 mm mould
1.98 3.5% bitumen, 150 mm mould
1.96 3.5% bitumen, 100 mm mould
1.94
1.92
1.90
1.88
Dry 1.86
density 1.84
(t/m3) 1.82
1.80
1.78
1.76
1.74
1.72
1.70
1.68
1.66
1.64
3 4 5 6 7 8 9 10 11 12 13
Standard Proctor MDD and OMC values were also measured for the granular base stabilised in the
laboratory with 3.5% foamed bitumen and 1.5% quicklime and compacted into 100 mm and 150 mm
diameter size moulds. Figure E 10 shows that the MDD decreased to about 1.89 t/m3 after treatment with
lime and foamed bitumen, about 6% lower than the MDD of the untreated material. By comparison, in
subsequent testing of field-mixed stabilised materials retrieved during construction, the MDD was 1.84 t/m3
and the OMC was 11.1% (Table E 11). In this case, there was a 3.5 to 6 hour delay between sampling the
field-mixed material and commencement of laboratory compaction. This time delay may have resulted in
reduced ability to compact the materials and hence the lower MDD and higher OMC measured during
construction. The effect of a 24-hour delay is discussed in Section 6.
To prepare the modulus test specimens, water was firstly added to untreated base. The mixing moisture
content was about 75–80% of standard Proctor OMC of the untreated material (10% water by mass), which
was similar to the moisture content of material sampled from the roadbed (Table E 11). An additional
0.5% moisture was added to slake the quicklime.
Quicklime (1.5%) was then added and the lime-treated material was left to cure for 45 to 60 minutes before
addition of the foamed bitumen.
The foamed bitumen was then added to the lime-treated material in a laboratory pugmill. The Class 170
bitumen used to stabilise the material was optimised for foaming by addition of 0.6% by mass of Teric 311
foaming agent. Table E 4 shows the trial bitumen contents used.

Table E 4: Downer Infrastructure mix design results of 150 mm diameter specimens compacted with
Servopac
Mean indirect tensile modulus (MPa) Cured wet

Mean
Bitumen Quicklime Moisture at 25 °C Ratio indirect
dry
content content content wet/dry tensile
density(4) Initial Cured dry Cured wet
(%) (%) (%) modulus strength at
(t/m3) modulus(1) modulus(2) modulus(3) 25 °C (kPa)
3.0 1.5 7.6 2.08 970 2650 1480 0.56 430
3.5 1.5 7.7 2.06 920 2860 1510 0.53 500
3.7 1.5 7.8 1.93 – 1360 550 0.41 –
4.0 1.5 7.9 2.06 910 2420 1620 0.68 515
Downer minimum requirements – – 1500 – –
Modulus test specimens (150 mm diameter) were prepared using a Servopac gyratory compactor with an
applied stress of 235 kPa and 3° angle of gyration. The specimens were then tested for indirect tensile
modulus and indirect tensile strength. In this initial modulus testing, using bitumen contents of 3.0%, 3.5%
and 4.0%, the target dry density was 2.10 t/m3, which corresponded to about 105% standard Proctor MDD of
the untreated material rather than the intended 105% of the MDD of the treated field-mixed material
(i.e. 1.844 x 1.05 = 1.94 t/m3). Due to this error in specimen density, an additional three specimens, with
nominal 3.5% bitumen content, were compacted to a target dry density of 1.94 t/m 3 close to the mean field
density (Table E 11).
The results are given in Table E 4. The test specimens were compacted to a mean density of 1.93 t/m3 near
the mean field density and 3.7% bitumen content. The mean cured wet modulus was 550 MPa, well below
Downer’s minimum modulus for mix acceptance of 150 mm diameter gyratory specimens (1500 MPa).
In addition to testing the 100 mm diameter specimens using gyratory compaction, specimens where
compacted by applying 50 blows of a 4.9 kg Marshall hammer. Table E 5 lists the mean moduli of testing
100 mm diameter Marshall and gyratory specimens. Differences in moduli due to compaction methods are
discussed in Appendix 4.3.
Table E 5: Downer Infrastructure mix design results of Marshall and gyratory 100 mm diameter specimens
Mean indirect tensile

Bitumen Quicklime Moisture Mean dry modulus (MPa) at 25 °C Ratio
Compaction method content content content density(3) wet/dry
(%) (%) (%) (t/m3) Cured dry Cured wet modulus
Marshall 100 mm
3.5 1.5 7.7 2.11 2570 1710 0.67
diameter
Gyratory 100 mm
3.5 1.5 7.7 2.11 2220 1390 0.63
diameter

Queensland Department of Transport and Main Roads (TMR) Testing
In January 2014, mix design testing was also undertaken by TMR using untreated material sampled from the
roadbed during construction. This testing was undertaken in accordance with the proposed Austroads mix
design process.
The grading of the untreated material was similar to the sample tested by Downer (Figure E 8) and the
standard Proctor MDD of 2.01 t/m3 and OMC of 10.5% of the untreated material were also similar to the
values obtained by Downer (Appendix E.3.2). However, whereas the Downer testing had indicated the
material was non-plastic, the TMR testing measured a plasticity index (PI) of 5% and a linear shrinkage of
5%. Downer advised that a possible reason for this difference in PI was as follows:
This may be due to allowance in the Australian Standard method, in line with most
international methods, to classify material not able to be rolled to an initial 3 mm thread as
non-plastic. The TMR test method varies from this and states that at any point, if it cannot
be rolled to a 3 mm thread, this should be taken as its plastic limit. This needs to be
reviewed as it will affect classifications of marginal materials.
The same issue occurred in testing of the Calder Freeway, Woodend project (Appendix D.3).
Following TMR practice, the modulus test specimens were compacted into 150 mm diameter moulds by
applying 50 blows to each end of the specimen with a heavy (9.92 kg) Marshall hammer.
The FBS mixes were all mixed at a target moisture content of 70% standard Proctor OMC of the untreated
material, that is a target moisture content of 7.4%, which was significantly lower than the field moisture
content when compaction commenced (8.0%, Table E 11). The moduli, densities, bitumen, hydrated lime
and moisture contents are shown in Table E 6. Note that at bitumen contents of 3% and 4%, the mean dry
density of the Marshall specimens was about 106% of standard Proctor MDD (1.87 t/m 3, Appendix E.3.2).
Table E 6: TMR mix design testing
Hydrated Mean indirect tensile modulus (MPa)

Bitumen Moisture Mean dry at 25 °C Ratio
Compaction lime
method content Initial Cured dry Cured wet
150 mm 2.0 2.0 7.3 2.03 1080 2070 870 0.42
Marshall mould
3.0 2.0 7.1 1.99 1160 2720 1090 0.40
and large
Marshall 4.0 2.0 7.2 1.98 1150 2910 1260 0.43
hammer
150 mm mould 3.0 2.0 7.1 1.98 440 1110 480 0.43
gyratory
compaction
TMR minimum base requirements (> 1000 ESA/day) for 700 4000 2000 0.50
specimens compacted in 150 mm diameter mould using heavy
(9.9 kg) Marshall hammer
Based on these test results, TMR considered the material not suitable for foamed bitumen stabilisation.
In addition to testing Marshall specimens, 3% FBS specimens were compacted in a 150 mm diameter mould
using gyratory compaction with the number of gyratory compaction cycles selected to result in similar dry
density as the Marshall specimens. Table E 6 shows that the gyratory specimens had less than half the
modulus of the Marshall specimens as discussed further in Section 4.3.1.

Note that:
• The dry densities of the Marshall specimens manufactured with 3% and 4% bitumen were about 2%
above the mean field density (Table E 11). However, the presaturation method was used to measure the
dry density of the Marshall specimens. Had the mensuration method been used, the densities would have
been 2–3% lower (Section 4.3.3) and similar to the field values.
• The TMR moduli of 3% FBS specimens compacted into a 150 mm diameter mould using gyratory
compaction were less than half the Downer values (Table E 4) which, at least in part, could be due to
TMR test specimens being about 5% lower in density than the Downer specimens.
• In terms of the TMR 150 mm diameter Marshall specimens, the mean density and mean moduli are
similar to the values measured for the 100 mm diameter specimens prepared by Roads and Maritime
from the field-mixed material using a Marshall hammer (Table E 13).
E.4 FBS Thickness and Predicted Allowable Traffic Loading

Except for the trial section, a foamed bitumen stabilisation depth of 250 mm was used for the 8 km Newell
Highway rehabilitation project. In order to increase the likelihood of FBS distress within the period of the
research project (2012–17), Roads and Maritime agreed to reduce the FBS thickness for the trial site to a
target of 175 mm. Based on coring post-construction (Appendix E.6.3), the mean constructed thickness was
slightly less, 165 mm. The resulting pavement composition after FBS is illustrated in Figure E 11.
Figure E 11: Newell Highway nominal pavement composition after FBS
Using the interim thickness design method (Austroads 2011a), the allowable traffic loading was calculated.
This calculation was undertaken despite the mix design testing concluding that the Newell Highway mix had
modulus values below the minimum acceptable values (Table E 6).
The interim thickness design method (Austroads 2011a) states that the FBS design modulus is determined
and then soaked in water under vacuum for 10 minutes. Consequently, the FBS design modulus was
determined from the mix design cured wet modulus (Table E 6). Using the Austroads (2011a) method, the
cured wet modulus was adjusted to the in-service temperature (WMAPT of 31 °C) and in-service loading rate
(design traffic speed of 80 km/h). The resulting FBS design modulus was low, 870 MPa (Table E 7).
All the design input values are listed in Table E 8.
Table E 7: Newell Highway calculation of FBS design modulus
Mix design
Heavy vehicle
IDT modulus at 25 °C 40 Mean dry Mean dry WMAPT FBS design
design speed
ms rise time density(1) density(2) (°C)
(km/h)
modulus (MPa)
(MPa) (t/m3) (t/m3)
1090 (wet) 1.99 1.94 31 80 870

The predicted allowable traffic loading to FBS fatigue cracking (2.2 x 106 ESA) was calculated using:
• Equation 4 adjusted to 50% design reliability (Section 8.4.3)
• the FBS design modulus of 870 MPa.
Note that from the measured surface deflections a high modulus (> 1000 MPa) layer was back-calculated
under the FBS layer (Figure E 26), whereas the allowable loadings have been calculated assuming a
granular subbase with a top modulus of 200 MPa. If there is such a high modulus layer under the FBS, the
predicted allowable traffic loadings will be considerably higher. A future pavement investigation will provide
further information in this regard.
Table E 8: Data used to predict fatigue life of the Newell Highway pavement
Pavement Available
Design parameter Technical basis for data selection
component data
Subgrade Subgrade type Silty clay Based on in situ CBR estimated from dynamic cone
penetrometer measurements (Appendix E.2.4)
Select fill Fill thickness 870 mm Pavement investigation at chainage 46.225 km (Appendix E.2)
Fill type Gravelly clay Pavement investigation at chainage 46.225 km (Appendix E.2)
Design modulus 60 MPa 10-day soaked laboratory CBR of 6% (Appendix E.2.5)
(top sublayer)
Subbase Subbase thickness 360 mm Pavement investigation at chainage 46.225 km (Appendix E.2)
Design modulus 200 MPa Assume normal standard base material from Table 6.5 of GPT
(top sublayer) Part 2 (Austroads 2012)
FBS base FBS thickness 165 mm Based on 12 cores taken along the edge line (Appendix E.6.3)
Design modulus 870 MPa (wet) Table E 7
FBS volume of binder 7% Assuming 3.5% bitumen and a FBS dry density of 1.98 t/m3
(Table E 17)
Wearing Thickness N/A Sprayed bituminous seal
course
Design modulus N/A
E.5 Construction and Testing of the FBS Trial Section
Construction of the FBS Pavement
The construction was undertaken by Roads and Maritime and Downer Infrastructure. Downer was
responsible for the supply of all stabilisation materials (e.g. quicklime, bitumen, and foaming agent) and in
situ mixing with the stabiliser. All other activities were undertaken directly by Roads and Maritime.
The trial section was 100 m in length and 5.5 m in width, covering the northbound lane and 2 m of the sealed
shoulder.
The trial section was constructed in October 2013.
The construction sequence and sampling of materials were as follows:

• About a week before the stabilisation works, a 50 mm thick re-sheet of size 20 dense graded base was
placed on the existing sprayed seal in order to correct its shape and improve the particle angularity of the
existing base material (Figure E 12).

• On 4 October 2013, in order to provide untreated samples for laboratory mix design testing
(Appendix E.3.2, Appendix E.3.3), Wirtgen Stabiliser WR2400 was used to pulverise the re-sheet
material, the existing sprayed seal and base material to a nominal total depth of 175 mm. Although the
total width to be stabilised was 5.5 m, this untreated material mixing was limited to the outer 2.4 m width.
The resulting untreated material comprised 50 mm thickness of re-sheet and 125 mm thickness of the
existing seal and base. (For the remainder of Newell Highway works, pulverisation of the untreated
materials was not undertaken prior to stabilisation).
• Shortly after the mixing, about 400 kg of the untreated mixed material was sampled over the length of the
trial section for use in mix design testing.
• The untreated material was then shaped with a grader and lightly compacted with a multi-tyred roller in
preparation for stabilisation the next day (Figure E 13).
Figure E 12: Stockpiles of untreated materials
Size 20 dense graded base Existing sandy gravel base
Figure E 13: Shaping and compaction of the untreated materials on the day before stabilisation

• On 5 October 2013, quicklime was spread using three passes of the spreader – widths of 2.4 m (nearest
centreline), 1.56 m and 1.54 m. For a target quicklime content of 1.5%, a stabilisation depth of 175 mm
and a mean in situ density of 1.94 t/m3 (Table E 11), the target lime spread rate was about 5.16 kg/m2.
The spread rate was measured by placing three trays between the wheel paths to receive the lime
discharged from the spreader (Figure E 14) and weighing each tray to get the lime weight for rate
estimation. The measured spread rates were 8.31, 5.37 and 5.36 kg/m 2, respectively. Consequently, the
2.4 m width near the centreline had a spread rate of 8.31 kg/m2, equivalent to 2.6% by mass for the mean
constructed thickness of 165 mm (Appendix E.6.3), well in excess of the design value. However, from the
total quantities of lime used as measured on the spreader by the contractor, average spread rates of
5.54, 5.51 and 5.70 kg/m2 were calculated. Note that the mean measured lime content of the mixed
material sampled in the outer wheel path was 1.7% (Appendix E.5.2), which equates to a spread rate of
5.44 kg/m2 for the mean constructed thickness of 165 mm (Appendix E.6.3).
• Quicklime samples were also collected, one at about halfway through the spreading and another at the
end of the last run, and later tested for active lime content (Appendix E.5.2).
Figure E 14: Sampling to estimate lime spread rate

Figure E 15: Stabiliser mixing slaked lime with basecourse
Lime slaking Mixing the lime through the base
• Immediately after the lime was spread, it was slaked (Figure E 15). The slaked lime was then mixed
through the untreated material to a nominal depth of 175 mm using three passes of the stabiliser of the
following widths: 2.4 m (near centreline), an adjacent 2.4 m width, and a 0.9 m width in the shoulder. In
the second and third passes, there was 100 mm overlapping with the previously mixed width.
• The lime-treated material was then sampled at full layer depth at an offset of 3.0 m from the road
centreline (roughly in the outer wheel path) at chainages 20, 40, 60 and 80 m. Four 15 kg buckets of
material each were collected from these locations to enable the lime content to be measured
(Appendix E.5.2).
• The surface was then reshaped and lightly compacted with a 15.5 t multi-tyred roller in preparation for
mixing with the foamed bitumen.
• The lime reacted with the basecourse for about 1.5 hours (from 9 am to 10.30 am) while the bitumen was
being heated to 180 °C in the tanker, and samples of bitumen were taken (4 x 10 kg samples). The
foaming agent (0.6% Teric 311, about 85 litres) was added to 14 000 litres of bitumen in the tanker and
then connected to the stabiliser. The bitumen temperature in the tanker was measured at the start and
end of each of the three runs; the average temperatures of the runs were 182 °C, 187 °C and 188 °C.
• With the tanker and stabiliser connected (Figure E 16), the foamed bitumen was then mixed through the
lime-treated base, starting from the road centreline, with 2.4, 2.4 and 0.9 m run widths. The nominal depth
of stabilisation was 175 mm. From the quantity of bitumen used, the contractor advised the average
bitumen application rates were 12.05, 13.1 and 12.86 kg/m2, compared to the target 2.25 kg/m2. For the
mean constructed thickness of 165 mm (Appendix E.6.3), the estimated bitumen contents by mass were
3.8, 4.1 and 4.0% higher than the target value of 3.5%. Four 15 kg samples of loose foamed bitumen
treated base material were collected close to the locations where the lime-treated material was sampled,
namely at chainages 20, 40, 60 and 80 m in the outer wheel path (Figure E 16).
• Immediately after mixing with foamed bitumen, the treated basecourse was compacted using two 18.6 t
padfoot rollers (Figure E 16).
• The FBS base was then shaped with a grader and compacted (Figure E 16) using:
– two passes of 18.6 t padfoot rollers without vibration
– two passes of a 24.6 t smooth drum roller using full vibration
– a required number of passes of a 15.5 t multi-tyred roller.

Figure E 16: Loose basecourse after mixing and being compacted
Sampling behind the stabiliser immediately after mixing Vibration compaction using 18.6 t padfoot rollers
foamed bitumen
Figure E 17: Pavement after compaction

Lime Content
The quicklime was sampled from the spreader and tested for active lime content using the ASTM C25-11e1
test method. The available lime index (as CaO) was 87.5%.
Samples of the lime-treated granular base were obtained from the roadbed at four locations in the outer
wheel path and tested for lime content (Table E 9) in accordance with ASTM C25-11e1. The mean measured
quicklime content was 1.7%, close to the target 1.5%.
Note that samples were not taken in the inner wheel path, where an excessive amount of lime was spread
(Appendix E.5.1).
Table E 9: Lime content of roadbed samples
Location Project chainage Available lime content Quicklime content

(km) (m) (%) (%)
46.075 22 1.42 1.62
46.095 42 1.79 2.04
46.115 62 0.95 1.09
46.135 82 1.82 2.08
Mean 1.7
Residual Bitumen Contents
Figure E 16 shows samples of loose FBS base being taken from the roadbed at chainages 46.127 km and
46.145 km to determine the residual bitumen added during stabilisation. The samples were taken before and
after the foamed bitumen was mixed through the granular base and tested by Roads and Maritime for
bitumen content using the reflux method.
At chainage 46.127 km, the measured bitumen contents were 0.2% and 3.3% before and after treatment,
respectively, indicating 3.1% bitumen was added during the stabilisation. At chainage 46.145 km, the
measured bitumen contents were 0.1% and 2.8%, respectively, indicating 2.7% bitumen was added during
the stabilisation.
In addition, samples were taken from another four chainages and tested for bitumen content by ARRB using
the ignition oven method (Austroads 2005). The measured bitumen contents are listed in Table E 10.
Table E 10: Measured bitumen contents of samples from the roadbed
Bitumen content (%)

Location Project chainage
(km) (m) Roadbed material FBS material from Residual bitumen
prior to FBS roadbed added
46.075 22 0.9 3.85 2.95
46.095 42 3.13 2.23
46.115 62 3.36 2.46
46.135 82 3.61 2.59
Mean 2.6
Considering the Roads and Maritime test results of 3.1% and 2.7%, the estimated bitumen content added
was 2.7%, well below the target 3.5% bitumen.

The results were at variance with the amount of bitumen the contractor advised was used (Appendix E.5.1).
From the quantity of bitumen and considering the mean constructed thickness of FBS was 165 mm
(Appendix E.6.3) rather than the specified thickness of 175 mm, the bitumen contents in excess of 3.5%
were calculated.
Field Density and Moisture Testing
Following compaction, the in situ wet densities were measured using a nuclear gauge in direct transmission
(175 mm depth) (Figure E 18) using the Roads and Maritime test method (Roads and Maritime 2012d). This
data is listed in Table E 11.
Figure E 18: Nuclear gauge direct transmission field density measurements

Table E 11: FBS field wet densities, moisture contents, dry densities and relative densities
Field Standard Standard

In situ wet In situ dry Density Moisture
Chainage Offset moisture Proctor Proctor
density density ratio(1) ratio(1)
(km) (m) content MDD(1) OMC(1)
(t/m3) (t/m3) (%) (%)
(%) (t/m3) (%)
46.059 1.1 2.085 8.4 1.923 1.817 12.7 105.8 66
46.069 0.2 2.085 7.9 1.932 1.872 10.9 103.2 72
46.079 3.3 2.142 7.2 1.998 1.827 10.1 109.3 71
46.089 4.4 2.074 7.8 1.924 1.840 11.4 104.6 68
46.099 5.3 2.122 8.3 1.959 1.883 11.4 104.1 73
46.109 2.2 2.130 7.0 1.991 1.906 8.3 104.5 84
46.119 1.1 2.121 7.3 1.977 1.847 10.2 107.0 72
46.129 0.2 2.085 8.1 1.929 1.840 11.3 104.8 72
46.139 3.3 2.095 8.0 1.940 1.798 10.6 107.9 75
46.149 4.4 2.039 9.6 1.860 1.806 14.0 103.0 69
Mean 2.10 8.0 1.94 1.84 11.1 105.4 72
1 Standard Proctor compactive effort, 100 mm diameter moulds, field-mixed material.
As the gauge moisture contents are influenced by the added bitumen and the pulverised sprayed seal,
Roads and Maritime practice is to take samples from the pavement just prior to commencement of field
compaction and measure oven-dried moisture contents (Roads and Maritime 2012c). Consequently, these
samples are commonly taken several hours prior to compaction being completed and the subsequent in situ
density measurements. These oven-dried moisture contents and the resulting dry densities are listed in
Table E 11. The mean field moisture content at commencement of compaction was 8.0% and the mean in
situ dry density was 1.94 t/m3.
Table E 12 lists the mean dry densities obtained by coring (Appendix E.6.3). Pooling the mensuration
method density results, the upper portion had a mean density of 2.00 t/m3 and the lower portion had a mean
density of 1.92 t/m3, the average of which is 1.96 t/m3. As discussed in Section 5.2, the mensuration method
may underestimate the density of field cores if they have significant surface voids. Hence, the densities were
also measured using the presaturation method. It is noted that the mean dry density using the presaturation
method is about 2.01 t/m3. The nuclear gauge dry densities lie between the core results measured by the two
methods.
Table E 12: Mean dry density of field cores
Mean dry density (t/m3)

Cores taken at four Cores taken at eight
chainages two months after chainages 18 months after
Test method for density
construction construction
Upper Lower Upper Lower
portion portion portion portion
AS/NZS 2891.9.3:2014 Mensuration Method 2.02 1.99(1) 1.99 1.90
AS/NZS 2891.9.2:2014 Presaturation Method – – 2.04 1.98
1 Testing limited to two chainages.

At each of the density testing sites, loose FBS material was sampled prior to compaction (Figure E 16) and
transported to the Moree laboratory to measure the standard Proctor MDD and OMC (Roads and
Maritime 2012b). It was noted that due to the time required to transport the samples, there was a 3.5 to
6 hour time delay between field mixing and completion of laboratory compaction. The effect of such a time
delay on laboratory MDD and OMC has yet to be determined. However, as discussed in Section 6, a 24-hour
delay between mixing and compaction results in about a 2% reduction in dry density. It was also noted that
mean MDD of 1.84 t/m3 was 2–3% lower than the MDD determined in the mix design (Appendix E.3.2). The
mean OMC of the field-mixed material (11.1%) was also higher than the mix design value of 10.2%.
Using the MDD of the field-mixed material given in Table E 11, the density ratios were calculated at each test
site. The mean density ratio was 105% of standard MDD. However, these density ratios may have been
overestimated due to the effect of time delay on the MDD values.
There was considerable variation in the MDD values, which reinforced the Roads and Maritime practice of
measuring the MDD value for each field density site in order to calculate density ratios.
The mean field moisture content of 8.0% was about 72% of standard Proctor OMC of the field-mixed material
and about 80% of the OMC obtained in the mix design.
Differences between Nuclear Gauge and Oven-dried Moisture Contents
As the gauge moisture contents are influenced by the added bitumen, it was of interest to investigate the
differences in nuclear gauge and oven-dried moisture contents. Nuclear gauge moisture contents were
measured at four locations in the shoulder, and samples were excavated at the test sites to enable
measurement of associated oven-dried moisture contents. Figure E 19 data confirms that the gauge
moisture contents need significant adjustment in order to be consistent with oven-dried values.
Figure E 19: Comparison of nuclear gauge and oven-dried moisture contents

9.0
8.5
MCoven = 0.472MCgauge + 0.96
R² = 0.73
8.0
Oven-dried
moisture
7.5
content
(%)
7.0
6.5
6.0
12.0 12.5 13.0 13.5 14.0 14.5 15.0 15.5 16.0
Nuclear gauge moisture content (%)

Figure E 16 shows loose FBS being sampled from the roadbed at the same locations as samples taken for
determination of MDD and OMC (Appendix E.5.4). These samples were quickly transported to the Roads
and Maritime laboratory at Moree, and cylinders for indirect tensile modulus testing were prepared by
compacting the material into 100 mm diameter moulds using two compaction methods:
• by applying 50 blows to each end of the specimen with a 4.54 kg Marshall hammer, as applied to
compact asphalt specimens (Roads and Maritime 2012e) at field moisture content
• using standard Proctor compaction effort, as applied to untreated soils and granular materials (Roads and
Maritime 2012b) at field moisture content.
The following process was used to prepare and test the 100 mm diameter Marshall test cylinders:
• On reaching the laboratory from the field sampling, the loose FBS material was placed in an oven and
held at a temperature of 40 °C. Table E 13 gives the time that the material was sampled in the field and
the time at which the material was removed from the oven for compaction.
• The loose FBS material was compacted in the mould by applying 50 blows of the Marshall hammer to
each end of the specimen.
• The compacted test cylinders were then covered and cured at 20–23 °C for three hours.
• The cylinders were then uncovered and returned to the 40 °C oven for three days.
• The cylinders were then cooled, weighed, and their bulk densities calculated using the mensuration
method.
• The cylinders were then wrapped and transported to the Roads and Maritime laboratory in Sydney for
indirect tensile modulus testing.
• About a month after construction, the dry and wet moduli of each cylinder were measured (Figure E 13).

Table E 13: Modulus of Marshall hammer (4.54 kg) compacted specimens
Cured dry Cured wet

Time Time Density indirect indirect Ratio
Chainage Offset Sample sampled Moisture Dry density
lab compact. ratio(1) tensile tensile wet/dry
(km) (m) number from content (%) (t/m3)
(%) modulus modulus modulus
roadbed
(MPa) (MPa)
46.059 1.1 1A 1200 1332 8.4 2.026 111.5 3330 1820 0.55
1B 1200 1336 1.994 109.7 2500 1410 0.56
46.069 0.2 2A 1100 1230 7.9 2.037 108.8 3100 1440 0.46
2B 1100 1238 2.025 108.2 2880 1200 0.42
46.079 3.3 3A 1319 1450 7.2 1.972 107.9 1920 870 0.45
3B 1319 1456 1.933 105.8 1850 730 0.39
46.087 4.4 4A 1328 1501 7.8 1.981 107.7 1950 840 0.43
4B 1328 1507 1.972 107.2 2270 890 0.39
46.097 5.3 5A 1336 1520 8.3 2.052 109.0 3750 1600 0.43
5B 1336 1525 2.075 110.2 3780 1830 0.48
46.107 2.2 6A 1230 1351 8.4 2.029 106.5 2440 1060 0.43
6B 1230 1358 2.019 105.9 2100 1280 0.61
46.117 1.1 7A 1240 1410 7.3 2.035 110.2 3110 1700 0.55
7B 1240 1416 2.029 109.9 3500 1790 0.51
46.127 0.2 8A 1135 1259 8.1 2.037 110.7 3240 1650 0.51
8B 1135 1305 2.028 110.2 2200 1180 0.54
46.137 3.3 9A 1351 1531 8.0 1.959 109.0 2280 1280 0.56
9B 1351 1537 1.945 108.2 2190 1130 0.52
46.147 4.4 10A 1412 1541 9.6 1.975 109.4 1810 940 0.52
10B 1412 1552 1.987 110.0 1770 820 0.46
Mean 8.1 2.01 109.2 2600 1270 0.49
1 Using the standard Proctor MDD (Figure E 11).

A similar process was adopted for the cylinders compacted in the standard Proctor compaction mould with
the standard compaction hammer, except that 25 blows of a 2.7 kg standard hammer were applied to each
of three layers. As the standard Proctor specimens were compacted after the Marshall specimens, the time
delay between field mixing and laboratory compaction was 3.5 to 6 hours, well outside the 3-hour limit used
by TMR. To enable modulus testing, the 116 mm high specimens were then cut to a height of 75 mm. The
moduli are listed in Table E 14.
It should be noted that the average density ratio of the Marshall specimens (109.2%) was 3.9% higher than
the average in situ density ratio (Table E 11). In contrast, the average density ratio of test cylinders
compacted using the standard compaction hammer and mould (Table E 14) was 5.5% lower than the
average in situ density ratio. It is considered that this difference in density of the Marshall and standard
compaction test specimens was a source of the large differences in moduli between the two compaction
processes (Figure E 20) and the additional time delay in compacting the standard Proctor specimens.
Table E 14: Modulus of specimens compacted using standard Proctor drop hammer
Cured
Cured dry
wet
Standard Dry Density indirect Ratio
Chainage Offset Sample indirect
MDD density ratio(1) tensile wet/dry
(km) (m) number tensile
(t/m3) (t/m3) (%) modulus modulus
modulus
(MPa)
(MPa)
46.057 1.1 1 1.817 1.871 103.0 1450 780 0.54
46.067 0.2 2A 1.872 1.829 97.7 1250 590 0.47
46.087 4.4 4 1.840 1.822 99.0 1100 380 0.35
46.117 1.1 7A 1.847 1.829 99.0 1470 680 0.46
7B 1.847 1.810 98.0 1190 630 0.53
46.127 0.2 8A 1.840 1.867 101.5 1470 750 0.51
8B 1.840 1.863 101.3 1460 670 0.46
46.147 4.4 10 1.806 1.800 99.7 690 294 0.43
Mean 1.84 99.9 1260 600 0.47
1 Using the standard Proctor MDD (Table E 11).

Figure E 20: Cured dry modulus variation with density ratio and compaction method
4000
3500
Marshall cylinders
3000
Standard cylinders
2500
Dry cured
indirect
tensile 2000
modulus
(MPa)
1500
1000
500
0
97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113
Density ratio (%)
Note that for roads with a daily traffic loading of 1000 ESA or more, the TMR mix design criteria (TMR 2012)
based on laboratory-manufactured 150 mm diameter Marshall specimens (9.92 kg hammer) are:
• a minimum cured dry modulus of 4000 MPa
• a minimum cured wet modulus of 2000 MPa
• a minimum wet to dry modulus ratio of 0.50.
None of the field-mixed laboratory-compacted specimens given in Table E 13 complies with these criteria,
which is consistent with the mix design findings (Appendix E.3.3) but varies from the initial Roads and
Maritime mix design (Appendix E.3.1).
Opening to Traffic and Placement of Sprayed Seal
The FBS layer was left to cure for two days prior to opening to traffic on 7 October 2013.
A size 14 sprayed seal was placed on 10 October 2013, five days after the FBS construction and after
3 days trafficking. The bituminous binder was Class 170 bitumen with 5% cutter oil applied with a bitumen
application rate of 1.2 l/m2. A size 7 mm scatter coat was applied on top of the size 14 mm aggregate without
application of additional bitumen. This seal type is commonly called a single/single seal with scatter coat
(Figure E 21). The FBS surface was not primed before the application of this initial seal.
Figure E 21: Single/single seal with scatter coat

On 11 March 2015, a double/double (14 mm/7 mm) final sprayed seal was placed with a total bitumen
application rate of 1.7 l/m2. An S45R (rubber) polymer modified binder was used.
E.6 Performance Monitoring
Surface Deflections
Falling weight deflectometer (FWD) deflection testing (Figure E 22) of the trial section was undertaken two
months after construction (December 2013) and at about 18 months (April 2015). Measurements were taken:
• along the edge line
• along the outer wheel path
Figure E 22: FWD testing of Newell Highway about two months after construction
of 566 kPa. During the deflection measurements, the pavement temperatures were 32–34 °C in
December 2013 and 29–32 °C in April 2015. The maximum deflections and curvatures were corrected from
the pavement temperature at the time of testing to the weighted mean annual pavement temperature
(WMAPT) of 31 °C using the interim deflection-temperature method previously developed (Appendix D.15).
The maximum deflections and curvatures are shown in Figure E 23 and Figure E 24 and indicate a very
high-strength pavement with no apparent structural deterioration after 18 months trafficking. On the contrary,
the curvature results have decreased, consistent with an increase in FBS modulus.

Figure E 23: Newell Highway maximum deflections

0.50
0.45
0.40 Outer wheel path 2 months
Outer wheel path 18 months

0.35
0.30
Maximum
deflection at
0.25
566 kPa
(mm)
0.20
0.15
0.10
0.05
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
0.50
0.45
0.40 Between wheel path 2 months
Between wheel path 18 months

0.35
0.30
Maximum
deflection at
0.25
566 kPa
(mm)
0.20
0.15
0.10
0.05
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
0.50
0.45
0.40 Edge line wheel path 2 months
Edge line wheel path 18 months

0.35
0.30
Maximum
deflection at
0.25
566 kPa
(mm)
0.20
0.15
0.10
0.05
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

Figure E 24: Newell Highway FBS curvatures

0.10
0.09
0.08 Outer wheel path 18 months
0.07
0.06
Curvature
(D0-D200) at
0.05
566 kPa
(mm)
0.04
0.03
0.02
0.01
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
0.10
0.09
Between wheel path 2 months
0.08 Between wheel path 18 months
0.07
0.06
Curvature
(D0-D200) at
0.05
566 kPa
(mm)
0.04
0.03
0.02
0.01
0.00
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
0.10
0.09
Edge line 2 months
0.08
Edge line 18 months
0.07
0.06
Curvature
0.05
(D0-D200) at
566 kPa 0.04
(mm)
0.03
0.02
0.01
0.00
0 10 20 30 40 50 60 70 80 90 100
-0.01
Chainage (m)

Moduli Back-calculated from Measured Surface Deflections
Using the measured surface deflection bowls at two months and 18 months after construction, and the mean
constructed FBS thickness of 165 mm from the coring (Appendix E.6.3), the pavement and subgrade layer
moduli were back-calculated. The results are listed in Appendix G.2.
Of particular interest is the FBS modulus variation with time and also the differences between values in the
unloaded pavement areas (edge line and between wheel paths) and loaded pavement area (outer wheel
path). To enable this comparison, the back-calculated moduli from the measurement temperature during
FWD testing to the WMAPT for Moree ( 31 °C) using the Austroads (2011a) modulus temperature
dependency.
The temperature-corrected FBS moduli are plotted in Figure E 25. Figure E 26 is a plot of the
back-calculated moduli of the layer under the FBS material which the pavement investigation
(Appendix E.2.3) suggested was unbound granular subbase.
Figure E 25: Newell Highway back-calculated FBS moduli

2 months after construction
13000
12000
11000
Outer edge line
Between wheel paths
9000
8000
FBS
modulus
7000
at 31°C
(MPa) 6000
5000
4000
3000
2000
1000
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

12000
11000
10000
9000
8000
FBS
7000
modulus
at 31°C 6000
(MPa)
5000
4000
Outer edge line
3000
Outer wheel path
1000
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

Figure E 26: Newell Highway back-calculated moduli of thick layer under the FBS
9000
8000 Outer edge line

Outer wheel path
7000
Between wheel paths
6000
Modulus 5000
layer under FBS
(MPa) 4000
3000
2000
1000
0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

7000
Outer edge line

Between wheel paths
5000
4000
Modulus
layer under FBS
(MPa)
3000
2000
1000
0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)
To illustrate the unusually high moduli back-calculated for the layers under the foamed bitumen layer,
Table E 15 lists the mean layer moduli calculated from deflections taken at 18 months in the outer wheel
path, between wheel paths and along the edgeline.
It was concluded that the very low deflections resulted in part due to the very high moduli of the layers under
the foamed bitumen stabilised layer. In the visual inspection in December 2014, cracking was observed in
pavement areas near the trial site. These two findings bring into question whether the pavement had been
previously cement stabilised, despite this not being observed in the pavement investigation (Appendix E.2.3).
However, Roads and Maritime have no record of cement stabilisation of this section of the Newell Highway.

Unlike the findings for the Calder Highway (Appendix D.10.1), there is no indication as yet that the FBS in the
outer wheel path is reducing in modulus due to loading, in fact, it roughly doubled in value between 2 months
and 18 months. This is not surprising given the very low strains being applied to the FBS layer due to it being
well-supported by underlying layers. Based on this finding, it is unlikely that FBS fatigue will be observed.
Table E 15: Newell Highway mean back-calculated moduli of outer wheel path 18 months after opening to
traffic
Layer thickness Mean back-calculated modulus

under in back- (MPa)
Assumed material type
calculation Outer wheel Between Edge line
(mm) path wheel paths
Foamed bitumen stabilised granular base 165 8200 7260 7490
Granular subbase 180 1650 1710 2650
Granular subbase 180 810 480 1160
Gravelly clay fill 435 840 470 840
Gravelly clay fill 435 380 480 570
Top of subgrade 400 250 230 280
Lower subgrade Semi-infinite 250 220 260
Testing of Field Cores
Two months after construction, 100 mm diameter cores of FBS material were taken along the edge line at
four chainages along the trial section (Figure E 27). Another eight cores were taken along the edge line at
18 months. Surface deflections were measured at all coring locations prior to coring. All 12 cores were
extracted intact without breakage during the coring (Figure E 28). The cores were transported to ARRB’s
laboratory in Victoria for testing.
The heights of the cores were between 160 mm and 173 mm, with a mean thickness of 165 mm (close to the
target layer thickness of 175 mm).
Figure E 27: Coring of Newell Highway two months after construction

Figure E 28: Four FBS cores taken along the edge line at two months

In the case of the two-month-old cores, each core was sawn into upper and lower portions using dry ice to
cool the specimens rather than water, to enable the field moisture contents to be determined. The dry density
(by the mensuration method) and indirect tensile modulus were measured. The results are listed in
Table E 16 and plotted in Figure E 29. The moduli after curing in situ for 2 months were generally in excess
of 5000 MPa.
Table E 16: Modulus and density of 2 month old field cores
Bulk
Indirect Back-
Project density at Field
Upper Dry tensile calculated
field moisture
Date of coring chain- Offset or lower density(1) modulus at modulus at
moisture content
age (m) portion (t/m3) 25 °C 25 °C
content (%)
(MPa) (MPa)
(t/m3)
10 December 20 Outer Upper 2.112 2.0 2.038 10 650 14 500
2013 edge
Lower 2.060 1.6 1.969 6 140
2 months after line
construction
40 Outer Upper 2.085 2.0 2.000 7 220 10 380
edge
Lower 2.084 2.0 2.011 6 990
line
60 Outer Upper 2.085 2.3 2.002 9 090 2 850
edge
Lower 2.046 N/A N/A 5 720
line
80 Outer Upper 2.096 2.2 2.027 7 800 3 640
edge
Lower 2.018 N/A N/A 2 840
line
Upper 2.09 2.2 2.02 8 690
Mean 7 840
Lower 2.06 1.8 1.99 6 280
When the 18-month-old cores were sawn into upper and lower portions, it was no longer possible to use dry
ice with a new saw, and hence the specimens’ moisture contents were affected by the wet-sawing process.
For these cores, the dry densities were measured using both the mensuration and presaturation methods.
The measured dry densities and indirect tensile moduli are given in Table E 17 and plotted in Figure E 30.
Comparing the data in Figure E 29 and Figure E 30, for a given dry density, the 18-month-old cores have
about twice the modulus of the 2-month-old cores, reasonably consistent with the back-calculated modulus
findings.

Table E 17: Modulus and density of field cores extracted at 18 months
Indirect tensile modulus at 25 °C Back-

Project Dry (MPa) calculated
Upper or Dry density(1)
Date of coring chainage Offset density(2) modulus at
lower portion (t/m3)
(m) (t/m3) 25 °C
As received Cured Cured (MPa)
from roadbed dry modulus(3) wet modulus(4)
14 April 2015 15 Outer edge Upper 2.054 2.095 17 020 16 450 9 660 12 260
18 months after line
Lower 1.990 2.046 11 750 11 410 4 770
construction
25 Outer edge Upper 1.962 2.022 11 050 11 880 5 990 13 260
line
Lower 1.903 1.968 6 640 6 050 2 720
35 Outer edge Upper 1.984 2.036 12 390 10 930 4 970 12 530
line
Lower 1.904 1.985 7 750 7 590 3 500
45 Outer edge Upper 1.991 2.05 12 410 11 820 6 970 10 530
line
Lower 1.924 1.991 9 590 9 340 3 930
55 Outer edge Upper 1.979 2.021 11 720 10 760 5 640 5 110
line
Lower 1.923 1.996 8 440 10 160 3 870
65 Outer edge Upper 1.999 2.041 13 750 13 140 8 120 6 260
line
Lower 1.839 1.921 8 340 6 480 3 700
75 Outer edge Upper 1.967 2.009 12 200 13 150 5 140 4 490
line
Lower 1.908 1.971 8 290 8 730 3 680
85 Outer edge Upper 1.977 2.018 13 150 13 450 7 060 8 950
line
Lower 1.828 1.943 7 630 7 790 3 270
Upper 1.99 2.04 12 960 12 700 6 690
Mean 9 170
Lower 1.90 1.98 8 560 8 440 3 680
2 AS/NZS 2891.9.2:2014 Presaturation Method
4 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under partial vacuum then tested for modulus at 25 °C.

Figure E 29: Variation in field core modulus at 2 months with dry density
Modulus of 2 month old field cores as-received

11000
Top half core
10000 Bottom half core
9000 E = 71060DD - 134888
R² = 0.77
8000
Indirect 7000
tensile
modulus
(MPa) 6000
5000
4000
3000
2000
1.92 1.94 1.96 1.98 2.00 2.02 2.04 2.06
Dry density (t/m3) using mensuration method
Figure E 30: Variation in field core modulus at 18 months with dry density
Modulus of 18 month old field cores as-received
18000
Top half core
16000 Bottom half core
14000
Indirect 12000
tensile
modulus
(MPa) 10000
8000
6000
4000
1.80 1.85 1.90 1.95 2.00 2.05 2.10
Dry density (t/m3) using mensuration method

It was of interest to compare the 2-month core moduli (Table E 16) with the mix design moduli (Table E 6).
As the mix design dry densities were measured using the presaturation method, the mean dry density of the
mix design specimens in Table E 6 needed to be reduced by 2.5% (Section 4.3.3) to an equivalent dry
density using the mensuration method. Plotted in Figure E 29 is the mix design mean cured dry moduli
compared to the core moduli. It was concluded that when allowance is made for the difference in density
in situ compared to that of laboratory specimens, the cured dry modulus was similar to the modulus of field
cores taken 2 months after construction.
Differences between moduli of laboratory specimens and field core values are discussed in Section 5.2.
Rutting and Roughness
On 28 November 2013, 22 April 2015 and 10 April 2017 about 6 weeks, 18 months and 42 months after
placement of the surfacing, rutting and roughness measurements were taken with a laser profiler.
The rutting results, shown in Figure E 31, indicate very minor deformation under initial trafficking. The 2015
and 2017 surveys indicated that the average rut depth had increased 2–3 mm compared to the 2013 survey
results. Placement of the final sprayed seal in March 2015 may have contributed to this increase in rutting.
Figure E 31: Newell Highway rutting results
10
9
Outer wheel path 6 weeks
8
7
6
Rut
depth 5
(mm)
4
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Chainage (m)

10
9
Inner wheel path 6 weeks
8 Inner wheel path 18 months

Inner wheel path 42 months
7
6
Rut
depth 5
(mm)
4
0
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100
Chainage (m)
The initial ride quality of the trial section was excellent, with a lane International Roughness Index (IRI) of
0.71 m/km in 2013 and 0.76 m/km in 2015 (Figure E 32). In the 2017 survey, the results were similar except
for an increase in roughness of the inner wheel path around chainage 30 m due to cracking.
Figure E 32: Newell Highway roughness results
2.5
2.0 6 weeks
18 months
42 months
1.5
Lane IRI
(m/km)
1.0
0.5
0.0
0 10 20 30 40 50 60 70 80 90 100
Chainage (m)

Visual Survey
The trial section was inspected by Roads and Maritime in December 2014, about 14 months after opening to
traffic and prior to the placement of the final seal. There was no cracking apart from minor crack sealing of a
break in the seal at the end of the test section (Figure E 33), which is unlikely to be due to fatigue of the FBS.
Figure E 33: View from the end of the test section looking south, December 2014
However, cracking was observed outside the trial area, including longitudinal cracking in the adjacent
southbound lane (Figure E 34). As mentioned in Appendix E.2, very high plasticity (PI = 49) soil exists more
than 1400 mm below the surface. This material is a highly reactive black soil that causes pavement
longitudinal cracking when it dries, as observed in the verge. The very low deflections at this site are
consistent with the black soil moisture content being currently low. Roads and Maritime have no record of
previous cementitious stabilisation at this site and such materials were not evident in the initial pavement
investigation (Appendix E.2.3).
Figure E 34: Cracking of the southbound lane adjacent to the northbound lane trial section
Nearby longitudinal cracking, December 2014 Cracking in the verge
It seems likely that longitudinal cracking due to reactive soil will occur in the trial section in the future. As
such, the usefulness of the site in terms of observing load-related cracking may be comprised by the
environmental cracking.
As mentioned in Appendix E.5.7, the final seal was placed in March 2015, and hence no cracking inside or
outside the trial section was apparent when the pavement was inspected in April 2015 during the FWD
testing and pavement coring.

Figure E 35: Condition in April 2015, one month after final sprayed seal was placed
In September 2016, after about 3 years trafficking, the northbound lane trial section was again visually
inspected. Figure E 36 is a record of the observed distress (Sept 2016 results marked in blue). There was
some minor additional transverse cracking when the pavement was inspected again in May 2017 (marked in
green).
Between chainages 0 to 75 m, the longitudinal cracking about 530 mm away from the edge line was barely
discernible. It seems unlikely this crack is associated with the construction joint between stabiliser passes
(Appendix E.5.1) as this joint is located about 1 m from the edge line.
Between chainages 75 m to 95 m, the combined longitudinal and transverse cracking was more pronounced
although still far from needing intervention (Figure E 37).
The most severe cracking was the small block cracking in the inner wheel path between 95 m and 100 m
(Figure E 38). As mentioned in Appendix E.5.2, the inner wheel path was mistakenly overdosed with lime
(2.4% quicklime compared to 1.5% target). This high lime content may have contributed to the development
of this block cracking.
When the severity of cracking develops further between chainages 0 to 95 m, pits and cores will be taken to
investigate whether the reactive subgrade soil at this site has contributed to the longitudinal cracking.

Figure E 36: Visual inspection September 2016 and May 2017
Figure E 37: Cracking between chainages 80–90 m observed September 2016

Figure E 38: Block cracking over a 5 m length observed September 2016
E.7 Comparison of Predicted and Observed Performance

As mentioned in Appendix E.4, the predicted allowable traffic loadings to FBS fatigue cracking was
2.2 x 106 ESA.
Using the cumulative traffic loading shown in Figure E 2, the pavement is predicted to have reached this
loading after about 2.5 years of trafficking, that is early 2016. As discussed in Appendix E.6.5, fine cracking
was observed after 3 years trafficking but there are doubts about whether fatigue is the cause.
E.8 Future Monitoring

A further visual inspection and coring of the trial section is planned in late 2017 under Austroads project
TT2046 Improved Cost-effectiveness of Foamed Bitumen Stabilised Pavements. The coring will assess
whether the very low measured deflections are due to a bound layer under the FBS base.
E.9 Newell Highway Trial Summary

In October 2013, a trial section of 3.5% FBS mixes was constructed on the Newell Highway, Bellata, New
South Wales. Based on preliminary testing to date, the findings are:
• The mix design moduli were below TMR minimum requirements. However, the mix did comply when
tested by Roads and Maritime using an alternative mix design method which included heating the
material to 35–40 °C during mixing and compaction.
• The dry densities of the Marshall mix design specimens were similar to the mean field densities, despite
the lower moisture content used in the laboratory.

• The trial section was constructed to a high standard of compaction, about a mean 105% standard Proctor
MDD. However, this compaction assessment was based on laboratory compaction of field-mixed FBS
more than 3 hours after mixing in the roadbed. The time delay between mixing and compaction may be
lead to the laboratory MDD being underestimated and hence the relative compaction overestimated.
• There was a reduction of density with depth, less than for the Calder Freeway trial but exceeding that of
the Western Freeway trial.
• The mean field moisture content (8.0%) was well above the value used in the Marshall mix design testing,
7.2% corresponding to 70% standard Proctor OMC.
• The average depth of stabilisation was 165 mm, close to the 175 mm design thickness.
• The measured bitumen contents of materials sampled during construction indicated the amount of
bitumen added was about 2.7%, significantly below the 3.5% design value.
• Early-life FBS cores taken along the edge line two months after construction had mean indirect tensile
moduli in excess of 6000 MPa at a temperature of 25 °C. For a given density, these early-life moduli were
similar to the mix design cured dry modulus.
• For a given density, these early-life field core moduli roughly doubled after 18 months in situ curing.
• The moduli of the pavement layers and subgrade under the FBS layer were high and well in excess of
expectation for the pavement investigation prior to construction. This may be related to the extended
period of low rainfall. In its current condition, fatigue cracking of the FBS is not anticipated due to the very
low strains being induced in the FBS layer.
• Very fine longitudinal and transverse wheel path cracking were observed after about 3 years trafficking
over a significant portion of the trial, with block/crocodile cracking over a short length. As the trial
pavement is on a reactive black soil subgrade, there are doubts as to whether this cracking is solely
load-induced.

Appendix F Western Freeway, Ballan
F.1 Introduction
In March 2015, VicRoads and the stabilisation industry agreed to support the Austroads project by
constructing FBS pavements on the Western Freeway, Ballan, Victoria. The Western Freeway at Ballan is
part of the heavily-trafficked national road carrying freight between Victoria and South Australia. Figure F 1
shows the location of the trial sections.
As described in Table F 1, two FBS pavements were constructed of the same thickness with varying bitumen
content of FBS layer: a nominal 2.5% bitumen mix and a nominal 3% bitumen mix. Both mixes included
about 20% of recycled asphalt patching material. The nominal 200 mm thickness of the FBS layer was
selected so that there was a reasonable likelihood of fatigue cracking within the period of the research
project.
The annual average rainfall in the nearby township of Ballan is 570 mm.
Table F 1: Description of the Western Freeway FBS pavement
Description
Location Western Freeway, Ballan, westbound carriageway 77.93 to 78.23 km
Job size Length 266 m comprising 138 m with 2.5% bitumen mix and 128 m of
3% bitumen mix
Number of lanes 1 (outer lane, westbound)
Lane width 4.0 m (3.5 m slow lane and 0.5 m shoulder)
Total area 1060 m2
Stabilisation depth Nominal 200 mm FBS base: actual 185 mm for 2.5% mix,
205 mm for 3% mix
Foamed stabilisation Mix design Undertaken by Downer Infrastructure
specification
Host materials VicRoads Class 1 crushed rock (size 20 mm), plasticity index 2 to
6 with about 20% asphalt millings
Supplementary 1.5% hydrated lime
binder
Bitumen Specified: 2.5% Class 170 bitumen, actual 2.3%
Specified: 3.0% Class 170 bitumen, actual 2.7%
Construction method Work specifications VicRoads specifications
and QA testing
Construction dates Milling of asphalt patches and mixing untreated materials,
2 March 2015
Foamed bitumen stabilisation, 3 March 2015
Double/double seal, 4 March 2015
Surfacing Size 14/7 primerseal, includes 10 parts cutter oil and 5 parts
rubber
Figure F 2 is a view of the trial site during construction. The total length of the slow lane stabilised was 266 m
and comprised a nominal 2.5% bitumen FBS mix between chainages 0 and 138 m, with a nominal
3.0% bitumen mix between chainages 138 and 266 m.

Figure F 1: Locality plan, Ballan, Victoria
Figure F 2: Western Freeway FBS site during construction

F.2 Site Investigation Prior to Stabilisation

To assist in the mix design and structural design, a pavement investigation was undertaken in
December 2014 and the pavement surface deflections measured.
The Western Freeway, Ballan pavement was originally constructed in the 1970s. In the mid-1990s the
pavement was strengthened with a 200 mm thick re-sheet of crushed rock base. Based on the two pits dug
at project chainages 80 and 190 m, the pavement structure prior to stabilisation comprised:
• multiple sprayed seal surfacings
• 200 mm Class 1 (size 20 mm) crushed rock base
• 100 mm crushed rock upper subbase
• 200 mm sandstone lower subbase
• 300 mm brown gravelly clay select fill
• subgrade clay.
From the results of dynamic cone penetrometer (DCP) testing from the top of the brown clay select fill, the in
situ California Bearing Ratios (CBR) were estimated (Figure F 3). The estimated CBRs were 5% or more.
The relatively low deflectograph maximum deflections (D0) also confirmed that subbase/subgrade strength
was adequate for foamed bitumen stabilisation (Figure F 4).
Figure F 3: Estimated in situ CBR of the clay select fill and underlying subgrade
Western Freeway Ballan

50
45
40
35
30
In situ CBR
25
(%) Chainage 80 m
20 Chainage 190 m
15
10
0
560 610 660 710 760 810 860 910 960 1010 1060 1110 1160
Depth below the surface (mm)

Figure F 4: Deflectograph maximum deflections prior to treatment
1.0
OWP
0.9
IWP
0.8
0.7
0.6
D0
0.5
(mm)
0.4
0.3
0.2
0.1
0.0
0 20 40 60 80 100 120 140 160 180 200 220 240 260
As shown in Figure F 5, the pavement included 150 mm deep asphalt patches over about 30% of the area
between chainages 0 and 138 m (2.5% bitumen) and about 25% of the area between chainages
138 and 266 m (3.0% bitumen). The patches were generally in the outer wheel path. Prior to construction,
cores were excavated to determine the thickness of the asphalt patches. This testing confirmed that the
asphalt patches were about 150 mm thick. As discussed in Appendix F.3, prior to stabilisation these patches
were removed then backfilled with imported crushed rock. The asphalt millings were spread along the trial
sections and mixed with the crushed rock base.

Figure F 5: Locations of 150 mm deep asphalt patches in the two trial areas
The measured particle size distributions of the crushed rock base sampled from the two test pits are given in
Figure F 6. As asphalt patch millings were subsequently mixed with this crushed rock base, the in situ
material stabilised on the project had a different distribution of sizes (Figure F 29). As the unblended crushed
rock base tended to be on the fine side of the Austroads limits, the size distribution was improved by the
addition of asphalt millings (Figure F 8).

Figure F 6: Particle size distribution of the untreated crushed rock base prior to addition of asphalt patch
millings
100
Lower limit proposed Austroads
90 Upper limit proposed Austroads
80 m
80 190 m
70
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)
F.3 Inclusion of Asphalt Millings in Material to be Stabilised

As described in Appendix F.2, there is a significant amount of asphalt patching (150 mm deep) in the trial
section: about 260 m2 in area, which is about 28% of the trial area. VicRoads advised that such patching is
common on heavily trafficked rural roads which could be considered for stabilisation in the future. As such,
VicRoads was interested in incorporation of the asphalt patch millings in the works.
In considering the inclusion of the millings in the works, it was noted that:
• there has been limited Australian experience on the use of high (> 20%) recycled asphalt
pavement (RAP) content in FBS mixes for heavily trafficked roads
• early-life rutting occurred when asphalt patches were stabilised in a previous Austroads trial (Port
Wakefield Road, South Australia – Appendix C.6).
After consultation with VicRoads and industry, it was agreed that:

• the existing asphalt patches would be milled from the pavement using a profiler and cast to the shoulder
• the milling would then be homogeneously mixed with the crushed rock base to limit the RAP content to a
maximum of 20%.
Appendix F.6.2 describes the construction process used.

F.4 Mix Design
Testing Prior to Construction
Mix design testing was undertaken by Downer Infrastructure at the request of ARRB. As described in
Appendix F.2, crushed rock base was sampled from the pavement prior to the works for use in the mix
design.
A significant issue in undertaking the mix design was that the material proposed to be stabilised would
include up to 20% asphalt millings. For the mix design, it would have been desirable to use millings from the
trial area. This was considered impractical prior to construction; therefore, it was decided to use dense
graded asphalt millings obtained from a major road in the Melbourne metropolitan area. Figure F 7 shows the
particle size distribution of these millings, the crushed rock base, and a blend of 80% crushed rock base and
20% millings. Note the inclusion of 20% millings resulted in a material blend well within the required
Austroads limits. The distributions were similar to those from sampling the roadbed during construction
(Figure F 8).
Figure F 7: Particle size distributions of materials used in mix design undertaken prior to construction
100
90 Upper limit proposed Austroads
asphalt millings
80 100% crushed rock base
80% crushed rock base: 20% millings
70
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)
Over the last two years, there has been increasing use of lower bitumen content FBS mixes with bitumen
contents around 3% rather than 3.5% commonly used in the past. Given this trend and with the inclusion of
20% asphalt millings, in consultation with Austroads and industry, it was decided to trial two mixes at Ballan:
2.5% and 3.0% bitumen.
Using 100 mm diameter moulds, the modified Proctor MDD and OMC of the untreated and 3% bitumen
treated materials were measured and are reported in Table F 2. Note these MDD and OMC values are
markedly different from those measured on materials sampled from the roadbed (Appendix F.4.2) and
compacted into a 150 mm diameter mould.

Table F 2: Modified Proctor maximum dry densities and optimum moisture contents prior to construction
80% crushed rock base: 20% asphalt

100% crushed rock base
millings
Binder Maximum dry Optimum Maximum dry Optimum
density moisture density moisture content
(t/m3) content (%) (t/m3) (%)
Untreated 2.12 6.8 2.12 6.6
1.5% hydrated lime and 3% bitumen 2.02 6.4 2.00 6.1
The first step in the mix design testing was to add moisture to the untreated material to increase it to a target
80% of modified Proctor OMC of the untreated material, in accordance with current practice. Then the
hydrated lime was added and mixed through. The lime-treated material was then left to cure for one hour
before application of the foamed bitumen. A foaming agent (0.6% Teric 311) was added to the Class 170
bitumen to assist with foaming characteristics. The foamed bitumen was then mixed with the lime-treated
materials. The resulting contents of bitumen, hydrated lime and moisture are shown in Table F 3. The mixing
moisture contents were 5.0–5.5%, significantly wetter than the mean field moisture content measured during
construction (4.6%, Table F 9).
Modulus test specimens (152 mm diameter) were then prepared by compacting the loose treated material
into the moulds by applying 50 blows to each specimen end with a 10.2 kg Marshall hammer. The specimens
were then tested for indirect tensile modulus (initial, cured dry and cured wet) at a temperature of 25 °C and
40 milliseconds rise time. The results are given in Table F 3.
At a bitumen content of 3.1%, the 100% crushed rock base mixture meets the minimum TMR requirement for
mix acceptance. The 80% crushed rock base and 20% millings mixture also meets the minimum
requirements, except that the cured dry modulus is about 5% below the TMR minimum (4000 MPa). These
modulus specimens were mixed and compacted at lower moisture content (5.0%) than the other two mixes.
Had this mix been compacted at slightly higher moisture content, the density would have increased, and
more than likely the cured dry modulus would have met the TMR requirements.
Table F 3: Mix design results prior to construction
Mean indirect tensile modulus (MPa)

Hydrated at 25 °C
Bitumen Moisture Mean dry Ratio
Material to lime
be stabilised content Initial Cured dry Cured wet modulus
(%) (%) (t/m3)
100% 3.1 1.5 5.3 2.12 990 5480 3460 0.63
crushed rock
80% crushed 3.1 1.5 5.0 2.09 1010 3800 2800 0.74
rock/20%
3.6 1.5 5.5 2.15 1000 6930 4750 0.69
asphalt
millings
TMR minimum requirements for base with > 1000 ESA/day 700 4000 2000 0.50
Based on this testing, it was decided to use 3% bitumen and 1.5% hydrated lime in the trial. In addition, the
trial included a mix with 2.5% bitumen and 1.5% hydrated lime to investigate how the field performance
varied with bitumen content.

Testing after Construction
During the first day of construction, samples of the crushed rock/asphalt milling blend were obtained every
5 m along the 2.5% bitumen and 3.0% bitumen trial areas prior to stabilisation with lime and foamed bitumen
(Appendix F.7.2). For each mix type (Figure F 5), the samples at each chainage were combined to produce a
bulk sample for mix design testing.
As the asphalt millings in each bulk sample still had conglomerates greater than 20 mm in size yet to be fully
broken down by the stabiliser (Figure F 16), these conglomerates were broken in the laboratory using a
hammer to more accurately reflect the particular size distribution after two further passes of the stabiliser in
the field. Figure F 8 shows the mean particle size distributions of the materials to be stabilised, which were
similar to those later obtained from sampling the roadbed after mixing with lime and foamed bitumen
(Figure F 29). The material to be stabilised was non-plastic (i.e. plasticity index less than 2).
Figure F 8: Particle size distributions of materials used in mix design undertaken after construction
100
90
Upper limit proposed Austroads
80 2.5% mix mean sampled from road bed
3% mix mean sampled from the road bed

70
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)
Using the 100 mm diameter moulds, the modified Proctor MDD and OMC of the untreated and treated
materials were measured and are reported in Table F 4. Note that these MDD and OMC values are markedly
different from those measured in the preliminary mix design (Table F 2).
Table F 4: Modified Proctor maximum dry densities and optimum moisture contents of bulk samples from
the roadbed
Maximum dry Optimum moisture

Section Binder density content
(t/m3) (%)
Chainages 10–138 m Untreated 2.29 6.8
2.5% bitumen/ 1.5% lime 2.17 7.2
Chainages 139–265 m Untreated 2.30 7.0
3.0% bitumen/ 1.5% lime 2.14 6.9

As described in Section 3, the moisture content at which mix design specimens are mixed and compacted
has a major effect on both the test specimen dry density and modulus. As shown in Table F 9, the mean
moisture content in the field was 4.6%, which equates to about 62% and 66% of the modified Proctor OMCs
listed in Table F 4. It is common practice to mix and compact materials in the laboratory-mix designs at
higher moisture contents. For example, TMR use 70% of standard Proctor OMC of the untreated material
and Downer Infrastructure use 80–85% of modified Proctor OMC of the untreated material. Against this
background, it was decided to assess the modulus at two mixing moistures:
• the mean field moisture content (4.6%) at commencement of field compaction
• 80% of modified Proctor OMC of the untreated material (about 5.6%).
The first step in the mix design was to add moisture to the untreated material to increase it to the target
moisture content. Then the hydrated lime was added and mixed through. The lime-treated material was then
left to cure for one hour before application of the foamed bitumen. A foaming agent (0.6% Teric 311) was
added to the Class 170 bitumen to assist with foaming characteristics. The foamed bitumen was then mixed
into the lime-treated materials. The resulting contents of bitumen, hydrated lime and moisture are shown in
Table F 5.
Table F 5: Mix design results after construction
Mean indirect tensile modulus (MPa)

Moisture Dry at 25 °C Ratio
Specimen
Mix type content density(4) wet/dry
no. Initial Cured dry Cured wet
(%) (t/m3) modulus
2.5% 4.6 A 2.130 785 3490 1990 0.57
bitumen/1.5%
B 2.139 725 3220 1890 0.59
hydrated lime
C 2.117 705 3340 1950 0.58
Mean 2.13 738 3350 1943 0.58
5.8 1 2.169 695 3750 2280 0.61
2 2.139 800 4120 2500 0.61
3 2.179 725 4450 2690 0.60
Mean 2.16 740 4110 2490 0.61
3.0 % 4.6 J 2.093 685 2900 1710 0.59
bitumen/1.5%
K 2.101 725 2740 1630 0.60
hydrated lime
L 2.083 680 2830 1840 0.65
Mean 2.09 695 2820 1730 0.61
5.5 M2 2.108 835 3010 1940 0.64
N 2.146 580 4220 2650 0.63
O 2.142 845 3730 2330 0.62
Mean 2.13 753 3650 2310 0.63
TMR minimum requirements for base with 700 4000 2000 0.50
> 1000 ESA/day compacted at 70% standard Proctor
OMC
The loose treated mixture was then compacted into a 152 mm (6 inch) diameter mould by applying 50 blows
to each specimen end with a 10.2 kg Marshall hammer. The specimens were then tested for indirect tensile
modulus (initial, cured dry and cured wet) at a temperature of 25 °C and 40 milliseconds rise time. Table F 5
lists the results.

For each mix, the moduli at the higher moisture contents (5.8% and 5.5%) were compared to the TMR mix
acceptance criteria. The moduli for the 2.5% bitumen mix exceeded the minimum acceptable values. In
addition, the moduli for the 3% bitumen mix exceeded the minimum requirement, except that the mean cured
dry modulus was low due a low result for specimen M2.
The specimens compacted at the mean field moisture (4.6%) have lower mean dry densities than achieved
in the field (Table F 9). This contributed to the mean cured dry moduli being below the TMR acceptance
criteria (4000 MPa).
Figure F 9 shows the variation in modulus with density and moisture content. For the 2.5% mix, the mean
field dry density achieved was about 2.22 t/m3 (Table F 9, Table F 18). At this field dry density, the
extrapolated cured dry modulus was above the TMR minimum (4000 MPa). For the 3% mix, the mean field
dry density of the field cores was about 2.20 t/m3. Figure F 9 indicates that the mean cured dry modulus of
the 3% mix at this density also exceeded 4000 MPa.
Given the scatter of results in Figure F 9, for a given density, the two mixes were not significantly different in
cured dry modulus.
Figure F 9: Variation in modulus with density and moisture content

4600
2.5% bitumen mix at 4.60% moisture
4400 2.5% bitumen mix at 5.8% moisture

3% bitumen mix at 4.6% moisture
4200 3% bitumen mix at 5.5% moisture

2.5% bitumen
E = 14301DD - 26955
4000
3% bitumen
3800 E = 21414DD - 41992
Cured dry
indirect tensile
3600
modulus
(MPa)
3400
3200
3000
2800
2600
2.06 2.08 2.1 2.12 2.14 2.16 2.18 2.2
Dry density (t/m3)
F.5 FBS Thickness Design and Predicted Allowable Traffic Loadings

An objective of the project was to obtain pavement performance data within the project duration. To increase
the likelihood of fatigue cracking, a nominal FBS thickness of 200 mm was agreed based on the experience
of previous trials. However, coring after construction (Table F 17) indicated the mean thicknesses
constructed were 185 mm and 205 mm for the 2.5% and 3.0% mixes respectively. Figure F 10 lists the
resulting pavement compositions.

Figure F 10: Western Freeway pavement compositions
The interim thickness design method (Austroads 2011a) requires the FBS design modulus be determined
and then soaked in water under partial vacuum for 10 minutes.
Consequently, the FBS design moduli were calculated using the mix design mean cured wet moduli
(Table F 6) adjusted to the weighted mean annual pavement temperature (WMAPT) of 20 °C and a design
traffic speed of 80 km/h using the Austroads (2011a) method. In this case, the design modulus of each mix
was limited by the maximum value of 2500 MPa provided in the interim method.
Table F 6: Western Freeway calculations of FBS design moduli
Mix design
IDT modulus at Heavy vehicle FBS design
Mean dry WMAPT
Mix type 25 °C 40 ms rise design speed modulus
Condition density (°C)
time (km/h) (MPa)
(t/m3)
(MPa)
2.5% bitumen Wet 2490 2.16 20 80 2500
3% bitumen Wet 2310 2.13 20 80 2500
For 3% bitumen mix, the predicted FBS fatigue life was 3.3 x 106 ESA using:
• Austroads interim method (Austroads 2011a), including Equation 4 adjusted to 50% design reliability
(Section 8.4.3)
• Table F 7 design inputs.
• As discussed in Appendix F.8, the estimated annual traffic loading was about 1.4 x 106 ESA.
Consequently, the design model predicted there was a 50% chance of observing FBS fatigue cracking
after about 2½ years of trafficking (mid 2017).

Table F 7: Data used to predict FBS fatigue life of the FBS mixes
Pavement Available
Design parameter Technical basis for data selection
component data
Subgrade Subgrade type Clay A presumptive subgrade design CBR of 5% was adopted
for this clay, despite the very high CBR estimated from
DCP results (Figure F 3)
Select fill Thickness 300 mm Based on DCP results (Figure F 3) a design CBR of 10%
was adopted
(top sublayer)
Subbase Subbase thickness 315 mm Assume high standard base material from Table 6.5 of
2.5% bitumen mix GPT Part 2 (Austroads 2012)
Subbase thickness 295 mm
3.0% bitumen mix
(top sublayer)
FBS base FBS thickness 2.5% 175 mm 10 mm less than mean core thickness (Table F 17)
bitumen mix
FBS thickness 3.0% 195 mm
bitumen mix
Design modulus 2.5% 2500 MPa Using mix design modulus of Table F 6
Design modulus 3% 2500 MPa
Volume of bitumen 5.2% Using 2.3% by mass and an FBS dry density of 2.28t/m3
2.5% bitumen mix
Volume of bitumen 6.2% Using 2.7% by mass and an FBS dry density of 2.25t/m3
3% bitumen mix
Wearing Thickness N/A 14/7 mm double/double seal
course
Design modulus N/A
Following a similar process for the 2.5% bitumen mix, an FBS fatigue life of 1.1 x 106 ESA at 50% design
reliability was predicted (Table F 8). For this mix, the design model predicted there was a 50% chance of
observing FBS fatigue cracking after about 1 year of trafficking. The mean core thickness of 2.5% bitumen
mix was 20 mm lower than 3% bitumen mix value. This was the most significant cause of the differences in
the predicted allowable traffic loading of the two mixes.
Table F 8: Predicted allowable traffic loadings
Allowable traffic loading at 50% reliability

FBS mix type
(ESA)
2.5% bitumen mix 1.1 x 106
3.0% bitumen mix 3.3 x 106
F.6 Construction of the FBS Trial Sections
Introduction
Construction of the trial sections was undertaken by VicRoads and the stabilisation industry as a contribution
to this national research project. Viva Energy and Sibelco Australia kindly contributed the stabilising binders.
Two adjacent trial areas were constructed:

• at chainages 0–138 m, the pavement was stabilised with 1.5% hydrated lime and 2.5% bitumen
• at chainages 138–266 m, the pavement was stabilised with 1.5% hydrated lime and 3.0% bitumen.

In addition, a 98 m length of 300 mm thick FBS was constructed for VicRoads about 500 m east from the
Austroads trial area. Construction of this VicRoads patch was similar to that described below.
The construction occurred over three days:

• 2 March 2015: the pavement material was prepared for stabilisation
• 3 March 2015: construction of stabilised pavements
• 4 March 2015: application of sprayed seal surfacing.
The slow-lane pavement area stabilised was a total length of 266 m. The total width stabilised was 4.0 m,
comprising 3.5 m width of the traffic lane, about 0.3 m of the outer shoulder and about 0.2 m of the fast lane.
Preparation of the Material to be Stabilised
As described in Appendix F.2, the existing pavement had 150 mm deep asphalt patches covering about 28%
of the slow lane area (Figure F 11). Based on the observed premature rutting of the stabilised asphalt patch
on the Port Wakefield Road trial (Figure C 5) and in order to improve the uniformity of the material along the
two test areas, it was decided to remove these asphalt patches by milling, cast the millings to the shoulder
and later mix the millings with the crushed rock base.
Figure F 11: Asphalt patches along existing pavement
Figure F 12 illustrates the process to mill and cast the asphalt to the shoulder and also pulverisation of the
remaining area with the profiler.

Figure F 12: Milling of asphalt patches and pulverisation of the remaining area
Source: Prepared by Klaus Kiesel of VicRoads.
Although the patch millings cast to the shoulder included aggregations greater than 20 mm in size
(Figure F 13), these aggregations were later broken down in size by the stabiliser during the lime and
foamed bitumen mixing.
Figure F 13: Asphalt patch millings including coarse aggregations

Imported VicRoads Class 2 wet-mixed crushed rock from Dunnstown quarry was then used to fill the areas
where the asphalt patches had been removed (Figure F 14). This wet-mixed crushed rock was delivered with
higher moisture content than the existing crushed rock base. The moisture content was suitable for
compaction in the untreated state rather than after FBS. It would have been preferable for the crushed rock
to have been delivered at a similar moisture content to the existing base as this would have provided a
material with more uniform characteristics to stabilise. A total of 52 tonnes of imported crushed rock was
placed. It should be noted that after 2 years trafficking, cracking was more prevalent in these areas of wetter
imported crushed rock (Figure F 47, Figure F 48). This may or may not be due to the higher moisture
content.
Figure F 14: Placement of imported crushed rock in the milled areas
After the milled areas had been filled, a grader was used to shift most of the side-cast asphalt millings onto
the pavement and to shape the pavement (Figure F 15). After light rolling with a 16 tonnes smooth drum
roller, the resulting pavement was 20–30 mm high and ready for stabilisation the next day. Excess material
was loaded in a truck using a bobcat. The pavement was left untrafficked overnight.
Figure F 16 shows the pavement at the end of day 1, ready for stabilisation the next day.

Figure F 15: Incorporation of millings and shaping of pavement prior to light rolling
Figure F 16: Pavement ready for stabilisation the next day

Lime Stabilisation
On 3 March 2015, the prepared pavement was firstly mixed with 1.5% of hydrated lime. A spreader was used
to uniformly distribute the lime on the surface in two passes: firstly the 2.4 m wide strip adjacent to the fast
lane and then the 1.6 m wide strip adjacent to the outer shoulder (Figure F 17). Care was taken to minimise
excessive lime at the joint between the two spread runs.
Figure F 17: Spreading of the hydrated lime
To achieve the specified lime content of 1.5% for a target 200 mm depth (plus an allowance of 20 mm cut to
waste), the calculated spread rate was 6.6 kg/m2 assuming a dry density of 2.00 t/m3. It is common to add
0.2 kg/m2 for variability, hence the target spread rate used by the contractors was 6.8 t/m2. The total amount
of lime spread was 7.26 tonnes as measured by the spreader. For a total area of 1064 m2, this equated to an
average spread rate of 6.8 kg/m2, which was similar to the initial target spread rate. This spread rate was
also checked by measuring the amount of lime spread using a mat placed on the surface.
Following construction, the average percentage lime applied was calculated using:
• the 7.26 tonnes of lime spread
• the mean measured dry density in situ (2.20 t/m3, Table F 9 and Table F 18)
• the mean FBS layer thicknesses of 195 mm of the total 265 m trial length (Table F 17), plus 20 mm
assumed cut to waste.
The calculated average lime content was 1.4%, close to the specified target of 1.5%.

The lime was then mixed with the pavement material using Wirtgen Stabiliser WR2400 (Figure F 18) to a
depth of 180 mm, which was 20 mm less than the target depth of 200 mm to reduce the risk of leaving
disturbed material under the foamed bitumen stabilised material. Water was added during mixing from the
water tanker connected to the stabiliser.
The pavement was then shaped using a grader and lightly-rolled using an 18 tonne padfoot roller and a 16 tonne
vibrating smooth drum roller (Figure F 19) to provide a surface for the application of the foamed bitumen.
Figure F 18: Mixing of crushed rock/RAP, lime and water
Figure F 19: Preparing the pavement surface for application of foamed bitumen

Foamed Bitumen Stabilisation
About three hours after the lime was mixed through, foamed bitumen stabilisation commenced.
To provide adequate foaming characteristics, Teric 311 foaming agent was mixed with the
Class 170 bitumen in the tanker at a concentration of 0.6%.
With the bitumen tanker and the stabiliser connected, foamed bitumen was produced and mixed through the
lime-treated crushed rock/RAP base (Figure F 20). The two temperature gauges on the bitumen tanker
indicated that the bitumen temperature was about 172 °C during foaming of the 2.5% mix, and 164 °C during
foaming of the 3.0% mix.
Figure F 20: Mixing the foamed bitumen through the lime-treated material
As described above, sections with the following bitumen contents were constructed:
• chainages 0–138 m: the specified bitumen content was 2.5%
• chainages 138–266 m: the specified bitumen content was 3.0%.
To achieve a specified bitumen content of 2.5% at the target 200 mm depth (plus an allowance of 20 mm cut
to waste), the initial target bitumen application rate at the field temperature was 12.1 litres/m2, assuming a
dry density of 2.00 t/m3 (later measured to be 2.20 t/m3).
In this section, the total volume of bitumen used was 6150 litres, as read from dipping of the tanker. The
actual percentage by mass of bitumen added to the crushed rock/RAP was calculated assuming:
• 6150 litres equates to 5590 kg of bitumen
• an average depth of stabilisation of 185 mm as estimated by coring (Table F 17), plus an allowance of
20 mm cut to waste
• given a trial length of 138 m and width 4 m, the total area stabilised was 552 m2
• the mean measured dry density of 2.21 t/m3 (Table F 9).

The calculated average bitumen content was 2.3%, slightly below the target 2.5%.
Similarly, for the section with a specified content of 3.0%, the total volume of bitumen used was 7550 litres.
The actual percentage by mass of bitumen added to the crushed rock/RAP was calculated assuming:
• 7550 litres equates to 6864 kg of bitumen
• the average depth of stabilisation was 205 mm as estimated by coring (Appendix F.9.3), plus an
allowance of 20 mm cut to waste
• given a trial length of 138 m and width 4 m, the total area stabilised was 552 m2
• the mean measured dry density was 2.23 t/m3 (Table F 9).
The calculated average bitumen content was 2.7%, slightly below the target 3.0%.
The foaming characteristics of the bitumen were checked during the operation.
The FBS crushed rock base was then compacted using an 18 tonne padfoot roller without the application of
any additional water (Figure F 21).
Figure F 21: Initial compaction using padfoot roller
The grader, a 16 tonne vibrating smooth drum roller and an 18 tonne multi-tyred roller then completed the
compaction with addition of water to the surface as required (Figure F 22). The pavement was left slightly
high (up to 30 mm) in level to allow reshaping the next day, prior to sealing.

Figure F 22: Shaping and rolling of the surface at end of day 2
The pavement was not opened to traffic that night due to the late finish and concern that the pavement had
insufficient time to cure prior to traffic loading.
Preparation of the Surface for Sealing
Early on day 3, the pavements were inspected, and it was noted that due to the late finish the previous day,
there were still some slight padfoot indentations on the pavement surface. Accordingly, water was applied to
the surface and the top 20 mm of the surface was reshaped with a grader and cut to level. The surface was
then rolled using smooth drum and multi-tyred rollers (Figure F 23).
Figure F 23: Reshaping and final compaction prior to sealing
The transverse pavement shape surface was then assessed by measuring the deviation from a 3 m straight
edge (Figure F 24). Deviations up to 15 mm were measured but were commonly 10 mm or less. As
described in Appendix F.9.4, these deviations may be the cause of the rutting measured soon after opening
to traffic, rather than rutting due to the applied traffic loading.

Figure F 24: Check of surface shape
Double/Double Sprayed Bituminous Seal
About two hours after the above rolling was completed, the surface had dried and needed to be lightly
watered prior to light brooming to remove loose aggregates whilst retaining the surface fines (Figure F 25).
Figure F 25: Preparation of the base for application of the sprayed seal
A double/double seal (Figure F 26), consisting of size 14 mm aggregate followed by 7 mm aggregate, was
then placed using a total binder application rate of 1.8 l/m2 (1.2 l/m2 and 0.6 l/m2). The sprayed binder
included 10 parts cutter oil and 5 parts rubber (Figure F 27). As mentioned later the inclusion of the cutter oil
may have contributed to bleeding of the bitumen and/or sealing aggregate embedment.

Figure F 26: Double/double seal
Figure F 27: Sprayed bituminous sealing
F.7 Sampling and Testing During Construction
Field Density Testing
The field wet densities were measured at four sites in each area at the commencement of day 3, prior to the
final surface trim (Figure F 28). At each test site, a sample of FBS material was excavated for determination
of oven-dried moisture content. The wet density, moisture contents and dry densities are given in Table F 9.

Figure F 28: Nuclear gauge direct transmission field density measurements
Table F 9: FBS field wet densities, moisture contents and dry densities
Modified
Wet Moisture Dry Modified Density Moisture
Project Proctor
Mix density content density Proctor ratio ratio
chainage MDD(1)
(t/m3) (%) (t/m3) OMC (%) (%) (%)
(t/m3)
2.5% 20 2.26 4.5 2.16 2.17 7.0 99.7 64
40 2.29 3.5 2.21 2.14 6.7 103.5 52
60 2.35 5.2 2.24 2.12 7.0 105.3 74
80 2.34 5.1 2.22 2.23 7.0 100.0 73
Mean 2.31 4.6 2.21 2.16 6.9 102.1 66
3.0% 180 2.29 5.2 2.18 2.17 7.5 100.5 69
200 2.32 4.4 2.22 2.16 7.2 102.7 61
220 2.36 4.9 2.25 2.08 7.7 108.1 64
240 2.34 3.9 2.25 2.13 7.5 105.6 52
Mean 2.33 4.6 2.23 2.14 7.5 104.2 62
1 The measured values were increased by 2% due to the time delay between field mixing and laboratory compaction.

Table F 10 compares the mean dry densities measured with the nuclear gauge to the mean densities
subsequently obtained by coring the pavement (Table F 18). The core densities were measured using two
methods (mensuration and presaturation, as described in Section 4.1). As early-life field cores were friable,
they tended to have more air voids on the surface than a laboratory prepared test specimen. As a result, the
volume by mensuration is overestimated and hence the dry density underestimated. Conversely, the
presaturation method excludes surface voids and overestimates the dry density. The nuclear gauge densities
were between-core densities measured using the mensuration and presaturation methods.
Table F 10: Comparison of nuclear gauge and coring dry densities
Mean dry density (t/m3)

Coring (from Table F 18)
FBS mix type
Nuclear gauge Mensuration method(1) Presaturation method(2)
Upper portion Lower portion Upper portion Lower portion
2.5% bitumen 2.21 2.22 2.18 2.29 2.26
3.0% bitumen 2.23 2.15 2.15 2.27 2.26
2 ASTM C642-13, measured dry densities are similar to AS/NZS 2891.9.2:2014 Presaturation Method.
The modified Proctor MDD and OMC results for each field site are also listed in Table F 9. These results
were derived from laboratory compaction of samples taken from the roadbed after mixing and prior to
compaction (Appendix F.7.6). As listed in Table F 13, there was a time delay between field sampling and
laboratory compaction. In a separate study, a time delay of 24 hours reduced the density of the Marshall
specimens by 2%. The MDD values determined on the field-mixed materials (Table F 13) were likely to be
similarly affected. Accordingly, the field-mixed modified Proctor MDDs were increased by 2% for use in the
density ratio calculations as given in Table F 9.
The mean field density equated to about 103% modified Proctor MDD. The mean field moisture content of
4.6% was about 65% of the modified Proctor OMC.
Properties of Untreated Material Sampled from the Roadbed
Samples of the untreated material were obtained from the roadbed at the end of day 1, after the material had
been prepared for stabilisation (Appendix F.6.2).
These samples were tested for particle size distribution and plasticity, and the modified Proctor MDD and
OMC were measured as reported in Appendix F.4.2.
These samples were also used in the mix design testing after construction, the results of which are also
described in Appendix F.4.2.
Lime Content
After mixing through the lime and prior to compaction, samples of the lime-treated material were obtained at
four chainages for determination of lime content in accordance with ASTM C25-11e1. Samples were also
obtained at the same chainages prior to addition of lime to measure the inherent lime content of the crushed
rock/asphalt millings mixture. A bulk sample of the hydrated lime was also obtained behind the spreader and
the available lime index as Ca(OH)2 was measured to be 94%. The results, presented in Table F 11, suggest
the average hydrated lime content was close to 1% which was below the target of 1.5% hydrated lime. This
result was inconsistent with the contractor’s record of the amount of lime used, which was calculated to be
1.4% (Appendix F.6.3).

Table F 11: Lime index and lime percentage of untreated and lime-treated materials
Material prior to lime treatment Lime-treated material

Sample chainage
(m) Lime index as Lime content Lime index as Hydrated lime
Ca(OH)2 (%) Ca(OH)2 content added (%)
45 0.046 0.08 0.702 1.16
100 0.034 0.06 0.280 0.46
180 0.034 0.06 0.550 0.91
230 0.034 0.06 0.710 1.18
Mean 0.93
Bitumen Content
As described in Appendix F.6.4, the contractor advised the volumes of bitumen applied to the two trial
sections and these were used to calculate the percentage bitumen by mass added by stabilisation. The
calculations indicated that the constructed bitumen contents were slightly below the design values, namely
2.3% bitumen for the nominal 2.5% bitumen mix and 2.7% for the nominal 3% bitumen mix.
To confirm these estimates, samples of FBS material were taken from the trial sections after mixing through
the foamed bitumen and prior to compaction at four chainages for each mix. As the material that was
stabilised included asphalt milling and old sprayed seals, it was necessary to determine the bitumen content
prior to the addition of the foamed bitumen. Samples were taken at two chainages for each mix after lime
treatment but prior to the application of foamed bitumen. To supplement these results, samples taken prior to
both lime and foamed bitumen treatment were also tested for bitumen content. The bitumen contents of all
samples were measured using the ignition oven method (Austroads 2005).
The results, shown in Table F 12, indicate high variability in recovered bitumen contents, most likely due to
material that was stabilised including variable amounts of multiple old sprayed seals and millings from the
asphalt patches. It was concluded that the estimated bitumen added by this method was unreliable due to
the variability in the bitumen content prior to foaming. However, it does highlight the potential to improve the
construction process used at Ballan to produce a more uniform material to be stabilised.
Table F 12: Bitumen contents of materials sampled from the roadbed
Recovered bitumen content (% by mass) Estimated

bitumen
Sample Untreated Lime-treated content added
Mix type Material after
chainage crushed material prior to by foamed
foamed bitumen
rock/asphalt foamed bitumen bitumen
application
millings application (% by mass)
2.5% bitumen/ 40 4.75 – 6.7
1.5% hydrated lime
45 – 5.0
60 4.5 – 6.9
80 0.6 – 4.5
100 1.35 1.8 4.4
Mean 3.0 5.6 2.6
3% bitumen/ 180 1.55 4.2 6.3
1.5% hydrated lime
200 1.65 – 6.0
220 3.10 – 5.8
230 – 3.3 –
240 3.15 6.5
Mean 2.8 6.2 3.3

Particle Size Distribution of Field-mixed Materials
Loose foamed bitumen stabilised crushed rock was sampled behind the stabiliser but prior to compaction to
determine the particle size distribution after completion of field mixing and prior to compaction. Note that at
this stage in construction, the stabiliser had broken down most of the large aggregations of asphalt patch
millings (Figure F 13) initially present.
The mean particle size distributions for the mixes were within the limits recommended for heavily trafficked
roads with greater than or equal to 1000 ESA per day (Austroads 2011b).
Figure F 29: Mean particle size distribution of roadbed material prior to compaction
100
90
Upper limit proposed Austroads
80 2.5% mix mean sampled from road bed
3% mix mean sampled from the road bed

70
60
Percentage
50
passing
40
30
20
10
0
0.01 0.1 1 10
Sieve size (mm)
Maximum Dry Densities and Optimum Moisture Contents of Field-mixed

Material
Loose FBS crushed rock was sampled from the roadbed after mixing and prior to compaction at or near the
chainages, at or near where the field densities were measured (Table F 9). The samples were transported to
Downer’s laboratory to determine modified Proctor MDD and OMC in 100 mm diameter moulds. While
laboratory resources allowed the laboratory compaction of 3% bitumen mixes on the day of field stabilisation,
the 2.5% bitumen mix specimens were compacted the following morning. The time delays between field
mixing and laboratory compaction are listed in Table F 13, together with the MDD and OMC values.
It is apparent that there is considerable variation in the MDD values along each trial. Variations in the
amounts of asphalt millings along the project may have contributed to this variation.
It is noted that the mean MDD of the field-mixed materials are about 2% lower than the values determined on
laboratory-mixed specimens (Table F 4). However, the MDD of the field-mixed materials were affected by
time delays between the field mixing and laboratory compaction. The effect of a 24-hour delay between
mixing and compaction was investigated in a separate laboratory investigation, as detailed in Section 6. It
was observed that a 24-hour delay between laboratory mixing and compaction also resulted in the densities
reducing by about 2%. Accordingly, in the field compaction assessment (Table F 9), the MDD values in
Table F 13 were reduced by 2% to allow for the time delay.

Table F 13: Modified Proctor MDD and OMC of field-mixed FBS material sampled from roadbed
Delay before Modified Proctor Modified Proctor

Mix type Sample chainage laboratory compaction MDD OMC
(hours) (t/m3) (%)
2.5% bitumen/ 40 18.5 2.095 6.7
1.5% hydrated lime
60 19.5 2.080 7.0
80 21.0 2.183 7.0
100 22.0 2.148 7.3
Mean 2.13 7.0
3% bitumen/ 180 3.25 2.127 7.5
1.5% hydrated lime
200 4.25 2.120 7.2
220 5.75 2.040 7.7
240 6.75 2.088 7.5
Mean 2.09 7.5
Loose foamed stabilised crushed rock from the roadbed after mixing and prior to compaction was also used
to prepare modulus test specimens. The samples were transported to Downer’s laboratory and compacted
the same day into 152 mm diameter Marshall moulds by applying 50 blows to each end with a 10.2 kg
Marshall hammer.
The mean measured cured dry, cured wet moduli and modulus ratio values are given in Table F 14, together
with the time delays between field mixing and laboratory compaction. The individual cured dry moduli results
are plotted in Figure F 30. As expected, there was a tendency for the densities and moduli of both mixes to
increase with moisture content.
Note the significantly higher moduli of the 3% FBS mix than the 2.5% mix, a difference that was not apparent
in the mix design testing (Appendix F.4.2). In part, the higher modulus of 3% mix may be due to the shorter
delay between field mixing and laboratory compaction for the 3% mix, as seen from the delay times listed in
Table F 14. The moisture contents of the 3% mix at the time of laboratory compaction were also higher.
Figure F 30: Modulus variation with density, field moisture content marked
4000
6.4% 6.4%
3500
2.5% bitumen
3.0% bitumen
3000
4.4% 6.4%
5.3%
4.2%
Cured dry 2500 5.3%
4.7% 4.2% 4.7%
indirect tensile
modulus 4.4%
4.2% 5.3%
4.4%
(MPa) 2000
4.1% 4.1%
4.1%
1500 4.7%
3.9% 4.0% 3.9%

1000
3.9%
4.0% 4.0%
500
1.96 1.98 2.00 2.02 2.04 2.06 2.08 2.10 2.12
Dry density (t/m3)

Table F 14: Modulus of laboratory-compacted field-mixed FBS material

Dry Moisture
Time delay Cured Cured Ratio
Mix Chainage density(3) content
(hours:min) dry modulus(1) wet modulus(2) wet/dry
(t/m3) (%)
modulus
2.5% 40 5:20 1.99 4.0 800 633 0.79
60 5:50 1.99 3.9 930 763 0.82
80 6:20 2.08 5.3 2490 1500 0.61
100 6:50 2.05 4.7 2420 1520 0.67
Mean 6:05 2.03 4.5 1660 1100 0.72
3.0% 180 3:05 2.05 4.4 2350 1560 0.66
200 3:35 2.06 4.2 2390 1490 0.63
220 4:05 2.04 6.4 3390 2370 0.70
240 4.35 1.98 4.1 1690 1280 0.76
Mean 3:50 2.04 4.8 2450 1670 0.69
TMR minimum base requirements (> 1000 ESA/day) for
specimens compacted in 150 mm diameter mould and large 4000 2000 0.50
Marshall hammer
Note that the moisture contents at which these specimens were mixed and compacted were generally well
below those commonly used in mix design (Appendix F.4.2). This complicates assessment against the
specified minimum mix design moduli. The results are discussed further in Section 5.3.
F.8 Traffic Monitoring

To evaluate field performance of the FBS mixes, it is necessary to quantify the cumulative heavy vehicle
traffic loading. VicRoads provided the 2012, 2013, 2014 and 2017 heavy vehicle volumes of Western
Freeway between Old Melbourne Road and Ballan-Daylesford Road, as given in Table F 15. This data
suggests a low growth rate in heavy vehicle volumes. Assuming an annual ESA growth rate of 1%,
Figure F 31 shows the predicted cumulative traffic loading from opening in March 2015.
Table F 15: Traffic volume for the Western Freeway westbound carriageway
Lane
Average Estimated annual ESA/heavy Annual ESA in
Year distribution
daily heavy vehicles number of heavy vehicles vehicle slow lane
factor
2012 1 900(1) 694 000 2.0(2) 0.98 1.4 x 106
2013 1 800(1) 657 000 2.0(2) 0.98 1.3 x 106
2014 1 900(1) 694 000 2.0(2) 0.98 1.4 x 106
2017 1 800(1) 657 000 2.0(2) 0.98 1.3 x 106
1 Western Freeway between Old Melbourne Road and Ballan-Daylesford Road, supplied by VicRoads.
2 Weigh in motion data for Western Highway Ballan westbound carriageway late 2017.

Figure F 31: Predicted cumulative traffic loading for trial section

4.0E+06
3.5E+06
3.0E+06
2.5E+06
Cumulative 2.0E+06
ESA
1.5E+06
1.0E+06
5.0E+05
0.0E+00
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 2.6 2.8 3
F.9 Performance Monitoring
Surface Deflections
Falling weight deflectometer (FWD) pavement deflection testing was undertaken one month (April 2015), one
year (March 2016) and two years (March 2017) after opening to traffic. Measurements were taken:
• in the outer wheel path
of 566 kPa. The FBS base mid-depth temperatures were measured during the deflection testing. During the
April 2015 deflection testing, the FBS average temperature was 18 °C during the between wheel path
measurements, and 19 °C for the outer wheel path measurements.
During the March 2016 and March 2017 deflection testing, the FBS average temperatures were 22 °C and
18 °C, respectively. The maximum deflections and curvature were corrected from the pavement temperature
at the time of measurement to a WMAPT of 20 °C using the interim deflection-temperature method
developed in Appendix D.15.
The temperature-corrected maximum deflections and curvatures are shown in Figure F 32 and Figure F 33.
One month after opening to traffic, the curvatures were similar to those measured on the Calder Freeway
trial section (Figure D 15, Figure D 16).
Comparing the 1 month and 1 year results, the maximum deflections and curvatures decreased less in the
outer wheel path than between wheel paths, consistent with Calder Freeway trial findings. After one year of
trafficking these results suggested that the FBS layer had accumulated fatigue damage.

After two years trafficking, consistent with the extensive surface cracking (Appendix F.9.5), outer wheel paths
deflections and curvatures increased in variability. The deflections and curvatures between the wheel paths
also tended to increase where there was significant cracking in the adjacent wheel paths.
Figure F 32: Western Freeway FBS maximum deflections

0.70
1 month outer wheel path
1 year Outer wheel path

0.60
2 years outer wheel path
0.50
Maximum 0.40
deflection at
566 kPa
(mm) 0.30
0.20
0.10
2.5% FBS mix 3% FBS mix
0.00
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (km)
0.50
1 month between wheel paths
1 year between wheel paths
0.40 2 years between wheel paths
0.30
Maximum
deflection at
566 kPa
(mm)
0.20

0.10
0.00
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (km)

Figure F 33: Western Freeway FBS curvatures

0.25
1 month outer wheel path
1 year outer wheel path
2 years outer wheel path
0.20

0.15
Curvature
(D0-D200) at
566 kPa
(mm)
0.10
0.05
0.00
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (km)
0.14
1 month between wheel paths
1 year between wheel paths

0.12
2 years between wheel paths
0.10
Curvature 0.08
(D0-D200) at
566 kPa
(mm) 0.06
0.04
0.02
0.00
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (km)
Back-calculated Moduli
Using the measured surface deflection bowls 1 month (April 2015), 1 year (March 2016) and 2 years
(March 2017) after opening to traffic and the mean constructed FBS thickness of 185 mm and 205 mm for
the 2.5% and 3.0% mixes respectively (Table F 17), the pavement and subgrade layer moduli were
back-calculated. The results are listed in Appendix G.3.

Of particular interest is the variation of FBS moduli with time, and also the differences between values in the
unloaded pavement areas (between wheel paths) and loaded pavement area (outer wheel path). For this
comparison, the back-calculated moduli were adjusted from the pavement temperature during FWD testing
to WMAPT for Ballarat of 20 °C using the Austroads (2011a) modulus-temperature dependency.
The temperature-corrected FBS moduli are plotted in Figure F 34 and Figure F 35. The mean FBS moduli
are given in Table F 16. Note that the Ballan trial back-calculated FBS moduli tend to be lower than the
Calder Freeway trial (Figure D 19), despite a substantially higher relative compaction being achieved.
In the untrafficked area between wheel paths, the FBS moduli after 1 year roughly doubled from the 1 month
values, reasonably consistent with the trend at the Calder Freeway trial (Figure D 20). After one year of
trafficking, the FBS moduli in the outer wheel path have increased less, consistent with the hypothesis that
the FBS is being damaged due to the applied loading.
After two years of trafficking, the FBS moduli in the outer wheel path were highly variable consistent with the
significant extent of surface cracking (Appendix F.9.5).
Table F 16: Mean FBS moduli back-calculated
Mean FBS modulus at 20 °C (MPa)

FBS mix Age
Between wheel paths Outer wheel paths
2.5% bitumen + 1.5% hydrated lime 1 month 2720 2210
1 year 6020 3790
2 years 5340 3180
3% bitumen + 1.5% hydrated lime 1 month 2590 1970
1 year 5150 2590
2 years 5400 3220
Figure F 34: FBS moduli of 2.5% mix back-calculated from surface deflections
2.5 % mix : 1 month after construction
5000
4500
Outer wheel path
3500
FBS
modulus 3000
at 20°C
(MPa)
2500
2000
1500
1000
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Chainage (m)

2.5 % mix : 1 year after opening to traffic

10000
9000
Outer wheel path

8000
Between wheel paths
7000
6000
FBS
modulus 5000
at 20°C
(MPa)
4000
3000
2000
1000
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Chainage (m)
2.5 % mix : 2 years after opening to traffic

10000
surface cracking
visible in 2017
9000
8000
7000
Asphalt patch
placed late 2016
6000
FBS
modulus 5000
at 20°C
(MPa)
4000
3000
2000
Outer wheel path
1000
Between wheel paths
0
0 10 20 30 40 50 60 70 80 90 100 110 120 130 140
Chainage (m)

Figure F 35: FBS moduli of 3% mix back-calculated from surface deflections

3% FBS mix : one month after construction
7000
6000 Outer wheel path 1 month

Between wheel paths 1 month
5000
FBS
modulus 4000
at 20°C
(MPa)
3000
2000
1000
140 150 160 170 180 190 200 210 220 230 240 250 260
Chainage (m)
3% FBS mix : 1 year after opening to traffic

9000
Outer wheel path 1 year

8000
Between wheel paths 1 year
7000
6000
FBS 5000
modulus
at 20°C
(MPa) 4000
3000
2000
1000
0
140 150 160 170 180 190 200 210 220 230 240 250 260
Chainage (m)

3% FBS mix : 2 years after opening to traffic
11000
surface cracking
visible in 2017
10000
9000
8000
7000
FBS
6000
modulus
at 20°C 5000
(MPa)
4000
3000
2000

Between wheel paths
0
140 150 160 170 180 190 200 210 220 230 240 250 260
Chainage (m)
Field Core Moduli
One month after construction, 100 mm diameter cores of both FBS mixes were taken between the wheel
paths at five chainages. The cores were taken using water to lubricate and cool the pavement and hence the
cores wetted up during extraction. Figure F 36 shows two of the extracted cores after air-drying in the
laboratory. In some cases, the entire FBS layer was not recovered due to low strength at this early age.
Figure F 36: Examples of cores taken one month after construction
One year after opening to traffic, an additional 9 cores were taken between wheel paths, followed by an
additional four cores after 2 years. In these cases, dry ice and compressed air were used to cool the
pavement (Figure F 37) so that the cores were not affected by saw-cutting water.

Figure F 37: Coring using dry ice and compressed air
The heights of each core were measured (Table F 17) to enable FBS thickness to be estimated. The cores
taken at 1 year and 2 years were more consistently near the design thickness of 200 mm. Based on this
coring at 1 year and 2 years, the mean constructed FBS thicknesses were about 185 mm and 205 mm for
the 2.5% and 3.0% mix sections respectively.
Table F 17: Ballan FBS layer thicknesses measured from field cores
FBS mix Age Chainage (m) FBS thickness (mm) Comment
2.5% mix 1 month 21.0 163
40.0 167
60.0 189
80.0 176
100.0 176
1 year 29.5 180
49.5 195
69.5 193
89.5 185
109.5 180
2 years 119.0 183
129.0 184
Mean 1 year and 2 year cores 185

FBS mix Age Chainage (m) FBS thickness (mm) Comment

3.0% mix 1 month 160.0 172
180.0 167
200.0 136 Lower portion not
retrieved
220.0 149 Upper portion not
retrieved
240.0 147 Lower portion not
retrieved
1 year 169.5 205
189.5 205
209.5 210
229.5 210
2 years 149.0 200
249.0 198
Mean 1 year and 2 year cores 205
It was of interest to measure the dry densities and moduli of the field cores. The dry densities were
measured using two methods: the mensuration method and presaturation method as described in
Section 4.1.
Early-life cores
There was a 2-month delay between the field coring 1 month after opening to traffic and the modulus testing,
resulting in the cores being about 3-months old at the time of testing. The cores were stored in a
temperature-controlled laboratory in sealed plastic bags between sampling and testing.
After trimming each end of the cores, they were cut into upper and lower halves.
The results are listed in Table F 18. It is noted that differences in the dry density results using the
mensuration and pre-saturation method clearly exceeded:
• the 2.5% density difference estimated from parallel testing during mix design (Section 4.3.3)
• about 1% obtained in testing of 1-year-old cores (Table F 19).
This casts doubt on the use of the mensuration method for these early-life cores. Accordingly, the mix moduli
were compared using the dry densities determined using the presaturation method (Figure F 38). The results
suggest that the two mixes have similar moduli in situ. However, for a given dry density, the 3% bitumen mix
has a higher modulus than the 2.5% bitumen mix.
It was of interest to compare the mix design moduli (Table F 5) with those of the 3-month-old cores
(Table F 18). As water was used in the saw cutting, the cured dry moduli of the cores were compared to the
mix design mean cured dry modulus. For this comparison, the Table F 5 mix design mean dry densities
measured using the mensuration method were increased by 2.5% (Section 4.3.3) to equivalent presaturation
method values.
As shown in Figure F 38, the densities of the mix design specimens were below the high compaction levels
achieved in this field trial. This hinders an assessment on whether the mix design values are consistent with
the values measured on 3-month-old cores. Nevertheless, cured dry modulus of individual test cylinders
measured in the mix design appear consistent, except for one of the 2.5% mix results. This finding supports
that observed in the Calder Freeway trial (Figure D 24 ) and Newell Highway trial (Figure E 29), that the
moduli measured in the mix design reflect early-life in situ moduli.

Table F 18: Modulus and density of three-month-old field cores taken between wheel paths
Project Upper or Age when cured dry and Indirect tensile modulus at 25 °C (MPa) Ratio
Dry density(3) Dry density(4)
FBS mix chainage lower wet moduli measured As received Cured Cured wet/dry
(t/m3) (t/m3)
(m) portion (days) from field dry modulus(1) wet modulus(2) modulus
2.5% mix 21 Upper 77 2.141 2.252 3 230 3 870 2 090 0.54
Middle 77 2.216 2.284 4 900 6 820 4 080 0.60
Lower 77 2.190 2.274 7 740 7 810 6 290 0.80
40 Upper 76 2.219 2.291 3 730 8 430 7 040 0.83
Lower 76 2.176 2.259 4 460 5 370 3 440 0.64
60 Upper 93 2.247 2.292 5 230 6 970 6 720 0.96
Lower 93 2.159 2.23 2 890 3 860 2 660 0.69
80 Upper 104 2.221 2.295 8 300 8 580 4 880 0.57
Lower 91 2.205 2.283 7 250 6 970 4 080 0.59
100 Upper 93 2.294 2.344 12 310 13 300 10 730 0.81
Lower 93 2.172 2.274 4 930 4 440 2 460 0.55
Upper 89 2.22 2.29 6 560 8 230 6 290 0.74
Mean
Lower 86 2.18 2.26 5 450 5 690 3 780 0.65
3% mix 160 Upper 77 2.133 2.239 4 160 5 500 3 600 0.65
Lower 77 2.167 2.262 6 160 7 980 5 550 0.70
180 Upper 104 2.131 2.264 9 600 9 260 6 100 0.66
Lower 104 2.141 2.257 4 490 5 860 3 850 0.66
200 Upper 104 2.222 2.289 11 390 11 230 9 080 0.81
220 Upper 104 2.136 2.303 7 140 9 300 7 240 0.78
240 Upper 77 2.123 2.234 6 230 5 780 5 040 0.87
Upper 93 2.15 2.27 7 700 8 210 6 210 0.75
Mean
Lower 91 2.15 2.26 5 330 6 920 4 700 0.68
4 AS/NZS 2891.9.2:2014 Presaturated Method.

Figure F 38: Western Freeway field-core moduli variation with density using presaturation method
(a) core moduli as received from the field
13000
12000 2.5% bitumen
3% bitumen
11000
10000
y = 61487x - 132183
R² = 0.34 y = 79097x - 174419
9000
R² = 0.68
Indirect 8000
tensile
modulus
7000
(MPa
6000
5000
4000
3000
2000
2.22 2.24 2.26 2.28 2.30 2.32 2.34 2.36
Dry density using presaturation method (t/m3)
(b) cured dry moduli
14000
2.5% bitumen
13000 3% bitumen
y = 73996x - 159686
12000
Mix design modulus : 2.5% bitumen
R² = 0.70
Mix design modulus: 3% bitumen mix
11000 y = 86921x - 191216
R² = 0.86
10000
Cured dry 9000

indirect
tensile 8000
modulus
(MPa) 7000
6000
5000
4000
3000
2000
2.14 2.16 2.18 2.20 2.22 2.24 2.26 2.28 2.30 2.32 2.34 2.36
Dry density using the presaturation method (t/m3)

(c) cured wet moduli

12000
11000 2.5% bitumen
10000 3% bitumen
9000
y = 61934x - 134439 y = 76172x - 168708
R² = 0.64 R² = 0.74
8000
Cured wet
indirect 7000
tensile
modulus 6000
(MPa)
5000
4000
3000
2000
1000
2.22 2.24 2.26 2.28 2.30 2.32 2.34 2.36
Dry bulk density using the presaturation method (t/m3)
Cores taken at 1 year
There was a 2-month delay between the field coring and the modulus testing, resulting in the cores being
about 14-months old at the time of testing. The cores were stored in a temperature-controlled laboratory in
sealed plastic bags between sampling and testing.
After trimming each end of the cores, the cores were cut into upper and lower portions each 60 mm high.
The upper portion related to the material from about 20 mm to 80 mm below the surface. Given the layer
thickness was about 200 mm, this portion represents the top half of the pavement layer. The lower portion
related to material 90 mm to 150 mm below the surface. As such, the cores overestimated the densities and
moduli of the lower half of the FBS layer.
The results are listed in Table F 19 and Figure F 39. Both the densities and the moduli are considerably
higher than the values obtained on the early-life cores. For instance, plotted in Figure F 39 are the early-life
and 1 year core results for the 3% mix. For a given density, the 3% bitumen mix core moduli have increased
about 50% between the values tested at 3 months and 14 months. This is less than the increase (about
100%) in the modulus back-calculated from the deflections between testing at 1 month and 12 months.

Figure F 39: Effect of age on untrafficked moduli of 3% bitumen mix cores extracted from the Western
Freeway trial
22000
sampled at 1 month tested at 3 months

20000
y = 39753x - 74536
sampled at 12 months tested at 14 months R² = 0.65
18000
sampled at 24 months tested at 25 months
16000
y = 65870x - 136574
R² = 0.60
Cured dry 14000
indirect
tensile modulus
(MPa) 12000
10000
8000 y = 73996x - 159686

R² = 0.70
6000
4000
2.1 2.15 2.2 2.25 2.3 2.35
Dry density (t/m )3

Table F 19: Modulus and density of 1-year-old field cores taken between wheel paths
Indirect tensile modulus at 25 °C (MPa)

Project Upper or Age when cured dry and Dry Ratio
Dry density(4)
FBS mix chainage lower wet moduli measured density(3) As received Cured Cured wet/dry
(t/m3)
(m) portion (months) (t/m3) from field dry modulus(1) wet modulus(2) modulus
2.5% mix 29.5 Upper 14 2.285 2.319 12 530 14 460 10 580 0.73
Middle 14 2.244 2.291 5 020 5 520 2 990 0.54
49.5 Upper 14 2.287 2.286 7 210 6 930 7 640 1.10
Lower 14 2.195 2.216 5 830 6 040 5 680 0.94
69.5 Upper 14 2.305 2.319 15 970 15 740 13 620 0.86
Lower 14 2.231 2.27 11 210 11 260 8 390 0.75
89.5 Upper 14 2.298 2.329 16 590 17 020 15 820 0.93
Lower 14 2.252 2.303 10 900 11 900 6 550 0.55
109.5 Upper 14 2.318 2.343 18 240 20 470 17 220 0.84
Lower 14 2.270 2.31 16 050 17 320 15 310 0.88
Upper 2.30 2.32 14 110 14 920 12 970 0.89
Mean
Lower 2.24 2.28 9 800 10 410 7 780 0.73
3% mix 169.5 Upper 14 2.264 2.293 12 990 13 120 10 860 0.83
Lower 14 2.225 2.268 10 870 11 830 7 030 0.59
189.5 Upper 14 2.262 2.284 10 260 13 430 10 040 0.75
Lower 14 2.247 2.258 12 280 13 540 9 200 0.68
209.5 Upper 14 2.253 2.286 15 360 16 080 15 570 0.97
Lower 14 2.226 2.259 11 660 11 150 8 450 0.76
229.5 Upper 14 2.247 2.268 13 880 13 390 10 050 0.75
Lower 14 2.186 2.22 10 340 9 500 7 090 0.75
Upper 2.25 2.28 12 460 13 280 10 860 0.81
Mean
Lower 2.22 2.25 11 290 11 500 7 940 0.70

Cores taken at 2 years
After 2 years trafficking, four between wheel path cores were extracted using dry ice coring. Unexpectedly,
the bottom portion of each core was wet (Figure F 40). The fact that this water was not observed when
coring the pavements after 1 year of trafficking suggested the water may have infiltrated the pavement after
the surface cracked. However, as three of the four cores taken at 2 years were well away from areas of
surface cracking, ground water may also be the cause.
Figure F 40: Example of moisture in the bottom half of field cores
Following a process similar to that used to test the 1-year cores, the top and bottom half core densities and
moduli were measured (Table F 20). The 2-year core moduli are higher than the 1-year values, indicating
curing continued beyond the first 12 months as illustrated in Figure F 39.

Table F 20: Modulus and density of 2-year-old field cores taken between wheel paths
Age when Indirect tensile modulus at 25 °C (MPa)

Project cured dry and Ratio
Upper or Dry density(3) Dry density(4) As received Cured Cured
FBS mix chainage wet moduli wet/dry
lower portion (t/m3) (t/m3)
(m) measured from field dry modulus(1) wet modulus(2) modulus
(months)
2.5% mix 119 Upper 25 2.320 2.306 17 890 20 640 18 200 0.88
Middle 25 2.261 2.214 14 690 12 440 10 870 0.87
129 Upper 25 2.329 2.332 16 460 16 840 18 110 1.08
Lower 25 2.270 2.219 13 020 15 300 11 380 0.74
Upper 2.32 2.32 17 180 18 740 18 155 0.98
Mean
Lower 2.26 2.22 13 865 13 870 11 125 0.81
3% mix 149 Upper 25 2.289 2.261 12 980 13 100 12 030 0.92
Lower 25 2.245 2.163 11 150 10 720 8 040 0.75
249 Upper 25 2.292 2.245 15 310 14 460 15 390 1.06
Lower 25 2.232 2.146 11 290 11 590 8 800 0.76
Upper 2.29 2.25 14 145 13 780 13 710 0.99
Mean
Lower 2.24 2.15 11 220 11 155 8 420 0.75
2 After drying for three days at 40 °C, cores were soaked in water for 10 minutes under vacuum of 95 kPa then tested for modulus at 25 °C.

Roughness and Rutting
About three weeks after opening to traffic, rutting and roughness measurements were taken with a laser
profiler. These measurements were repeated after one and two years of trafficking.
The rutting results, shown in Figure F 41, indicate generally low amounts of rutting after 1 year of trafficking.
Most likely the areas with rut depths in excess of 5 mm have not developed due to trafficking but may relate
to the constructed shape. As discussed in Appendix F.6.5, before sealing deviations up to 15 mm from the
straight edge were measured before sealing.
After 2 years trafficking, high rut depths were measured in areas with severe block cracking (Figure F 47,
Figure F 48). Away from these cracked areas there was minimal increase in rutting despite the inclusion of
asphalt patching material in the material stabilised.
Initially the ride quality of the trial section was good, with an average lane International Roughness Index
(IRI) of 1.4 m/km. As shown in Figure F 42, the roughness peaked at the joint between the two trial mixes. If
this measurement near the joint is discounted, the average roughness reduces to a low value of 1.3 m/km.
Figure F 41: Rutting three weeks, 1 and 2 years after opening to traffic
20
3 weeks Outer wheel path
18
1 year Outer wheel path
16 2 years outer wheel path
14
2.5% FBS 3.0% FBS
12
Rut
depth 10
(mm)
8
0
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (m)

20
3 weeks Inner wheel path
18
1 year Inner wheel path
16 2 years inner wheel path
14 2.5% FBS 3.0% FBS
12
Rut
depth 10
(mm)
8
0
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (m)
Figure F 42: Roughness three weeks, 1 and 2 years after opening to traffic
6.0
3 weeks asphalt patch and
5.5 block cracking
1 year
5.0 2 years
4.5
4.0
3.5
Lane IRI
3.0
(m/km)
2.5
2.0
1.5
1.0
0.5
2.5% FBS 3% FBS
0.0
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Chainage (m)
The cracking of the 2.5% mix between chainages 109 m to 114 m was so severe that in December 2016, the
FBS mix was removed and replaced with an asphalt patch. Due to this patch and the nearby severe block
cracking (Figure F 47), in March 2017 high roughness values were measured between chainages 110 m
to120 m.

Visual Inspection
Condition after 1 year of trafficking
In March 2016 the trial sections were inspected; Figure F 43 and Figure F 44 record the observed distress.
The most apparent distress is the loss of the surface texture of the sprayed bitumen seal over a significant
portion of the trial areas (Figure F 45). The loss of texture is a result of either flushing of the bituminous seal
binder or embedment of the sealing aggregate into the FBS layer. The seal included a nominal 10 parts of
cutter oil (Appendix F.6.6), compared to TMR’s general approach of avoiding cutter oil where possible.
TMR’s experience is that there is a risk that the cutter oil will penetrate and soften the bitumen in the upper
10 mm of the foamed bitumen base. The softening of the FBS layer increases the risk of embedment of the
sealing aggregates, causing flushing/loss of texture and reduced skid resistance. TMR advise that in
Queensland the commonly applied seal surface over an FBS base is now:
• an initial armour-coat seal, comprising about a binder application of 0.6 l/m2 of C170 or C320 (without
cutter oil) and 7 mm cover aggregate spread at 200 m2/m3
• final seal, a 14/7 mm double/double seal using polymer modified binder S0.7B.
In terms of cracking after 1 year of trafficking, surface cracking was limited to a 1–2 m length of the outer
wheel path of the 2.5 % mix, near chainage 112 m as shown in Figure F 46.

Figure F 43: Visual condition after one year of trafficking: 2.5% bitumen mix
Foamed Bitumen Stabilisation Trial Inspection : 2 March 2016
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1
2
3
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
1
2
3
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
1
2
3
75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
1
2
3
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125
1
2
3
125 126 127 128 129 130 131 132 133 134 135 136 137 138
1
2
3
Crocodile cracking Longitudinal cracking Transverse cracking Block cracking
Binder flushing Core holes March 2016

Figure F 44: Visual condition after one year of trafficking: 3% bitumen mix
3% bitumen mix
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163
1
2
3
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
1
2
3
188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213
1
2
3
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
1
2
3
238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263
1
2
3
263 264 265 266

1
2
3
Binder flushing Core holes March 2016

Figure F 45: Example of flushing/embedment of the sprayed seal surface after one year of trafficking
Figure F 46: Fatigue cracking of 2.5% bitumen mix over 1–2 m length

Condition after two years of trafficking
In March 2017 after two years of trafficking, the severity and extent of surface cracking had increased
markedly (Figure F 47, Figure F 48). Note that the cracking of the 2.5% and 3% mixes are reasonably similar
after two years, whereas the 3% mix was predicted (Appendix F.5) to have about three times the fatigue life
of the 2.5% mix.
The cracking of the 2.5% mix between chainages 109 to 114 m was so severe that in December 2016, the
FBS mix was removed and replaced with an asphalt patch.
Figure F 47: Visual condition after two years of trafficking: 2.5% bitumen mix
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
1
2
3
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
1
2
3
50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
1
2
3
75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
1
2
3
100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125
1
2 Asphalt patch
3
125 126 127 128 129 130 131 132 133 134 135 136 137 138
1
2
3
Binder flushing/embedment cores March 2017

Figure F 48: Visual condition after two years of trafficking: 3% bitumen mix
3% bitumen mix
138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163
1
2
3
163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188
1
2
3
188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213
1
2
3
213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238
1
2
3
238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263
1
2
3
263 264 265 266

1
2
3
Binder flushing/embedment cores March 2017
Figure F 49 shows examples of the cracking observed after two years trafficking. In general, the cracking
appears to commence as wheel path transverse cracking, followed by longitudinal cracking intersecting the
transverse cracks, and soon after forming block cracking.

Figure F 49: Development of cracking
initially transverse cracking
then longitinal cracks between transverse cracks

finally block cracking
At four chainages, the cracked areas in the outer wheel path were cored to investigate the cracking below
the surface. Before coring, a small quantity of food dye was poured on the cracked surface to aid observation
of the cracking path. The cores indicated that the cracking extended the full depth of the FBS layer
(Figure F 50).
Figure F 50: Full-depth cracking

In April 2017 more than 50% of the area in both the 2.5% and 3% bitumen sections were cracked to the
extent that the FBS was removed and replaced with 200 mm thick asphalt inlays.
F.10 Comparison of Observed and Predicted Performance

A key objective of the project was to improve the Austroads interim thickness design procedure for FBS
pavements (Austroads 2011a).
As stated in Appendix F.5, the allowable traffic loadings predicted using this method were 1.1 x 106 ESA and
3.3 x 106 ESA at 50% design reliability for 2.5% bitumen and 3% bitumen mixes, respectively. As discussed
in Appendix F.8, the estimated annual traffic loading was about 1.8 x 106 ESA. Consequently, the design
model predicted there was a 50% chance of observing FBS fatigue cracking after 1 year (early 2016) for the
2.5% mix and after 2 years trafficking (early 2017) for the 3% mix.
As mentioned in Appendix F.9.5, after one year of trafficking (March 2016) there was only a minor amount of
cracking of the 2.5% bitumen mix, but cracking was observed over about 20% of the length after a further
year of traffic loading.
For the 3% mix, there was no cracking at 1 year, but cracking was observed over about 20% of the length
after 2 years. The predicted difference in fatigue life of the 2.5% and 3% mixes was not observed.
F.11 Western Freeway Trial Summary

In March 2015, a trial section of 2.5% and 3% FBS mixes was constructed on the Western Freeway, Ballan,
Victoria. Based on preliminary testing to date, the findings are:
• The trial sections were constructed to a high standard of compaction, about a mean 103% modified
Proctor MDD. Unlike the previous Calder Freeway trial (Appendix D), there was minimal reduction in
density with depth, which was attributed to the compaction plant used.
• The mean field moisture content (4.6%) was well below the value used in mix design testing,
5.6% corresponding to 80% modified Proctor OMC.
• The average depth of stabilisation was about 185 mm for the 2.5% bitumen mix section and 205 mm for
the 3% bitumen mix compared to the 200 mm design thickness.
• The measured bitumen contents of samples taken during construction were variable, which most likely
was due to the inclusion of asphalt patch millings and old sprayed seals in the FBS mixes. There appears
to be potential to improve the construction process to reduce this variation.
• The dry densities of the mix design specimens were well below the mean field densities, despite the
higher moisture content used.
• Early-life FBS cores taken between wheel paths and tested three months after construction had mean
indirect tensile moduli in excess of 5000 MPa at a temperature of 25 °C. In addition, mean FBS moduli at
a temperature of 20 °C back-calculated from surface deflections measured one month after construction,
were in excess of 3000 MPa.
• After 1 year of trafficking, the FBS moduli between the wheel paths had increased as indicated from
extraction and testing of field cores and deflection measurements. The indirect tensile moduli of these
field cores were generally very high, in excess of 10 000 MPa. By comparison, the mix design cured dry
moduli were around 4000 MPa.
• During hotter summer periods, the sprayed seal surface had lost texture over a significant portion of the
trial length. The use of cutter oil in the bitumen and size 14 mm sealing aggregate may have led to
embedment of the sealing aggregate into the FBS base.

• After 1 year of trafficking, fatigue cracking was observed over only a 1–2 m length of the outer wheel path
of the 2.5% bitumen mix. However, after 2 years trafficking there was an extensive amount of cracking in
both mixes. The percentage of the area cracked in the two mixes was similar contrary to the fatigue-life
predictions.
• In general, the cracking progresses as follows: initially transverse wheel path cracking occurs, longitudinal
cracks then form linking the transverse cracks with further loading block cracks being formed.
• Due to the extent and rate of development of the cracking, it seems likely that these Austroads sections
will need to be rehabilitated in 2017. If this is the case, it is proposed that monitoring of the adjacent
thicker FBS layer constructed for VicRoads be monitored in 2017 under Austroads project TT2046
Improving the Cost-effectiveness of Foamed Bitumen Stabilised Pavements.

Appendix G Moduli Back-calculated from

Measured FWD Deflections
G.1 Calder Freeway Trial Section
Foamed Bitumen Stabilisation Section
Back-calculation of layer moduli from measured FWD surface deflections requires information about the
pavement layer types and their thicknesses. Based on the FBS cores extracted to date from the outer edge
line and between wheel paths (Appendix D.11), it is assumed the FBS thickness is 200 mm. If the FBS
thickness is 200 mm thick, there is nominal 300 mm thickness of granular material under the FBS layer.
Under the granular material, there is a nominal 300 mm thickness of capping (select subgrade material).
Based on this information, the pavement composition given in Table G 1 was used for all back-calculations.
Table G 1: Calder Freeway assumed FBS pavement composition
Layer Material Thickness (mm)

1 Foamed bitumen stabilised crushed rock 200
2 Crushed rock 150
3 Crushed rock 150
4 Selected subgrade material (capping layer) 300
5 Subgrade 500
6 Lower subgrade Semi-infinite
The moduli of the pavement and subgrade layers at eight days, one month, one, two and three years after
construction were back-calculated using the computer program EFROMD3. The results are given in
Table G 2 to Table G 16.

Table G 2: Calder Freeway back-calculated moduli along the lane edge line eight days after construction
Selected subgrade
FBS Granular Granular Subgrade Lower subgrade FBS FBS
Chainage (capping layer)
temperature modulus
(m) Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus (°C) at 24 °C
(mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa)
15 200 1691 150 261 150 97 300 61 500 120 Semi-infinite 186 29 1916
25 200 1725 150 332 150 110 300 55 500 122 Semi-infinite 198 29 1955
35 200 1934 150 347 150 114 300 53 500 123 Semi-infinite 204 29 2192
45 200 1481 150 422 150 147 300 44 500 118 Semi-infinite 200 29 1678
55 200 1637 150 368 150 118 300 45 500 116 Semi-infinite 196 29 1855
65 200 1228 150 357 150 130 300 38 500 106 Semi-infinite 183 29 1392
75 200 1203 150 499 150 130 300 40 500 107 Semi-infinite 186 29 1363
85 200 2259 150 282 150 81 300 56 500 113 Semi-infinite 182 29 2560
95 200 2318 150 385 150 117 300 44 500 105 Semi-infinite 178 29 2627
Mean 1720 360 115 50 115 190 1950
Table G 3: Calder Freeway back-calculated moduli outer wheel path eight days after construction
Selected subgrade
temperature modulus
15 200 1677 150 211 150 106 300 49 500 108 Semi-infinite 174 28 1853
25 200 1976 150 253 150 75 300 45 500 116 Semi-infinite 194 28 2184
35 200 1737 150 296 150 81 300 45 500 111 Semi-infinite 186 28 1920
45 200 1144 150 655 150 52 300 40 500 118 Semi-infinite 189 28 1264
55 200 1602 150 273 150 74 300 42 500 108 Semi-infinite 182 28 1770
65 200 1783 150 203 150 71 300 37 500 97 Semi-infinite 169 28 1971
75 200 1872 150 187 150 174 300 33 500 103 Semi-infinite 182 28 2069
85 200 1596 150 307 150 92 300 42 500 104 Semi-infinite 174 28 1764
95 200 2378 150 283 150 70 300 41 500 104 Semi-infinite 174 28 2628
Mean 1750 295 90 40 110 180 1935

Table G 4: Calder Freeway back-calculated moduli between wheel paths eight days after construction
Selected subgrade
temperature modulus
(m) Thickness Modulus Thicknes Modulu Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus (°C) at 24 °C
(mm) (MPa) s (mm) s (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa)
15 200 2024 150 378 150 134 300 50 500 110 Semi-infinite 180 29 2293
25 200 1605 150 475 150 118 300 40 500 109 Semi-infinite 191 29 1819
35 200 1762 150 412 150 137 300 40 500 109 Semi-infinite 187 29 1997
45 200 2222 150 221 150 161 300 37 500 104 Semi-infinite 185 29 2518
55 200 1791 150 406 150 137 300 37 500 107 Semi-infinite 187 29 2029
65 200 1894 150 422 150 130 300 35 500 106 Semi-infinite 185 29 2146
75 200 2437 150 383 150 153 300 36 500 112 Semi-infinite 192 29 2761
85 200 2281 150 484 150 138 300 35 500 109 Semi-infinite 190 29 2585
95 200 1822 150 305 150 119 300 44 500 101 Semi-infinite 167 29 2065
Mean 1980 390 135 40 110 185 2245

Table G 5: Calder Freeway back-calculated moduli along the lane edge line one month after construction
Selected subgrade
temperature modulus
15 200 3066 150 337 150 135 300 65 500 131 Semi-infinite 206 13 2329
20 200 4502 150 374 150 86 300 74 500 149 Semi-infinite 231 13 3420
25 200 4109 150 379 150 127 300 66 500 144 Semi-infinite 232 13 3121
30 200 4584 150 177 150 245 300 65 500 138 Semi-infinite 225 13 3482
35 200 3717 150 499 150 140 300 60 500 142 Semi-infinite 238 13 2823
40 200 4073 150 499 150 160 300 50 500 138 Semi-infinite 242 13 3094
45 200 3908 150 490 150 185 300 48 500 136 Semi-infinite 234 13 2968
50 200 4480 150 485 150 173 300 41 500 143 Semi-infinite 257 13 3403
55 200 3536 150 417 150 151 300 49 500 130 Semi-infinite 222 13 2686
60 200 4304 150 457 150 85 300 44 500 133 Semi-infinite 231 13 3269
65 200 3784 150 390 150 136 300 44 500 125 Semi-infinite 215 13 2874
70 200 4202 150 457 150 158 300 48 500 127 Semi-infinite 217 13 3192
75 200 2968 150 431 150 236 300 47 500 124 Semi-infinite 214 13 2254
80 200 4213 150 455 150 160 300 49 500 131 Semi-infinite 221 13 3200
85 200 4353 150 437 150 161 300 47 500 132 Semi-infinite 226 13 3306
90 200 5016 150 447 150 137 300 52 500 131 Semi-infinite 220 13 3810
95 200 3940 150 494 150 176 300 48 500 125 Semi-infinite 209 13 2993
Mean 4040 425 155 55 135 225 3070

Table G 6: Calder Freeway back-calculated moduli outer wheel path one month after construction
Selected subgrade
temperature modulus
15 200 2069 150 412 150 81 300 62 500 122 Semi-infinite 197 11 1495
20 200 3364 150 304 150 123 300 53 500 129 Semi-infinite 220 11 2431
25 200 2866 150 396 150 191 300 47 500 130 Semi-infinite 222 11 2071
30 200 3927 150 195 150 103 300 64 500 133 Semi-infinite 211 11 2837
35 200 3750 150 240 150 136 300 51 500 128 Semi-infinite 216 11 2709
40 200 3515 150 257 150 124 300 49 500 128 Semi-infinite 218 11 2540
45 200 4722 150 152 150 91 300 57 500 129 Semi-infinite 215 11 3412
50 200 3898 150 242 150 107 300 47 500 131 Semi-infinite 230 11 2816
55 200 4343 150 154 150 83 300 49 500 129 Semi-infinite 210 11 3138
60 200 3915 150 150 150 97 300 47 500 117 Semi-infinite 198 11 2829
65 200 3358 150 270 150 83 300 42 500 112 Semi-infinite 190 11 2426
70 200 4151 150 333 150 123 300 38 500 119 Semi-infinite 209 11 2999
75 200 3549 150 154 150 164 300 46 500 109 Semi-infinite 195 11 2564
80 200 3841 150 245 150 108 300 51 500 112 Semi-infinite 194 11 2775
85 200 3549 150 161 150 83 300 62 500 123 Semi-infinite 192 11 2564
90 200 5343 150 300 150 104 300 38 500 124 Semi-infinite 215 11 3860
95 200 3774 150 332 150 140 300 37 500 119 Semi-infinite 209 11 2727
Mean 3760 255 115 50 125 210 2715

Table G 7: Calder Freeway back-calculated moduli between wheel paths one month after construction
Selected subgrade
temperature modulus
15 200 3943 150 436 150 132 300 54 500 128 Semi-infinite 209 12 2921
20 200 4593 150 499 150 104 300 52 500 134 Semi-infinite 226 12 3403
25 200 4248 150 421 150 118 300 46 500 131 Semi-infinite 227 12 3147
30 200 4530 150 372 150 123 300 50 500 121 Semi-infinite 210 12 3356
35 200 5023 150 260 150 85 300 53 500 126 Semi-infinite 209 12 3721
40 200 4830 150 361 150 128 300 36 500 128 Semi-infinite 221 12 3578
45 200 4185 150 370 150 116 300 47 500 123 Semi-infinite 206 12 3100
50 200 4676 150 323 150 101 300 42 500 123 Semi-infinite 214 12 3464
55 200 4290 150 473 150 134 300 39 500 121 Semi-infinite 212 12 3178
60 200 5447 150 150 150 83 300 44 500 122 Semi-infinite 213 12 4035
65 200 5039 150 200 150 81 300 56 500 124 Semi-infinite 205 12 3733
70 200 5101 150 398 150 126 300 45 500 124 Semi-infinite 213 12 3779
75 200 4926 150 498 150 84 300 54 500 125 Semi-infinite 212 12 3649
80 200 4639 150 384 150 127 300 49 500 118 Semi-infinite 207 12 3437
85 200 5031 150 438 150 144 300 39 500 126 Semi-infinite 219 12 3727
90 200 6365 150 357 150 131 300 45 500 124 Semi-infinite 213 12 4715
95 200 4691 150 155 150 84 300 58 500 119 Semi-infinite 194 12 3475
Mean 4795 360 110 50 125 3555

Table G 8: Calder Freeway back-calculated moduli along the lane edge line one year after construction
Selected subgrade
temperature modulus
15 200 2 889 150 495 150 191 300 48 500 120 Semi-infinite 202 25 2 925
20 200 6 070 150 487 150 100 300 56 500 139 Semi-infinite 228 25 6 146
25 200 7 909 150 499 150 81 300 52 500 139 Semi-infinite 243 25 8 008
30 200 5 540 150 499 150 116 300 69 500 138 Semi-infinite 216 25 5 610
35 200 8 718 150 450 150 83 300 44 500 145 Semi-infinite 266 25 8 939
40 200 7 292 150 493 150 80 300 46 500 134 Semi-infinite 249 25 7 477
45 200 7 812 150 499 150 84 300 39 500 141 Semi-infinite 262 25 8 010
50 200 7 693 150 499 150 81 300 43 500 145 Semi-infinite 270 25 7 888
55 200 6 771 150 433 150 102 300 41 500 132 Semi-infinite 242 25 6 942
60 200 3 143 150 500 150 83 300 61 500 124 Semi-infinite 213 25 3 223
65 200 7 123 150 417 150 82 300 37 500 140 Semi-infinite 245 25 7 303
70 200 7 443 150 429 150 80 300 39 500 128 Semi-infinite 239 26 7 727
75 200 6 658 150 499 150 81 300 37 500 132 Semi-infinite 238 26 6 912
80 200 7 586 150 497 150 81 300 34 500 135 Semi-infinite 249 26 7 876
85 200 7 081 150 499 150 81 300 43 500 126 Semi-infinite 225 26 7 352
90 200 7 573 150 320 150 85 300 31 500 139 Semi-infinite 248 26 7 862
95 200 10 737 150 373 150 86 300 20 500 167 Semi-infinite 338 26 11 147
Mean 6 940 465 95 45 135 245 7 130

Table G 9: Calder Freeway back-calculated moduli outer wheel path one year after construction
Selected subgrade
temperature modulus
15 200 1725 150 489 150 92 300 42 500 107 Semi-infinite 179 24 1704
20 200 5260 150 401 150 119 300 45 500 113 Semi-infinite 194 24 5195
25 200 6546 150 166 150 83 300 30 500 124 Semi-infinite 222 24 6465
30 200 3236 150 303 150 81 300 46 500 105 Semi-infinite 172 24 3196
35 200 2952 150 499 150 122 300 29 500 107 Semi-infinite 194 24 2915
40 200 4116 150 186 150 85 300 29 500 102 Semi-infinite 185 24 4065
45 200 3427 150 499 150 92 300 30 500 109 Semi-infinite 191 24 3384
50 200 4123 150 156 150 83 300 28 500 114 Semi-infinite 199 24 4072
55 200 3238 150 150 150 83 300 32 500 115 Semi-infinite 176 24 3198
60 200 3565 150 166 150 85 300 32 500 102 Semi-infinite 188 24 3565
65 200 3346 150 152 150 82 300 35 500 99 Semi-infinite 172 24 3346
70 200 3931 150 190 150 94 300 20 500 88 Semi-infinite 176 24 3931
75 200 3094 150 248 150 80 300 33 500 92 Semi-infinite 161 24 3094
80 200 3722 150 150 150 80 300 29 500 90 Semi-infinite 166 24 3722
85 200 2387 150 235 150 83 300 38 500 91 Semi-infinite 158 24 2387
90 200 2978 150 175 150 240 300 30 500 89 Semi-infinite 161 24 2978
95 200 3182 150 309 150 100 300 27 500 145 Semi-infinite 183 25 3222
Mean 3580 265 100 35 105 180 3555

Table G 10: Calder Freeway back-calculated moduli between wheel paths one year after construction
Selected subgrade
temperature modulus
15 200 5112 150 333 150 80 300 46 500 114 Semi-infinite 205 20 4580
20 200 9210 150 410 150 81 300 37 500 135 Semi-infinite 243 20 8313
25 200 8425 150 404 150 80 300 35 500 130 Semi-infinite 241 20 7642
30 200 7395 150 214 150 89 300 37 500 129 Semi-infinite 223 20 6742
35 200 8359 150 369 150 82 300 32 500 133 Semi-infinite 240 20 7564
40 200 9317 150 174 150 80 300 20 500 164 Semi-infinite 308 20 8367
45 200 3745 150 500 150 89 300 53 500 112 Semi-infinite 190 20 3363
50 200 7198 150 174 150 81 300 20 500 129 Semi-infinite 365 20 6529
55 200 7459 150 444 150 86 300 20 500 155 Semi-infinite 292 20 6749
60 200 6734 150 150 150 80 300 20 500 130 Semi-infinite 344 20 6139
65 200 8154 150 164 150 90 300 20 500 165 Semi-infinite 305 20 7452
70 200 9126 150 336 150 80 300 20 500 145 Semi-infinite 307 20 8341
75 200 8988 150 263 150 81 300 31 500 137 Semi-infinite 252 21 8256
80 200 8233 150 197 150 81 300 20 500 152 Semi-infinite 354 21 7600
85 200 8445 150 236 150 91 300 20 500 164 Semi-infinite 347 21 7835
90 200 9165 150 350 150 81 300 33 500 120 Semi-infinite 229 21 8545
95 200 7293 150 150 150 80 300 28 500 140 Semi-infinite 259 21 6817
Mean 7785 285 85 30 140 275 7100

Table G 11: Calder Freeway back-calculated moduli along the lane edge line two years after construction
Selected subgrade FBS

FBS Granular Granular Subgrade Lower subgrade FBS
Chainage (capping layer) temperature
modulus
(m) Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus during
at 24 °C
(mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) testing (°C)
15 200 5575 150 499 150 84 300 56 500 123 Semi-infinite 208 18 4739
20 200 8753 150 496 150 83 300 54 500 140 Semi-infinite 245 18 7440
25 200 8546 150 469 150 81 300 51 500 134 Semi-infinite 230 18 7264
30 200 5011 150 498 150 96 300 71 500 133 Semi-infinite 201 18 4259
35 200 6915 150 499 150 103 300 56 500 140 Semi-infinite 229 18 5878
40 200 7936 150 489 150 89 300 44 500 135 Semi-infinite 230 18 6831
45 200 8202 150 499 150 91 300 37 500 129 Semi-infinite 230 18 7060
50 200 8146 150 449 150 90 300 33 500 141 Semi-infinite 264 18 7011
55 200 4164 150 500 150 97 300 59 500 123 Semi-infinite 199 18 3584
60 200 2955 150 500 150 80 300 63 500 115 Semi-infinite 189 18 2543
65 200 6287 150 471 150 84 300 39 500 124 Semi-infinite 221 18 5411
70 200 8258 150 435 150 80 300 40 500 119 Semi-infinite 217 18 7108
75 200 3175 150 500 150 83 300 57 500 109 Semi-infinite 177 18 2733
80 200 6375 150 499 150 83 300 37 500 121 Semi-infinite 211 18 5487
85 200 7436 150 494 150 80 300 39 500 119 Semi-infinite 211 18 6400
90 200 8232 150 436 150 80 300 37 500 127 Semi-infinite 225 18 7085
95 200 7596 150 492 150 83 300 49 500 125 Semi-infinite 215 18 6538
Mean 6680 485 85 50 130 220 5730

Table G 12: Calder Freeway back-calculated moduli outer wheel path two years after construction
Selected subgrade
temperature modulus
15 200 1574 150 497 150 91 300 40 500 104 Semi-infinite 171 17 1321
20 200 6114 150 421 150 89 300 29 500 131 Semi-infinite 235 17 5132
25 200 4685 150 310 150 81 300 54 500 120 Semi-infinite 186 18 3982
30 200 3071 150 269 150 82 300 45 500 103 Semi-infinite 160 18 2610
35 200 4617 150 163 150 80 300 33 500 105 Semi-infinite 179 17 3876
40 200 2873 150 223 150 80 300 39 500 95 Semi-infinite 162 17 2412
45 200 2563 150 499 150 90 300 36 500 97 Semi-infinite 167 18 2179
50 200 2481 150 305 150 80 300 34 500 102 Semi-infinite 184 18 2109
55 200 2177 150 493 150 91 300 28 500 101 Semi-infinite 172 17 1827
60 200 2242 150 338 150 80 300 32 500 94 Semi-infinite 167 17 1882
65 200 4135 150 199 150 80 300 21 500 128 Semi-infinite 214 18 3515
70 200 3357 150 469 150 80 300 20 500 87 Semi-infinite 174 18 2854
75 200 1513 150 361 150 80 300 37 500 82 Semi-infinite 135 18 1286
80 200 1888 150 335 150 87 300 20 500 73 Semi-infinite 142 18 1605
85 200 1491 150 368 150 81 300 26 500 86 Semi-infinite 150 18 1267
90 200 2819 150 165 150 80 300 27 500 91 Semi-infinite 147 18 2396
95 200 2739 150 159 150 86 300 32 500 118 Semi-infinite 174 18 2328
Mean 2960 330 85 30 100 170 2510

Table G 13: Calder Freeway back-calculated moduli between wheel paths two years after construction
Selected subgrade
temperature modulus
15 200 3 748 150 500 150 82 300 47 500 111 Semi-infinite 185 17 3 107
20 200 10 874 150 381 150 87 300 20 500 169 Semi-infinite 315 17 9 015
25 200 8 258 150 428 150 80 300 34 500 125 Semi-infinite 227 17 6 846
30 200 4 284 150 500 150 80 300 63 500 111 Semi-infinite 178 17 3 552
35 200 7 645 150 465 150 86 300 32 500 129 Semi-infinite 223 17 6 338
40 200 8 396 150 302 150 80 300 33 500 126 Semi-infinite 228 17 6 961
45 200 3 729 150 499 150 80 300 52 500 110 Semi-infinite 189 17 3 091
50 200 5 733 150 229 150 95 300 20 500 140 Semi-infinite 317 17 4 753
55 200 7 673 150 374 150 80 300 29 500 121 Semi-infinite 222 17 6 361
60 200 7 511 150 154 150 80 300 20 500 100 Semi-infinite 340 17 6 305
65 200 4 364 150 498 150 84 300 41 500 109 Semi-infinite 190 17 3 663
70 200 8 073 150 500 150 120 300 20 500 146 Semi-infinite 286 17 6 777
75 200 4 774 150 486 150 80 300 44 500 114 Semi-infinite 202 17 4 008
80 200 4 595 150 498 150 82 300 35 500 111 Semi-infinite 195 17 3 809
85 200 6 421 150 476 150 114 300 20 500 151 Semi-infinite 298 17 5 323
90 200 11 235 150 438 150 86 300 20 500 139 Semi-infinite 276 17 9 431
95 200 9 582 150 186 150 80 300 21 500 109 Semi-infinite 306 17 7 944
Mean 6 880 405 85 30 125 245 5 720

Table G 14: Calder Freeway back-calculated moduli along the lane edge line three years after construction
Selected subgrade FBS

FBS Granular Granular Subgrade Lower subgrade FBS
Chainage (capping layer) temperature
modulus
(m) Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus during
at 24 °C
(mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) testing (°C)
15 200 4 077 150 500 150 250 300 53 500 145 Semi-infinite 228 16 3 338
20 200 7 233 150 500 150 249 300 53 500 157 Semi-infinite 266 16 5 922
25 200 8 638 150 499 150 249 300 47 500 158 Semi-infinite 275 16 7 072
30 200 11 660 150 334 150 91 300 39 500 178 Semi-infinite 323 16 9 546
35 200 3 846 150 499 150 218 300 77 500 144 Semi-infinite 229 16 3 149
40 200 4 994 150 493 150 216 300 68 500 138 Semi-infinite 235 16 4 089
45 200 8 515 150 493 150 144 300 48 500 142 Semi-infinite 248 16 6 971
50 200 8 119 150 484 150 127 300 47 500 154 Semi-infinite 290 16 6 647
55 200 4 632 150 453 150 201 300 68 500 134 Semi-infinite 238 16 3 792
60 200 3 637 150 500 150 250 300 42 500 151 Semi-infinite 247 16 2 978
65 200 8 641 150 309 150 87 300 45 500 148 Semi-infinite 254 16 7 075
70 200 6 574 150 493 150 178 300 52 500 132 Semi-infinite 225 16 5 382
75 200 4 890 150 500 150 250 300 46 500 137 Semi-infinite 233 16 4 004
80 200 11 242 150 153 150 81 300 36 500 155 Semi-infinite 281 16 9 204
85 200 9 374 150 499 150 83 300 45 500 141 Semi-infinite 246 16 7 675
90 200 8 270 150 486 150 126 300 49 500 135 Semi-infinite 236 16 6 771
95 200 14 539 150 156 150 81 300 20 500 251 Semi-infinite 498 16 11 904
Mean 7 580 430 170 50 155 270 6 200

Table G 15: Calder Freeway back-calculated moduli outer wheel path three years after construction
Selected subgrade
temperature modulus
15 200 2 589 150 344 150 162 300 40 500 110 Semi-infinite 212 16 2 093
20 200 11 264 150 365 150 93 300 21 500 197 Semi-infinite 385 16 9 108
25 200 5 926 150 151 150 80 300 42 500 129 Semi-infinite 254 16 4 792
30 200 4 685 150 188 150 81 300 54 500 123 Semi-infinite 205 16 3 788
35 200 2 939 150 499 150 244 300 34 500 151 Semi-infinite 223 16 2 376
40 200 4 021 150 195 150 81 300 66 500 124 Semi-infinite 203 16 3 251
45 200 5 367 150 235 150 92 300 40 500 120 Semi-infinite 211 16 4 340
50 200 4 279 150 330 150 107 300 40 500 121 Semi-infinite 214 16 3 460
55 200 2 551 150 500 150 156 300 37 500 115 Semi-infinite 210 16 2 063
60 200 3 505 150 304 150 99 300 43 500 118 Semi-infinite 208 16 2 834
65 200 4 985 150 150 150 80 300 33 500 122 Semi-infinite 222 16 4 031
70 200 2 251 150 497 150 237 300 29 500 107 Semi-infinite 167 16 1 820
75 200 2 217 150 187 150 80 300 64 500 105 Semi-infinite 150 16 1 793
80 200 1 868 150 497 150 165 300 29 500 87 Semi-infinite 151 16 1 510
85 200 2 051 150 278 150 80 300 34 500 95 Semi-infinite 165 16 1 658
90 200 3 789 150 223 150 86 300 34 500 96 Semi-infinite 164 16 3 064
95 200 2 082 150 311 150 81 300 48 500 106 Semi-infinite 188 16 1 683
Mean 3 900 310 120 40 120 210 3 160

Table G 16: Calder Freeway back-calculated moduli between wheel paths three years after construction
Selected subgrade
temperature modulus
15 200 6 428 150 395 150 80 300 42 500 140 Semi-infinite 253 17 5 329
20 200 11 574 150 333 150 107 300 38 500 145 Semi-infinite 270 16 9 476
25 200 8 481 150 157 150 80 300 42 500 134 Semi-infinite 252 16 6 944
30 200 3 515 150 498 150 245 300 65 500 114 Semi-infinite 201 16 2 878
35 200 9 247 150 333 150 101 300 37 500 141 Semi-infinite 256 16 7 571
40 200 11 411 150 303 150 93 300 20 500 181 Semi-infinite 344 16 9 343
45 200 4 021 150 498 150 160 300 44 500 121 Semi-infinite 208 16 3 292
50 200 8 711 150 150 150 80 300 20 500 127 Semi-infinite 451 16 7 132
55 200 10 542 150 338 150 81 300 20 500 168 Semi-infinite 355 16 8 524
60 200 7 207 150 157 150 84 300 21 500 105 Semi-infinite 333 16 5 901
65 200 7 175 150 150 150 80 300 34 500 133 Semi-infinite 241 16 5 874
70 200 6 658 150 423 150 85 300 42 500 125 Semi-infinite 219 16 5 383
75 200 5 324 150 500 150 106 300 56 500 118 Semi-infinite 210 16 4 359
80 200 8 226 150 173 150 81 300 20 500 155 Semi-infinite 317 16 6 651
85 200 6 204 150 285 150 80 300 35 500 124 Semi-infinite 237 16 5 016
90 200 13 966 150 195 150 80 300 20 500 193 Semi-infinite 325 16 11 292
95 200 5 961 150 160 150 81 300 34 500 142 Semi-infinite 257 16 4 820
Mean 7 920 300 100 35 140 280 6 460

Asphalt Section
Back-calculation of layer moduli from surface deflections requires information about the pavement material
types and their thicknesses. Based on the measured thickness of asphalt cores extracted (Table D 14), it is
assumed the asphalt thickness is 100 mm for the outer wheel path and 110 mm between wheel paths. Using
the initial pavement composition (Figure D 4), the pavement composition given in Table G 17 was used for
all back-calculations.
Table G 17 Calder Freeway assumed pavement composition of asphalt section

1 Size 20 mm Class 320 dense graded asphalt 100 (OWP)
110 (BWP)
2 Crushed rock 150
3 Crushed rock 250
4 Selected subgrade material (capping layer) 300
5 Subgrade 500
The moduli of the pavement and subgrade layers two weeks and one year after construction were
back-calculated using the computer program EFROMD3. The results are given in Table G 18 to Table G 23.

Table G 18: Calder Freeway back-calculated moduli outer wheel path two weeks after construction
Selected subgrade
Asphalt Granular Granular Subgrade Lower subgrade Asphalt Asphalt
temperature modulus
(m) Thickness Modulus Thickness Modulus Thickness Modulus Thicknes Modulus Thickness Modulus Thickness Modulus (°C) at 24 °C
(mm) (MPa) (mm) (MPa) (mm) (MPa) s (mm) (MPa) (mm) (MPa) (mm) (MPa)
110 100 9 949 150 263 250 80 300 29 500 87 Semi-infinite 163 17 5 638
120 100 13 282 150 265 250 87 300 36 500 106 Semi-infinite 178 17 7 526
130 100 11 642 150 294 250 87 300 32 500 93 Semi-infinite 161 17 6 597
140 100 8 241 150 175 250 81 300 39 500 171 Semi-infinite 214 17 4 670
150 100 12 480 150 319 250 106 300 61 500 105 Semi-infinite 153 17 7 072
160 100 10 434 150 294 250 80 300 41 500 95 Semi-infinite 153 17 5 913
170 100 15 618 150 316 250 110 300 42 500 151 Semi-infinite 262 17 8 850
180 100 9 627 150 158 250 80 300 28 500 124 Semi-infinite 234 17 5 455
190 100 11 347 150 150 250 80 300 21 500 115 Semi-infinite 194 17 6 430
200 100 7 580 150 150 250 80 300 26 500 125 Semi-infinite 224 17 4 295
Mean 11 020 240 90 35 114 195 6 240
Table G 19: Calder Freeway back-calculated moduli between wheel paths two weeks after construction
Selected subgrade
temperature modulus
(m) Thickness Modulus Thickness Modulus Thickness Modulus Thicknes Modulus Thickness Modulus Thickness Modulus (°C) at 24 °C
(mm) (MPa) (mm) (MPa) (mm) (MPa) s (mm) (MPa) (mm) (MPa) (mm) (MPa)
110 110 10 219 150 425 250 114 300 24 500 96 Semi-infinite 175 19 6 581
120 110 12 303 150 448 250 117 300 34 500 103 Semi-infinite 178 19 7 924
130 110 8 642 150 359 250 93 300 37 500 92 Semi-infinite 155 19 5 566
140 110 8 548 150 237 250 87 300 37 500 121 Semi-infinite 236 19 5 505
150 110 10 463 150 365 250 105 300 46 500 100 Semi-infinite 161 19 6 739
160 110 6 477 150 326 250 86 300 34 500 91 Semi-infinite 154 19 4 171
170 110 13 921 150 469 250 237 300 22 500 191 Semi-infinite 404 19 8 966
180 110 7 247 150 308 250 85 300 29 500 139 Semi-infinite 226 19 4 667
190 110 7 728 150 260 250 82 300 30 500 105 Semi-infinite 178 19 4 977
200 110 8 508 150 180 250 81 300 29 500 125 Semi-infinite 219 19 5 479
Mean 9 410 340 110 30 115 210 6 060

Table G 20: Calder Freeway back-calculated moduli outer wheel path one year after construction
Selected subgrade
temperature modulus
105 100 2 945 150 304 250 82 300 34 500 84 Semi-infinite 152 21 2 226
110 100 7 873 150 330 250 81 300 33 500 85 Semi-infinite 154 21 5 950
115 100 11 494 150 197 250 81 300 38 500 89 Semi-infinite 158 21 8 687
120 100 11 986 150 365 250 94 300 33 500 97 Semi-infinite 161 21 9 059
125 100 4 221 150 298 250 83 300 40 500 88 Semi-infinite 146 21 3 190
130 100 10 958 150 369 250 94 300 44 500 96 Semi-infinite 154 21 8 282
135 100 11 013 150 233 250 83 300 44 500 92 Semi-infinite 155 21 8 323
140 100 9 447 150 181 250 83 300 56 500 163 Semi-infinite 241 21 7 140
145 100 6 449 150 361 250 110 300 51 500 99 Semi-infinite 157 21 4 874
150 100 10 193 150 366 250 108 300 52 500 104 Semi-infinite 160 21 7 704
155 100 4 172 150 150 250 82 300 67 500 100 Semi-infinite 149 21 3 153
160 100 3 707 150 151 250 80 300 54 500 112 Semi-infinite 144 21 2 802
165 100 10 232 150 407 250 108 300 47 500 100 Semi-infinite 159 21 8 049
170 100 14 492 150 171 250 82 300 60 500 163 Semi-infinite 264 21 11 400
175 100 4 890 150 150 250 80 300 60 500 149 Semi-infinite 242 21 3 847
180 100 3 816 150 150 250 80 300 37 500 142 Semi-infinite 222 21 3 002
185 100 4 152 150 159 250 84 300 31 500 97 Semi-infinite 171 21 3 266
190 100 3 641 150 150 250 80 300 26 500 80 Semi-infinite 148 21 2 864
195 100 3 596 150 162 250 82 300 28 500 85 Semi-infinite 160 21 2 829
200 110 2 897 150 108 250 80 300 34 500 133 Semi-infinite 203 22 2 372
Mean 7 110 240 85 45 110 175 5 450

Table G 21: Calder Freeway back-calculated moduli between wheel paths one year after construction
Selected subgrade
temperature modulus
105 110 5 871 150 218 250 85 300 30 500 92 Semi-infinite 150 22 4 807
110 110 10 016 150 452 250 83 300 33 500 89 Semi-infinite 150 22 8 200
115 110 9 244 150 388 250 80 300 43 500 90 Semi-infinite 148 22 7 568
120 110 11 342 150 493 250 85 300 42 500 95 Semi-infinite 154 22 9 286
125 110 5 301 150 310 250 82 300 44 500 89 Semi-infinite 150 22 4 340
130 110 9 544 150 498 250 91 300 57 500 99 Semi-infinite 150 22 7 814
135 110 9 494 150 439 250 82 300 41 500 90 Semi-infinite 155 22 7 773
140 110 7 690 150 318 250 80 300 88 500 140 Semi-infinite 234 22 6 296
145 110 5 748 150 500 250 88 300 58 500 105 Semi-infinite 159 22 4 898
150 110 9 552 150 402 250 84 300 62 500 107 Semi-infinite 156 22 8 140
155 110 6 541 150 158 250 83 300 60 500 111 Semi-infinite 153 22 5 574
160 110 4 134 150 341 250 83 300 56 500 95 Semi-infinite 149 22 3 523
165 110 11 398 150 486 250 93 300 62 500 111 Semi-infinite 164 22 9 713
170 110 10 069 150 405 250 90 300 53 500 189 Semi-infinite 281 23 8 930
175 110 8 045 150 209 250 80 300 47 500 163 Semi-infinite 259 23 7 135
180 110 5 123 150 154 250 84 300 32 500 137 Semi-infinite 214 23 4 544
185 110 4 671 150 209 250 85 300 30 500 92 Semi-infinite 168 23 4 143
190 110 4 671 150 213 250 88 300 26 500 90 Semi-infinite 151 23 4 143
195 110 6 083 150 218 250 81 300 25 500 96 Semi-infinite 175 23 5 395
200 110 4 306 150 203 250 83 300 28 500 140 Semi-infinite 199 23 3 975
Mean 7 440 330 85 45 110 175 6 310

Table G 22: Calder Freeway back-calculated moduli outer wheel path two years after construction
Selected subgrade
temperature modulus
105 100 2 947 150 304 250 82 300 34 500 84 Semi-infinite 152 22 2 511
110 100 7 080 150 400 250 81 300 32 500 86 Semi-infinite 154 23 6 279
115 100 11 389 150 208 250 81 300 37 500 89 Semi-infinite 158 23 10 101
120 100 14 832 150 178 250 81 300 41 500 92 Semi-infinite 157 22 12 639
125 100 3 893 150 291 250 95 300 37 500 88 Semi-infinite 147 23 3 453
130 100 9 430 150 456 250 108 300 40 500 95 Semi-infinite 156 23 8 364
135 100 11 042 150 230 250 83 300 45 500 92 Semi-infinite 155 23 9 793
140 100 9 370 150 167 250 86 300 56 500 161 Semi-infinite 242 23 8 310
145 100 5 481 150 428 250 122 300 46 500 100 Semi-infinite 160 23 4 861
150 100 8 578 150 453 250 127 300 45 500 105 Semi-infinite 163 23 7 608
155 100 4 148 150 150 250 81 300 68 500 100 Semi-infinite 148 23 3 679
160 100 3 481 150 151 250 81 300 57 500 100 Semi-infinite 146 23 3 087
165 100 8 896 150 491 250 124 300 41 500 99 Semi-infinite 163 23 7 890
170 100 14 435 150 181 250 84 300 56 500 178 Semi-infinite 264 23 12 803
175 100 4 367 150 156 250 80 300 63 500 130 Semi-infinite 247 23 4 031
180 100 3 727 150 153 250 80 300 34 500 171 Semi-infinite 219 23 3 440
185 100 3 656 150 186 250 89 300 29 500 102 Semi-infinite 173 23 3 375
190 100 3 715 150 150 250 80 300 24 500 95 Semi-infinite 148 23 3 429
195 100 3 082 150 198 250 84 300 27 500 87 Semi-infinite 162 23 2 845
200 100 2 546 150 150 250 80 300 34 500 118 Semi-infinite 207 23 2 350
Mean 6 800 255 90 40 110 175 6 040

Table G 23: Calder Freeway back-calculated moduli between wheel paths two years after construction
Selected subgrade
temperature modulus
105 110 4 116 150 278 250 81 300 32 500 83 Semi-infinite 143 23 3 651
110 110 8 404 150 491 250 90 300 32 500 82 Semi-infinite 140 23 7 454
115 110 8 670 150 489 250 80 300 39 500 87 Semi-infinite 144 22 7 388
120 110 11 620 150 497 250 89 300 38 500 92 Semi-infinite 151 22 9 902
125 110 7 479 150 359 250 82 300 40 500 89 Semi-infinite 147 22 6 373
130 110 8 679 150 482 250 101 300 51 500 94 Semi-infinite 144 22 7 396
135 110 8 606 150 462 250 80 300 37 500 86 Semi-infinite 140 22 7 334
140 110 7 073 150 408 250 80 300 127 500 158 Semi-infinite 195 22 6 027
145 110 8 088 150 496 250 89 300 62 500 105 Semi-infinite 148 22 6 892
150 110 9 266 150 495 250 84 300 60 500 96 Semi-infinite 146 22 7 586
155 110 3 981 150 305 250 83 300 52 500 97 Semi-infinite 153 22 3 259
160 110 3 775 150 448 250 83 300 48 500 88 Semi-infinite 140 22 3 091
165 110 11 594 150 498 250 90 300 57 500 106 Semi-infinite 156 22 9 880
170 110 8 973 150 413 250 193 300 27 500 199 Semi-infinite 346 22 7 646
175 110 6 905 150 194 250 80 300 49 500 197 Semi-infinite 251 22 5 653
180 110 3 634 150 160 250 83 300 23 500 169 Semi-infinite 197 22 3 097
185 110 2 979 150 229 250 80 300 27 500 98 Semi-infinite 153 22 2 539
190 110 3 428 150 185 250 82 300 22 500 74 Semi-infinite 138 22 2 921
195 110 5 109 150 212 250 81 300 20 500 71 Semi-infinite 162 22 4 354
200 110 3 380 150 151 250 80 300 24 500 119 Semi-infinite 191 22 2 880
Mean 6 790 365 90 45 110 170 5 770

G.2 Newell Highway Trial Section

pavement layer types and their thicknesses. Based on the FBS cores extracted to date from the outer edge
line and between wheel paths (Appendix E.6.3), it is assumed the FBS thickness is 165 mm. It is intended to
confirm this by additional coring in the wheel paths. If the FBS is 165 mm thick, there is nominal 360 mm
thickness of granular material under the FBS layer. Under the granular material, there is a nominal 870 mm
thickness of selected subgrade material. Based on this information, the pavement composition given in
Table G 24 was used for all back-calculations.
Table G 24: Newell Highway assumed FBS pavement composition

1 Foamed bitumen stabilised crushed rock 165
2 Crushed rock 180
3 Crushed rock 180
4 Selected subgrade material 435
5 Selected subgrade material 435
6 Upper subgrade 400
The moduli of the pavement and subgrade layers at two and 18 months after construction were
back-calculated using the computer program EFROMD3. The results are given in Table G 25 to Table G 30.

Table G 25: Newell Highway back-calculated moduli outer wheel path two months after construction
Selected subgrade Selected subgrade

FBS Granular Granular Upper subgrade Subgrade FBS
Chainage (capping layer) (capping layer)
modulus
(m) Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Thickness Modulus Modulus at 31 °C
(mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (mm) (MPa) (MPa)
5 165 3601 180 2017 180 944 435 892 435 561 400 268 262 3739
10 165 4383 180 1288 180 427 435 726 435 194 400 289 277 4550
15 165 4148 180 1078 180 1000 435 533 435 578 400 206 212 4307
20 165 8657 180 720 180 999 435 999 435 209 400 249 224 8988
25 165 6315 180 1997 180 985 435 682 435 293 400 196 196 6556
30 165 3248 180 959 180 524 435 998 435 209 400 234 221 3372
35 165 3046 180 1672 180 987 435 998 435 353 400 269 257 3123
40 165 2735 180 1668 180 603 435 1000 435 124 400 273 273 2804
45 165 4190 180 1053 180 1000 435 753 435 171 400 283 274 4296
50 165 3591 180 2073 180 687 435 689 435 599 400 234 221 3682
55 165 5460 180 1117 180 861 435 898 435 361 400 256 243 5598
60 165 2881 180 669 180 499 435 836 435 593 400 271 209 2954
65 165 5405 180 958 180 650 435 976 435 490 400 251 234 5612
70 165 2536 180 888 180 979 435 964 435 599 400 294 241 2633
75 165 3345 180 1477 180 893 435 1000 435 600 400 285 272 3473
80 165 4512 180 1241 180 1000 435 1000 435 591 400 330 297 4684
85 165 2195 180 1001 180 601 435 896 435 331 400 252 247 2251
90 165 3788 180 1012 180 454 435 418 435 600 400 164 162 3884
Mean 4110 1270 780 850 410 260 240 4250

Table G 26: Newell Highway back-calculated moduli between wheel paths two months after construction

modulus
10 165 3830 180 1522 180 639 435 855 435 189 400 258 260 3976
20 165 2572 180 3041 180 733 435 476 435 144 400 230 229 2670
30 165 4459 180 1270 180 425 435 421 435 249 400 201 197 4629
40 165 6051 180 1325 180 818 435 567 435 178 400 240 245 6204
50 165 4081 180 2206 180 680 435 726 435 235 400 245 246 4184
60 165 4861 180 715 180 1141 435 511 435 472 400 200 200 4984
70 165 3897 180 786 180 701 435 979 435 357 400 309 240 3996
80 165 4494 180 1899 180 561 435 917 435 203 400 283 280 4608
90 165 3682 180 861 180 269 435 300 435 174 400 246 219 3775
Mean 4210 1510 660 640 245 245 235 4340
Table G 27: Newell Highway back-calculated moduli between edge line two months after construction

modulus
10 165 11 697 180 1 665 180 401 435 1 000 435 1 000 400 208 190 12 144
20 165 12 000 180 2 456 180 2 925 435 163 435 993 400 414 313 12 459
30 165 3 039 180 3 632 180 845 435 580 435 491 400 260 200 3 155
40 165 8 606 180 1 078 180 425 435 704 435 470 400 211 209 8 935
50 165 3 974 180 2 441 180 416 435 992 435 760 400 233 230 4 126
60 165 2 362 180 1 944 180 503 435 934 435 985 400 283 223 2 452
70 165 5 013 180 3 082 180 934 435 975 435 998 400 259 236 5 205
80 165 3 702 180 8 395 180 1 076 435 933 435 979 400 386 310 3 843
90 165 3 015 180 1 659 180 764 435 998 435 562 400 206 194 3 130
Mean 5 930 2 930 921 810 800 275 235 6 160

Table G 28: Newell Highway back-calculated moduli outer wheel path 18 months after construction

modulus
10 165 9 807 180 609 180 999 435 999 435 102 400 282 293 11 415
15 165 8 844 180 1 064 180 503 435 579 435 287 400 263 260 10 552
20 165 7 091 180 3 161 180 612 435 583 435 338 400 231 231 8 260
25 165 9 605 180 1 745 180 563 435 673 435 381 400 194 194 11 415
30 165 4 939 180 1 547 180 433 435 900 435 107 400 271 280 11 558
35 165 10 524 180 2 641 180 998 435 998 435 450 400 254 254 10 786
40 165 5 995 180 2 494 180 863 435 671 435 271 400 242 240 6 359
45 165 10 423 180 1 539 180 1 000 435 1 000 435 233 400 253 254 9 064
50 165 8 917 180 2 464 180 1 000 435 660 435 470 400 254 254 7 963
55 165 9 544 180 1 815 180 931 435 995 435 315 400 251 250 4 396
60 165 8 803 180 1 460 180 972 435 820 435 335 400 256 255 7 820
65 165 6 281 180 1 798 180 999 435 791 435 607 400 246 246 5 392
70 165 5 685 180 1 144 180 810 435 1 000 435 521 400 275 257 4 505
75 165 7 335 180 1 261 180 978 435 988 435 559 400 287 282 3 862
80 165 9 325 180 1 609 180 993 435 1 000 435 803 400 300 281 4 283
85 165 6 163 180 1 302 180 670 435 938 435 382 400 240 240 7 701
90 165 12 000 180 465 180 424 435 660 435 230 400 207 183 1 934
Mean 8 310 1650 810 840 380 250 250 8200

Table G 29: Newell Highway back-calculated moduli between wheel paths 18 months after construction

modulus
10 165 5 886 180 1 598 180 357 435 591 435 168 400 263 222 5 886
15 165 5 549 180 1 895 180 220 435 395 435 271 400 252 236 5 549
20 165 9 510 180 1 527 180 237 435 162 435 351 400 213 193 9 510
25 165 10 715 180 610 180 321 435 186 435 172 400 175 175 10 715
30 165 11 985 180 1 665 180 476 435 370 435 143 400 228 215 11 985
35 165 11 152 180 729 180 556 435 999 435 122 400 264 261 11 152
40 165 6 140 180 4 337 180 262 435 371 435 825 400 223 245 6 140
45 165 7 814 180 773 180 1 359 435 660 435 937 400 198 205 7 814
50 165 6 511 180 2 104 180 706 435 729 435 419 400 236 236 6 511
55 165 5 499 180 2 109 180 436 435 269 435 876 400 204 207 5 499
60 165 6 470 180 1 745 180 693 435 395 435 362 400 202 203 6 551
65 165 6 052 180 2 739 180 624 435 525 435 672 400 211 210 6 128
70 165 5 353 180 1 247 180 450 435 292 435 1 000 400 278 239 5 420
75 165 11 338 180 859 180 322 435 879 435 636 400 267 255 11 481
80 165 3 761 180 2 785 180 556 435 527 435 219 400 266 265 3 808
85 165 6 633 180 1 350 180 379 435 344 435 310 400 209 209 6 716
90 165 2 533 180 997 180 97 435 314 435 587 400 164 164 2 565
Mean 7 230 1 710 475 470 475 230 220 7 260

Table G 30: Newell Highway back-calculated moduli between edge line 18 months after construction

modulus
10 165 12 000 180 3 584 180 974 435 553 435 293 400 245 265 11 415
15 165 11 093 180 1 022 180 1 244 435 508 435 464 400 258 240 10 552
20 165 8 684 180 2 879 180 1 504 435 996 435 518 400 271 268 8 260
25 165 12 000 180 4 885 180 1 111 435 681 435 489 400 210 214 11 415
30 165 12 000 180 3 167 180 1 278 435 891 435 300 400 215 217 11 558
35 165 11 198 180 4 182 180 1 795 435 1 000 435 606 400 254 254 10 786
40 165 6 602 180 1 634 180 758 435 656 435 357 400 258 254 6 359
45 165 9 410 180 1 781 180 326 435 655 435 604 400 296 236 9 064
50 165 8 267 180 774 180 1 008 435 786 435 527 400 248 243 7 963
55 165 4 507 180 1 856 180 1 697 435 991 435 688 400 329 266 4 396
60 165 8 018 180 1 499 180 865 435 975 435 962 400 258 239 7 820
65 165 5 528 180 3 272 180 1 687 435 996 435 576 400 270 257 5 392
70 165 4 619 180 2 763 180 2 489 435 739 435 997 400 320 279 4 505
75 165 3 960 180 1 832 180 594 435 905 435 582 400 347 323 3 862
80 165 4 391 180 6 651 180 1 261 435 997 435 422 400 393 369 4 283
85 165 7 896 180 836 180 642 435 1 000 435 835 400 251 250 7 701
90 165 1 983 180 2 402 180 533 435 1 000 435 517 400 249 206 1 934
Mean 7 770 2 650 1 160 840 570 275 260 7 490

G.3 Western Freeway, Ballan Trial Sections

pavement layer types and their thicknesses. Based on the FBS cores extracted to date between the wheel
paths (Table F 17), it is assumed the FBS thickness is 185 mm and 205 mm for the 2.5% bitumen and
3.0% bitumen mixes respectively. It is intended to confirm this in the future by additional coring in the wheel
paths. If the FBS are 185 mm and 205 mm, there are nominal granular thicknesses of 315 mm and 295 mm
under the FBS layer. Under the granular material, there is a nominal 300 mm thickness of selected subgrade
material. Based on this information, the pavement composition given in Table G 31 was used for all
back-calculations.
Table G 31: Western Freeway assumed FBS pavement composition
Thickness (mm)
Layer Material
2.5% bitumen 3% bitumen
1 Foamed bitumen stabilised crushed rock 185 205
2 Crushed rock 160 150
3 Crushed rock 155 145
4 Selected subgrade material 300 300
5 Upper subgrade 500 500
6 Lower subgrade Semi-infinite Semi-infinite
The moduli of the pavement and subgrade layers one month after construction were back-calculated using
the computer program EFROMD3. The results are given in Table G 32 and Table G 37.

Table G 32: Western Freeway back-calculated moduli of 2.5% bitumen mix one month after construction
Selected subgrade
temperature modulus
Outer wheel path
5 2503 160 234 155 96 300 87 500 254 2503 Semi-infinite 306 19 2411
10 1986 160 199 155 98 300 75 500 197 1986 Semi-infinite 272 19 1913
15 2910 160 251 155 88 300 64 500 200 2910 Semi-infinite 271 19 2803
20 2097 160 204 155 82 300 74 500 183 2097 Semi-infinite 262 19 2020
25 1791 160 173 155 86 300 75 500 235 1791 Semi-infinite 267 19 1725
30 1874 160 192 155 90 300 91 500 196 1874 Semi-infinite 266 19 1805
35 1988 160 178 155 93 300 92 500 196 1988 Semi-infinite 255 19 1915
40 2248 160 225 155 100 300 131 500 211 2248 Semi-infinite 253 18 2138
45 2402 160 249 155 97 300 170 500 217 2402 Semi-infinite 233 18 2285
50 2589 160 268 155 93 300 149 500 215 2589 Semi-infinite 251 18 2463
55 2166 160 210 155 104 300 196 500 495 2166 Semi-infinite 333 18 2060
60 1802 160 178 155 91 300 197 500 356 1802 Semi-infinite 329 18 1714
65 2133 160 194 155 91 300 183 500 312 2133 Semi-infinite 316 18 2029
70 2622 160 222 155 95 300 179 500 319 2622 Semi-infinite 309 18 2494
75 3321 160 239 155 127 300 196 500 308 3321 Semi-infinite 307 18 3159
80 2779 160 244 155 101 300 185 500 297 2779 Semi-infinite 291 18 2643
85 2412 160 194 155 81 300 200 500 231 2412 Semi-infinite 276 18 2294
90 1977 160 181 155 142 300 101 500 379 1977 Semi-infinite 282 18 1881
95 1899 160 164 155 87 300 152 500 312 1899 Semi-infinite 318 18 1806
100 285 1666 160 196 155 84 300 199 500 307 Semi-infinite 327 18 1585
105 285 1830 160 159 155 90 300 198 500 365 Semi-infinite 306 18 1741
110 285 2267 160 189 155 97 300 180 500 317 Semi-infinite 324 18 2156
115 285 1873 160 152 155 86 300 196 500 327 Semi-infinite 339 18 1782
120 285 3091 160 257 155 114 300 158 500 342 Semi-infinite 329 18 2940
125 285 2953 160 294 155 105 300 194 500 320 Semi-infinite 347 18 2809
130 285 2841 160 220 155 112 300 196 500 340 Semi-infinite 324 18 2702
135 285 2402 160 217 155 101 300 186 500 323 Semi-infinite 291 18 2285
Mean 2310 210 95 150 290 295 2205

Selected subgrade
temperature modulus
Between wheel paths
5 185 4713 160 335 155 121 300 102 500 241 Semi-infinite 315 19 4540
10 185 3470 160 272 155 92 300 86 500 201 Semi-infinite 259 19 3342
15 185 2880 160 223 155 87 300 83 500 184 Semi-infinite 253 19 2774
20 185 2181 160 174 155 90 300 78 500 187 Semi-infinite 284 18 2075
25 185 1978 160 181 155 80 300 98 500 204 Semi-infinite 267 19 1905
30 185 2482 160 212 155 95 300 134 500 210 Semi-infinite 265 19 2391
35 185 2796 160 273 155 101 300 194 500 242 Semi-infinite 266 19 2693
40 185 3103 160 323 155 117 300 197 500 248 Semi-infinite 255 19 2989
45 185 3135 160 273 155 135 300 198 500 252 Semi-infinite 236 19 3020
50 185 3041 160 281 155 108 300 150 500 209 Semi-infinite 228 19 2929
55 185 2426 160 298 155 105 300 198 500 347 Semi-infinite 308 19 2337
60 185 2623 160 165 155 136 300 117 500 368 Semi-infinite 322 19 2526
65 185 2825 160 238 155 101 300 124 500 304 Semi-infinite 307 19 2721
70 185 2695 160 181 155 89 300 121 500 241 Semi-infinite 287 19 2596
75 185 2556 160 281 155 92 300 90 500 262 Semi-infinite 306 19 2462
80 185 2306 160 196 155 83 300 102 500 259 Semi-infinite 304 19 2221
85 185 2465 160 183 155 82 300 197 500 303 Semi-infinite 300 19 2374
90 185 2414 160 215 155 91 300 153 500 280 Semi-infinite 287 19 2325
95 185 2240 160 169 155 93 300 154 500 300 Semi-infinite 317 19 2158
100 185 2160 160 182 155 82 300 176 500 297 Semi-infinite 320 19 2080
105 185 2564 160 168 155 93 300 145 500 295 Semi-infinite 319 19 2470
110 185 3143 160 215 155 152 300 127 500 309 Semi-infinite 318 19 3027
115 185 2883 160 166 155 137 300 120 500 368 Semi-infinite 325 19 2777
120 185 3069 160 218 155 108 300 156 500 342 Semi-infinite 343 19 2956
125 185 3778 160 308 155 123 300 175 500 326 Semi-infinite 336 19 3639
130 185 3853 160 313 155 124 300 197 500 350 Semi-infinite 313 19 3711
135 185 2509 160 209 155 97 300 197 500 296 Semi-infinite 299 19 2417
Mean 2825 230 155 105 300 145 500 275 295 2720

Table G 33: Western Freeway back-calculated moduli of 3% bitumen mix one month after construction
Selected subgrade
temperature modulus
Outer wheel path
140 205 2124 160 298 165 95 300 194 500 499 Semi-infinite 272 19 2046
145 205 1505 160 185 165 112 300 194 500 492 Semi-infinite 330 19 1450
150 205 2053 160 211 165 102 300 195 500 496 Semi-infinite 346 19 1977
155 205 1669 160 157 165 84 300 191 500 383 Semi-infinite 347 19 1608
160 205 1855 160 151 165 131 300 96 500 492 Semi-infinite 324 19 1787
165 205 1702 160 163 165 82 300 130 500 291 Semi-infinite 300 19 1639
170 205 1767 160 177 165 81 300 141 500 223 Semi-infinite 282 19 1702
175 205 2015 160 185 165 88 300 142 500 264 Semi-infinite 271 18 1917
180 205 1826 160 161 165 87 300 113 500 268 Semi-infinite 311 18 1737
185 205 1742 160 158 165 85 300 147 500 294 Semi-infinite 319 18 1657
190 205 1728 160 158 165 81 300 163 500 285 Semi-infinite 329 18 1644
195 205 1522 160 191 165 86 300 144 500 279 Semi-infinite 331 18 1448
200 205 1467 160 151 165 83 300 139 500 281 Semi-infinite 329 18 1395
205 205 1426 160 150 165 81 300 128 500 261 Semi-infinite 330 18 1356
210 205 1370 160 150 165 80 300 117 500 269 Semi-infinite 324 18 1303
215 205 1840 160 173 165 84 300 103 500 256 Semi-infinite 316 18 1750
220 205 1579 160 180 165 90 300 98 500 306 Semi-infinite 336 18 1502
225 205 1910 160 166 165 120 300 97 500 411 Semi-infinite 364 18 1817
230 205 1794 160 161 165 126 300 97 500 419 Semi-infinite 363 18 1707
235 205 2353 160 172 165 121 300 97 500 327 Semi-infinite 368 18 2238
240 205 4319 160 472 165 223 300 144 500 289 Semi-infinite 373 18 4108
245 205 3643 160 296 165 125 300 105 500 299 Semi-infinite 376 18 3465
250 205 1988 160 165 165 133 300 97 500 376 Semi-infinite 324 18 1891
255 205 2671 160 173 165 80 300 108 500 245 Semi-infinite 293 18 2541
260 205 3692 160 353 165 117 300 116 500 254 Semi-infinite 335 18 3512
Mean 2060 200 105 130 330 330 1970

Selected subgrade
temperature modulus
Between wheel paths
140 205 3100 160 266 165 116 300 198 500 373 Semi-infinite 306 19 2986
145 205 2632 160 209 165 114 300 199 500 381 Semi-infinite 338 19 2535
150 205 2661 160 229 165 94 300 198 500 379 Semi-infinite 363 19 2563
155 205 2520 160 227 165 88 300 194 500 384 Semi-infinite 354 18 2397
160 205 2457 160 152 165 98 300 129 500 314 Semi-infinite 306 19 2367
165 205 2277 160 160 165 98 300 137 500 288 Semi-infinite 288 19 2193
170 205 2568 160 194 165 85 300 124 500 213 Semi-infinite 284 19 2473
175 205 2332 160 180 165 82 300 147 500 238 Semi-infinite 276 19 2246
180 205 2018 160 161 165 89 300 100 500 277 Semi-infinite 299 19 1944
185 205 2151 160 154 165 85 300 89 500 284 Semi-infinite 327 19 2072
190 205 1978 160 165 165 89 300 113 500 307 Semi-infinite 323 19 1905
195 205 2615 160 176 165 114 300 116 500 331 Semi-infinite 322 19 2519
200 205 2425 160 177 165 112 300 117 500 371 Semi-infinite 321 19 2336
205 205 2509 160 197 165 123 300 119 500 377 Semi-infinite 323 19 2417
210 205 2666 160 189 165 114 300 118 500 369 Semi-infinite 331 19 2568
215 205 2832 160 218 165 118 300 127 500 334 Semi-infinite 326 19 2728
220 205 2985 160 283 165 88 300 164 500 323 Semi-infinite 352 19 2875
225 205 3193 160 289 165 97 300 148 500 328 Semi-infinite 373 19 3075
230 205 3087 160 247 165 101 300 158 500 379 Semi-infinite 447 19 2973
235 205 3010 160 249 165 86 300 175 500 354 Semi-infinite 387 19 2899
240 205 2767 160 276 165 85 300 196 500 342 Semi-infinite 384 19 2665
245 205 3284 160 262 165 87 300 196 500 366 Semi-infinite 380 19 3163
250 205 2694 160 177 165 107 300 113 500 301 Semi-infinite 337 19 2595
255 205 2927 160 200 165 87 300 105 500 223 Semi-infinite 311 19 2819
260 205 3572 160 294 165 86 300 95 500 239 Semi-infinite 321 19 3441
Mean 2690 215 100 145 325 335 2590

Table G 34: Western Freeway back-calculated moduli of 2.5% bitumen mix one year after construction
Selected subgrade
temperature modulus
Outer wheel path
5 185 2147 160 150 155 86 300 101 500 297 Semi-infinite 276 19 2068
10 185 3517 160 170 155 129 300 108 500 206 Semi-infinite 252 19 3388
15 185 5340 160 268 155 153 300 90 500 193 Semi-infinite 246 19 5143
20 185 3037 160 179 155 130 300 75 500 180 Semi-infinite 246 19 2925
25 185 2360 160 168 155 104 300 97 500 192 Semi-infinite 239 19 2273
30 185 2683 160 159 155 90 300 124 500 197 Semi-infinite 235 19 2584
35 185 3185 160 153 155 96 300 121 500 196 Semi-infinite 234 19 3068
40 185 3916 160 250 155 147 300 124 500 203 Semi-infinite 240 18 3725
45 185 3552 160 234 155 141 300 139 500 192 Semi-infinite 229 18 3379
50 185 3515 160 230 155 136 300 112 500 195 Semi-infinite 237 18 3344
55 185 3507 160 174 155 148 300 137 500 310 Semi-infinite 318 18 3336
60 185 3638 160 183 155 135 300 145 500 322 Semi-infinite 312 18 3461
65 185 4200 160 218 155 200 300 148 500 305 Semi-infinite 309 18 3995
70 185 5162 160 218 155 193 300 138 500 303 Semi-infinite 309 18 4910
75 185 4266 160 291 155 199 300 172 500 274 Semi-infinite 309 18 4058
80 185 3971 160 222 155 164 300 183 500 301 Semi-infinite 279 18 3777
85 185 4221 160 177 155 143 300 178 500 267 Semi-infinite 275 18 4015
90 185 3089 160 198 155 162 300 121 500 347 Semi-infinite 265 18 2938
95 185 2916 160 188 155 81 300 199 500 352 Semi-infinite 289 18 2774
100 185 2727 160 171 155 102 300 198 500 281 Semi-infinite 301 18 2594
105 185 3247 160 150 155 135 300 139 500 302 Semi-infinite 300 18 3089
110 185 3139 160 169 155 138 300 100 500 350 Semi-infinite 267 18 2986
115 185 2159 160 152 155 97 300 143 500 249 Semi-infinite 287 18 2054
120 185 4021 160 189 155 147 300 121 500 247 Semi-infinite 306 18 3825
125 185 5302 160 209 155 147 300 100 500 298 Semi-infinite 313 18 5043
130 185 4915 160 225 155 169 300 150 500 260 Semi-infinite 301 18 4675
135 185 4548 160 234 155 178 300 172 500 265 Semi-infinite 285 18 4326
Mean 3640 195 140 135 260 275 3470

Selected subgrade
temperature modulus
Between wheel paths
5 185 9015 160 308 155 281 300 116 500 224 Semi-infinite 291 19 8683
10 185 9009 160 293 155 218 300 94 500 184 Semi-infinite 252 19 8677
15 185 7727 160 216 155 188 300 65 500 175 Semi-infinite 253 19 7443
20 185 4119 160 150 155 130 300 83 500 165 Semi-infinite 232 18 3918
25 185 3627 160 157 155 130 300 97 500 187 Semi-infinite 234 19 3494
30 185 4287 160 152 155 156 300 134 500 219 Semi-infinite 231 19 4129
35 185 5269 160 236 155 210 300 153 500 220 Semi-infinite 246 19 5075
40 185 6197 160 267 155 232 300 177 500 227 Semi-infinite 243 19 5969
45 185 5785 160 259 155 223 300 198 500 223 Semi-infinite 234 19 5572
50 185 5306 160 258 155 208 300 124 500 204 Semi-infinite 236 19 5111
55 185 5252 160 234 155 208 300 141 500 313 Semi-infinite 309 19 5059
60 185 4869 160 184 155 184 300 139 500 260 Semi-infinite 313 19 4690
65 185 6206 160 257 155 211 300 129 500 282 Semi-infinite 309 19 5978
70 185 5744 160 207 155 183 300 126 500 243 Semi-infinite 309 19 5533
75 185 5551 160 152 155 169 300 118 500 267 Semi-infinite 299 19 5347
80 185 3748 160 165 155 168 300 107 500 272 Semi-infinite 293 19 3610
85 185 5179 160 157 155 186 300 158 500 294 Semi-infinite 291 19 4988
90 185 4824 160 189 155 164 300 127 500 338 Semi-infinite 275 19 4646
95 185 5001 160 167 155 194 300 141 500 286 Semi-infinite 299 19 4817
100 185 5208 160 165 155 133 300 143 500 301 Semi-infinite 305 19 5016
105 185 6860 160 180 155 161 300 117 500 300 Semi-infinite 306 19 6608
110 185 8087 160 151 155 153 300 83 500 197 Semi-infinite 288 19 7789
115 185 5479 160 208 155 199 300 129 500 303 Semi-infinite 280 19 5277
120 185 4794 160 191 155 162 300 112 500 279 Semi-infinite 285 19 4618
125 185 5457 160 197 155 166 300 116 500 264 Semi-infinite 294 19 5256
130 185 7112 160 265 155 242 300 131 500 221 Semi-infinite 286 19 6850
135 185 4793 160 186 155 162 300 117 500 280 Semi-infinite 286 19 4617
Mean 5720 205 185 125 250 275 5510

Table G 35: Western Freeway back-calculated moduli of 3% bitumen mix one year after opening to traffic
Selected subgrade
temperature modulus
Outer wheel path
140 205 2725 155 422 150 101 300 198 500 365 Semi-infinite 286 19 2794
145 205 1465 155 150 150 80 300 127 500 438 Semi-infinite 303 19 1502
150 205 2433 155 233 150 140 300 82 500 468 Semi-infinite 298 19 2495
155 205 1132 155 203 150 80 300 119 500 370 Semi-infinite 291 19 1161
160 205 1339 155 230 150 80 300 65 500 295 Semi-infinite 281 19 1390
165 205 1894 155 289 150 83 300 69 500 279 Semi-infinite 292 19 1966
170 205 2306 155 439 150 85 300 75 500 224 Semi-infinite 268 19 2394
175 205 2420 155 383 150 91 300 105 500 234 Semi-infinite 267 18 2512
180 205 2499 155 212 150 114 300 77 500 372 Semi-infinite 297 18 2594
185 205 2244 155 321 150 96 300 77 500 273 Semi-infinite 296 18 2330
190 205 2306 155 219 150 119 300 77 500 334 Semi-infinite 287 18 2394
195 205 2273 155 207 150 129 300 75 500 314 Semi-infinite 289 18 2360
200 205 1731 155 263 150 83 300 84 500 343 Semi-infinite 295 18 1797
205 205 1316 155 215 150 80 300 103 500 233 Semi-infinite 291 18 1349
210 205 1766 155 164 150 116 300 73 500 466 Semi-infinite 285 18 1811
215 205 1997 155 356 150 85 300 66 500 227 Semi-infinite 286 18 2073
220 205 1931 155 308 150 85 300 66 500 240 Semi-infinite 312 18 2005
225 205 2025 155 285 150 92 300 68 500 288 Semi-infinite 311 18 2102
230 205 1000 155 195 150 85 300 62 500 232 Semi-infinite 299 18 1038
235 205 2280 155 288 150 80 300 77 500 231 Semi-infinite 299 18 2367
240 205 4486 155 473 150 112 300 64 500 267 Semi-infinite 375 18 4657
245 205 6243 155 409 150 123 300 51 500 317 Semi-infinite 385 18 6401
250 205 2769 155 231 150 130 300 75 500 333 Semi-infinite 285 18 2875
255 205 3580 155 425 150 101 300 122 500 224 Semi-infinite 288 18 3717
260 205 6440 155 470 150 100 300 166 500 245 Semi-infinite 298 18 6686
Mean 2505 295 100 90 305 300 2590

Selected subgrade
temperature modulus
Between wheel paths
140 205 5062 155 233 150 203 300 194 500 343 Semi-infinite 290 19 5255
145 205 4730 155 175 150 146 300 134 500 301 Semi-infinite 326 19 4911
150 205 4233 155 173 150 140 300 105 500 305 Semi-infinite 315 19 4395
155 205 4734 155 161 150 152 300 110 500 270 Semi-infinite 330 18 4915
160 205 3259 155 162 150 121 300 88 500 208 Semi-infinite 287 19 3384
165 205 4416 155 158 150 137 300 74 500 240 Semi-infinite 282 19 4585
170 205 4809 155 202 150 201 300 88 500 198 Semi-infinite 269 19 4993
175 205 4985 155 169 150 151 300 131 500 231 Semi-infinite 277 19 5241
180 205 3694 155 153 150 150 300 113 500 238 Semi-infinite 287 19 3883
185 205 4728 155 156 150 150 300 119 500 232 Semi-infinite 285 19 4970
190 205 4160 155 177 150 148 300 106 500 276 Semi-infinite 293 19 4373
195 205 5397 155 223 150 155 300 110 500 250 Semi-infinite 299 19 5603
200 205 4645 155 191 150 146 300 108 500 265 Semi-infinite 305 19 4822
205 205 4859 155 197 150 156 300 114 500 272 Semi-infinite 307 19 5108
210 205 5789 155 189 150 210 300 120 500 233 Semi-infinite 321 19 6086
215 205 4980 155 222 150 167 300 116 500 281 Semi-infinite 303 19 5235
220 205 5386 155 167 150 141 300 89 500 222 Semi-infinite 320 19 5662
225 205 5330 155 192 150 151 300 110 500 244 Semi-infinite 305 19 5603
230 205 5087 155 186 150 147 300 87 500 221 Semi-infinite 323 19 5348
235 205 5021 155 194 150 134 300 106 500 250 Semi-infinite 309 19 5278
240 205 4043 155 188 150 160 300 81 500 340 Semi-infinite 337 19 4250
245 205 5903 155 207 150 164 300 102 500 249 Semi-infinite 345 19 6206
250 205 4555 155 210 150 166 300 98 500 280 Semi-infinite 319 19 4789
255 205 5314 155 215 150 150 300 109 500 244 Semi-infinite 304 19 5586
260 205 7825 155 256 150 215 300 98 500 204 Semi-infinite 318 19 8226
Mean 4920 190 160 110 255 305 5150

Table G 36: Western Freeway back-calculated moduli of 2.5% bitumen mix two years after opening to traffic
Selected subgrade
temperature modulus
Outer wheel path
5 185 1053 160 100 155 198 300 88 500 161 Semi-infinite 220 17 965
10 185 3864 160 176 155 107 300 64 500 166 Semi-infinite 222 16 3496
15 185 6292 160 200 155 122 300 75 500 172 Semi-infinite 263 16 5623
20 185 3141 160 183 155 108 300 103 500 178 Semi-infinite 218 16 2807
25 185 2182 160 100 155 80 300 150 500 199 Semi-infinite 197 15 1926
30 185 1084 160 163 155 157 300 55 500 140 Semi-infinite 202 15 957
35 185 3126 160 139 155 124 300 101 500 171 Semi-infinite 241 15 2724
40 185 5454 160 324 155 175 300 115 500 185 Semi-infinite 254 15 4753
45 185 5797 160 348 155 162 300 118 500 188 Semi-infinite 260 15 5052
50 185 5743 160 322 155 182 300 102 500 177 Semi-infinite 253 15 5005
55 185 5175 160 320 155 205 300 89 500 241 Semi-infinite 344 15 4510
60 185 5942 160 150 155 175 300 115 500 336 Semi-infinite 313 15 5179
65 185 8362 160 407 155 236 300 102 500 266 Semi-infinite 360 15 7288
70 185 1979 160 194 155 143 300 95 500 470 Semi-infinite 310 15 1725
75 185 2802 160 238 155 139 300 167 500 230 Semi-infinite 278 15 2442
80 185 3874 160 198 155 127 300 71 500 278 Semi-infinite 317 15 3376
85 185 1864 160 129 155 80 300 86 500 196 Semi-infinite 243 15 1625
90 185 1372 160 100 155 80 300 149 500 500 Semi-infinite 216 15 1196
95 185 823 160 100 155 148 300 111 500 500 Semi-infinite 228 15 717
100 185 589 160 259 155 80 300 173 500 465 Semi-infinite 216 14 507
105 185 771 160 100 155 80 300 154 500 500 Semi-infinite 233 15 672
110 185 4084 160 280 155 126 300 67 500 225 Semi-infinite 277 15 3559
115 185 313 160 100 155 109 300 200 500 500 Semi-infinite 315 15 273
120 185 4160 160 196 155 136 300 71 500 265 Semi-infinite 293 15 3626
125 185 5217 160 217 155 134 300 69 500 216 Semi-infinite 307 15 4604
130 185 5818 160 281 155 153 300 103 500 192 Semi-infinite 282 15 5134
135 185 6974 160 266 155 141 300 68 500 200 Semi-infinite 301 15 6155
Mean 3620 207 137 106 271 265 3180

Selected subgrade
temperature modulus
Between wheel paths
5 185 8708 160 204 155 117 300 69 500 151 Semi-infinite 240 16 7781
10 185 9558 160 213 155 93 300 31 500 179 Semi-infinite 296 16 8648
15 185 9232 160 200 155 120 300 61 500 175 Semi-infinite 277 16 8250
20 185 4248 160 207 155 111 300 72 500 183 Semi-infinite 247 16 3796
25 185 4224 160 226 155 114 300 105 500 186 Semi-infinite 246 16 3822
30 185 4072 160 278 155 122 300 124 500 182 Semi-infinite 242 16 3684
35 185 6494 160 488 155 88 300 199 500 225 Semi-infinite 241 16 5876
40 185 8980 160 448 155 157 300 147 500 208 Semi-infinite 270 17 8228
45 185 8447 160 348 155 152 300 190 500 214 Semi-infinite 260 17 7739
50 185 7101 160 328 155 175 300 107 500 295 Semi-infinite 227 17 6506
55 185 7116 160 343 155 188 300 130 500 254 Semi-infinite 343 17 6520
60 185 7585 160 261 155 147 300 90 500 249 Semi-infinite 385 17 6950
65 185 7913 160 337 155 155 300 140 500 267 Semi-infinite 342 17 7341
70 185 7334 160 189 155 156 300 146 500 275 Semi-infinite 326 17 6804
75 185 5671 160 237 155 137 300 55 500 259 Semi-infinite 335 17 5261
80 185 3675 160 152 155 93 300 115 500 186 Semi-infinite 293 18 3452
85 185 3005 160 150 155 81 300 165 500 233 Semi-infinite 294 18 2823
90 185 3251 160 150 155 89 300 194 500 315 Semi-infinite 264 18 3092
95 185 3353 160 150 155 80 300 162 500 235 Semi-infinite 291 18 3189
100 185 2983 160 160 155 80 300 130 500 213 Semi-infinite 302 18 2802
105 185 2912 160 172 155 80 300 58 500 206 Semi-infinite 287 18 2770
110 185 4704 160 196 155 112 300 83 500 193 Semi-infinite 257 19 4531
115 185 2643 160 150 155 122 300 55 500 237 Semi-infinite 247 19 2546
120 185 5351 160 188 155 106 300 54 500 186 Semi-infinite 298 18 5090
125 185 6683 160 232 155 116 300 81 500 190 Semi-infinite 317 19 6437
130 185 7835 160 283 155 144 300 57 500 202 Semi-infinite 305 19 7547
135 185 2732 160 169 155 82 300 103 500 206 Semi-infinite 273 19 2631
Mean 5770 239 119 108 219 285 5340

Table G 37: Western Freeway back-calculated moduli of 3% bitumen mix two years after opening to traffic
Selected subgrade
temperature modulus
Outer wheel path
140 205 3761 155 202 150 138 300 92 500 226 Semi-infinite 285 21 3856
145 205 1060 155 82 150 139 300 190 500 469 Semi-infinite 257 21 1087
150 205 2213 155 156 150 125 300 78 500 321 Semi-infinite 279 21 2269
155 205 1088 155 81 150 89 300 131 500 190 Semi-infinite 235 21 1116
160 205 493 155 80 150 239 300 86 500 499 Semi-infinite 231 21 505
165 205 1269 155 96 150 95 300 83 500 169 Semi-infinite 229 21 1301
170 205 2642 155 82 150 89 300 57 500 173 Semi-infinite 232 21 2709
175 205 3997 155 214 150 125 300 47 500 191 Semi-infinite 288 21 4098
180 205 3249 155 139 150 110 300 57 500 183 Semi-infinite 277 22 3373
185 205 4760 155 102 150 89 300 57 500 198 Semi-infinite 296 22 4942
190 205 3288 155 127 150 126 300 54 500 203 Semi-infinite 287 22 3414
195 205 1553 155 94 150 88 300 133 500 191 Semi-infinite 269 22 1612
200 205 1743 155 85 150 80 300 135 500 196 Semi-infinite 241 22 1832
205 205 1022 155 100 150 80 300 120 500 183 Semi-infinite 224 22 1061
210 205 459 155 80 150 140 300 200 500 500 Semi-infinite 282 22 477
215 205 2641 155 143 150 111 300 55 500 217 Semi-infinite 235 22 2742
220 205 2516 155 186 150 95 300 65 500 206 Semi-infinite 279 22 2612
225 205 1645 155 102 150 89 300 60 500 164 Semi-infinite 233 22 1708
230 205 3136 155 147 150 124 300 55 500 289 Semi-infinite 291 22 3256
235 205 3155 155 94 150 91 300 48 500 228 Semi-infinite 266 22 3317
240 205 8304 155 154 150 87 300 45 500 280 Semi-infinite 418 22 8730
245 205 6799 155 103 150 99 300 40 500 246 Semi-infinite 358 22 7059
250 205 3227 155 151 150 125 300 58 500 194 Semi-infinite 279 22 3392
255 205 3473 155 243 150 172 300 67 500 170 Semi-infinite 283 22 3651
260 205 9830 155 347 150 169 300 104 500 206 Semi-infinite 326 22 10334
Mean 3090 136 117 85 244 275 3220

Selected subgrade
temperature modulus
Between wheel paths
140 205 5421 155 216 150 127 300 121 500 218 Semi-infinite 282 19 5287
145 205 4927 155 394 150 169 300 20 500 343 Semi-infinite 677 19 4805
150 205 3159 155 156 150 124 300 103 500 230 Semi-infinite 310 20 3120
155 205 5031 155 221 150 110 300 60 500 271 Semi-infinite 335 20 4969
160 205 1845 155 150 150 81 300 90 500 163 Semi-infinite 278 20 1822
165 205 300 155 500 150 350 300 76 500 145 Semi-infinite 251 20 296
170 205 5335 155 207 150 128 300 62 500 185 Semi-infinite 279 20 5269
175 205 5463 155 239 150 135 300 72 500 203 Semi-infinite 283 20 5395
180 205 4277 155 202 150 124 300 86 500 205 Semi-infinite 286 20 4224
185 205 6100 155 173 150 119 300 71 500 195 Semi-infinite 302 20 6024
190 205 5443 155 202 150 194 300 76 500 205 Semi-infinite 266 20 5443
195 205 6087 155 160 150 117 300 50 500 209 Semi-infinite 347 20 6087
200 205 4666 155 222 150 94 300 70 500 239 Semi-infinite 295 20 4666
205 205 4417 155 200 150 146 300 65 500 258 Semi-infinite 311 20 4417
210 205 1218 155 382 150 163 300 112 500 216 Semi-infinite 274 20 1218
215 205 6811 155 187 150 114 300 57 500 200 Semi-infinite 342 20 6811
220 205 6650 155 169 150 114 300 49 500 219 Semi-infinite 370 20 6650
225 205 8473 155 150 150 81 300 21 500 377 Semi-infinite 732 21 8580
230 205 6282 155 216 150 165 300 50 500 223 Semi-infinite 378 21 6361
235 205 5733 155 186 150 115 300 61 500 193 Semi-infinite 311 20 5733
240 205 6515 155 235 150 159 300 60 500 225 Semi-infinite 338 20 6515
245 205 5724 155 196 150 104 300 81 500 220 Semi-infinite 364 21 5796
250 205 7627 155 333 150 155 300 77 500 202 Semi-infinite 323 21 7723
255 205 7615 155 296 150 153 300 57 500 189 Semi-infinite 330 21 7711
260 205 9781 155 336 150 166 300 81 500 189 Semi-infinite 315 21 9904
Mean 5400 237 140 69 220 340 5390


AP-T336-18 Design and Performance of Foamed Bitumen Stabilised Pavements-Reduced

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

AP-T336-18 Design and Performance of Foamed Bitumen Stabilised Pavements-Reduced

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

AP-T336-18 Design and Performance of Foamed Bitumen Stabilised Pavements-Reduced

Uploaded by

Copyright:

Available Formats

Technical Report

Design and Performance of Foamed

Geoff Jameson Austroads Ltd.

Abstract About Austroads

Keywords Austroads is governed by a Board consisting of senior

This final project report:

The major project conclusions were:

Austroads 2018 | page i

Austroads 2018 | page i

5.4 Western Freeway, Ballan ............................................................................................................... 31

Austroads 2018 | page ii

Appendix A Kwinana Freeway, Baldivis ............................................................................................ 65

Table 2.1: TMR early minimum modulus requirements ................................................................................ 3

Austroads 2018 | page iii

This final report for project TT1825 is structured as follows:

Austroads 2018 | page 1

2. Mix Design Procedures

2.2 Background to Indirect Tensile Modulus Requirements

2.2.1 Pioneering Work in the 1990s

Austroads 2018 | page 2

2.2.2 Early Development by TMR

Table 2.1: TMR early minimum modulus requirements

Parameter Minimum requirement

Source: Adapted from Ramanujam and Jones (2000).

Austroads 2018 | page 3

Table 2.2: TMR modulus requirements in 2004

Source: Jones and Ramanujam (2004).

Austroads 2018 | page 4

2.2.3 Current TMR Test Methods and Modulus Requirements

Minimum ‘cured’ indirect Minimum ‘soaked’

Minimum ‘cured’ indirect Minimum ‘soaked’

Austroads 2018 | page 5

2.3 Development of Austroads Mix Design Test Methods

The basis for these methods is described below.

Table 2.6: Austroads test methods for FBS mix design

Austroads 2018 | page 6

2.3.2 Determination of Foaming Properties

2.3.3 Mixing of Foamed Bitumen Materials

2.3.4 Compaction of Test Cylinders

Austroads 2018 | page 7

2.3.5 Curing of Modulus Specimens

Austroads 2018 | page 8

2.3.6 Resilient Modulus Measurement

2.4 Description of Procedure to Determine Binder Contents

Table 2.7: Selecting mix components

Table 2.8: Determination of design bitumen content

Austroads 2018 | page 9

3. Effect of Moisture Content and Lime Type on

Table 3.1: Variation in moisture contents used in mix designs

Moisture content (%)

Laboratory Newell Highway Western Freeway

Austroads 2018 | page 10

3.2 Influence of Moisture Content on Newell Highway Trial Mix

Figure 3.1: Influence of moisture on the Newell Highway material compactability

8.8% Moisture content

Cycles of gyratory compaction

Austroads 2018 | page 11