rubberised asphalt testing to uk standards 032 final report.pdfrubber type. thus variations in...

Laboratory Report - Final

Rubberised Asphalt Testing to UK Standards

Project code: TYR032-001 Research date: 15/10/07 – 31/03/08 Date: November 2008

WRAP helps individuals, businesses and local authorities to reduce waste and recycle more, making better use of resources and helping to tackle climate change.

Written by: Ronald Kibuuka, Daru Widyatmoko, Richard Elliott, James Grenfell, Gordon Airey, Andrew Collop

Front cover photography: Asphalt cores WRAP and Scott Wilson believe the content of this report to be correct as at the date of writing. However, factors such as prices, levels of recycled content and regulatory requirements are subject to change and users of the report should check with their suppliers to confirm the current situation. In addition, care should be taken in using any of the cost information provided as it is based upon numerous project-specific assumptions (such as scale, location, tender context, etc.). The report does not claim to be exhaustive, nor does it claim to cover all relevant products and specifications available on the market. While steps have been taken to ensure accuracy, WRAP cannot accept responsibility or be held liable to any person for any loss or damage arising out of or in connection with this information being inaccurate, incomplete or misleading. It is the responsibility of the potential user of a material or product to consult with the supplier or manufacturer and ascertain whether a particular product will satisfy their specific requirements. The listing or featuring of a particular product or company does not constitute an endorsement by WRAP and WRAP cannot guarantee the performance of individual products or materials. This material is copyrighted. It may be reproduced free of charge subject to the material being accurate and not used in misleading context. The source of the material must be identified and the copyright status acknowledged. This material must not be used to endorse or used to suggest WRAP's endorsement of a commercial product or service. For more detail, please refer to WRAP's Terms & Conditions on its web site: www.wrap.org.uk

Rubberised asphalt testing to UK standards 1

Executive summary Scott Wilson was commissioned to undertake a desk study and a suite of laboratory testing using a wide range of UK materials for the WRAP project TYR032: “Rubberised Asphalt Testing to UK Standards”. The outcome of the desk study on the use of RA and communication with relevant industry experts, leading to identification and provision of information on the practical issues associated with Rubberised Asphalt (RA), has been presented in a separate report. The laboratory testing was aimed at determining a number of RA mix designs that could be used in surface and binder courses, specifically to address some technical issues identified during the preparation stages of an earlier demonstration trial (that subsequently did not take place). This report presents the findings from the laboratory work carried out by Scott Wilson in collaboration with the Nottingham Transportation Engineering Centre (NTEC). The RA and the control mixtures were manufactured using a wide range of UK materials and assessed to UK Standards. The raw materials used in the study comprised 2 bitumen sources (each with 2 bitumen grades), 3 aggregate types, 3 rubber types and sources, 2 RA mixture designs and 2 control asphalt mixtures. The study comprised 10 separate tasks, each covering different target deliverables including project management, meetings, laboratory testing and final reporting. Prior to carrying out laboratory testing (Tasks 4 to 9), a preliminary assessment was carried out. It highlighted the sensitivity of rubberised bitumen to blending arrangements and test conditions and the extent to which these may affect the test results. Following the assessment, a blending protocol was agreed and it included high shear blending at 6000 rpm for 15 minutes, with rubber being fed in during the first 10 minutes, followed by reducing the shear rate to 1500 rpm for the rest of the remaining blending time. A Brookfield viscometer with spindle size No. 27 was used. Task 4 testing involved subjecting 48 rubberised bitumen blends, of different rubber contents/sources and base bitumen sources/grades, to viscosity (ASTM 6114), penetration (EN 1426) and softening point (EN 1427) testing from which the ‘best’ six blends meeting a target viscosity between 1500cP – 5000cP were selected. Under Task 6, the preferred six rubberised bitumen blends from Task 4 were assessed for the effect of rubber particle size on the blend properties, where it was found that finer particle sizes led to greater reaction and interaction with base bitumen hence a higher viscosity of the rubberised bitumen. The three “best” performing rubberised bitumen blends were then selected to move forward to RA testing. Tasks 5 and 7 gathered real information about the properties and performance characteristics of the three blends from Task 6 and from this information it was found that the selected blends satisfied the requirements for ageing index, softening point and viscosity. Venezuelan bitumen blends met the criterion for resilience whilst the Middle Eastern blend fell short of the recommended resilience value. However, the resilience value remained a lot higher than that of the base bitumen suggesting that there is significant improvement in resilience properties after rubber addition. In addition, the blends exhibited improvements in short and long term ageing, cohesion characteristics, Fraass breaking point, service temperature range and flash point values compared to typical penetration grade bitumen, and satisfactory residual properties after storage stability testing. Task 9 involved a two step process, the first being a screening assessment (workability assessment) of twelve RA mixtures and four control mixtures and the second being mechanical assessment of the eight best performing blends from the screening assessment and two control mixtures, i.e. Porous Asphalt (PA) and Stone Mastic Asphalt (SMA). Due to the limited time available to complete this study, a comprehensive mixture design exercise was not carried out; instead, designs typically adopted for these asphalt mixtures were used. From the screening assessment, it was found that almost all the rubberised asphalt mixtures had improved workability compared with the control mixtures, probably as a result of the higher mixing temperature and higher binder content. Bulk densities generally remained constant throughout mixing and holding, whereas the RA mixtures showed improved load spreading ability compared to the control mixtures. The tensile strength values of all the RA mixtures were greater than 600 kPa, indicating good quality material. The RA mixtures all showed good resistance to deformation and low temperature cracking. Test results showed that all the RA mixtures had better resistance to particle loss compared with their respective control samples and almost all had a low risk of binder drainage. In addition, the retained stiffness ratios of all the surface course RA

samples were above 1 and consistently greater than those for the PA control samples, indicating that the RA blends would not be expected to be susceptible to water related durability problems while in service. It should be noted that work carried out during this study was limited to one source (processing plant) for each rubber type. Thus variations in rubberised bitumen properties (due to use of rubber other than from these sources) remain unknown; this should be the subject of further research. Nevertheless, the study highlighted a number of benefits from using rubberised asphalt, specifically that the material exhibited laboratory performance appropriate for heavy duty long life surface course application. It is recommended that further design optimisation is carried out before adopting any of the mixtures investigated here for larger scale applications, to ensure that a greater level of enhanced performance can be obtained. Good quality control is always essential to ensure good quality rubberised asphalt can be manufactured and laid, and ultimately deliver the target level of performance. In summary, the comprehensive laboratory study described in this report has shown the overall performance of RA mixtures to be at least similar to that of premium asphalt binder course (e.g. SMA) and significantly better than a conventional porous surface course, due to improved tensile strength, and superior resistance to deformation, low temperature cracking and aggregate loss (fretting). Good quality RA mixtures can therefore be regarded as strong contenders for heavy duty long life surface course applications.


Contents

1.0 Introduction ............................................................................................................................. 6 1.1 Background .......................................................................................................................... 6 1.2 Scope of Work...................................................................................................................... 6

2.0 Materials for Testing ................................................................................................................ 8 2.1 Samples for Test................................................................................................................... 8 2.2 Properties of Base Bitumen ................................................................................................... 8 2.3 Grading of Reclaimed Tyre Rubber......................................................................................... 8 2.4 Aggregate Properties ............................................................................................................ 9

3.0 Preliminary Work ................................................................................................................... 10 3.1 Reproducibility Exercise........................................................................................................10

3.1.1 Blending Protocol 1 .................................................................................................10 3.1.2 Blending Protocol 2 .................................................................................................10 3.1.3 Blending Protocol 3 .................................................................................................11

3.2 Adopted Test Method...........................................................................................................12 4.0 Task 4: Preliminary ‘Screening’ Test ...................................................................................... 12

4.1 Test Matrix ..........................................................................................................................13 4.2 Grading of Crumb Rubbers ...................................................................................................13 4.3 Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre .................................15 4.4 Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre.....................................16 4.5 Viscosity of Rubberised Bitumen Blends with 50% Ambient Car/50% Ambient Truck Tyres ......17 Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre ...............................................18 4.6 Penetration and Softening Point Values .................................................................................19 4.7 The Best Blends...................................................................................................................19

5.0 Task 6: Assessment of Effect of Rubber Particle Size ............................................................ 20 5.1 Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre (Fine) ...........................21 5.2 Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre (Fine)........................22 5.3 Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre (Fine).........................23 5.4 Penetration and Softening Point Values .................................................................................24 5.5 Selected Blends ...................................................................................................................24

6.0 Task 5: Comprehensive Testing of the Most Suitable Blends................................................. 25 7.0 Task 7: EN 14023 Tests on the Three Best Performing Blends .............................................. 27 8.0 Workability Assessment......................................................................................................... 29

8.1 Workability of Loose Mixture.................................................................................................30 8.2 Bulk Densities......................................................................................................................33 8.3 Maximum Densities..............................................................................................................34 8.4 Stiffness..............................................................................................................................35 8.5 Tensile Strength ..................................................................................................................36

9.0 Mechanical Assessment.......................................................................................................... 38 9.1 Specimen Manufacturing ......................................................................................................38 9.2 Test Methodology ................................................................................................................38 9.3 Mixture Volumetrics .............................................................................................................38 9.4 Deformation Resistance........................................................................................................39 9.5 Crack Resistance..................................................................................................................41 9.6 Fatigue Resistance...............................................................................................................42 9.7 Retained Stiffness ................................................................................................................45 9.8 Binder drainage ...................................................................................................................45 9.9 Durability ............................................................................................................................46

10.0 Conclusions and Recommendations....................................................................................... 47 11.0 Acknowledgements ................................................................................................................ 51 12.0 References ............................................................................................................................. 51


Figures Figure 1: Grading of Crumb Rubbers (Standard Size) ...................................................................................... 14 Figure 2: Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre ............................................ 15 Figure 3: Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre................................................ 16 Figure 4: Viscosity of Rubberised Bitumen Blends with 50% Ambient Car and 50% Ambient Truck Tyres ........... 17 Figure 5: Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre ............................................. 18 Figure 6: Grading of Crumb Rubbers (Finer Size) ............................................................................................ 20 Figure 7: Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre................................................ 21 Figure 8: Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre ............................................ 22 Figure 9: Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre ............................................. 23 Figure 9 (Cont.): Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre .................................. 24 Figure 10: Asphalt Mixture Gradations............................................................................................................ 29 Figure 11: Asphalt Mixer................................................................................................................................ 31 Figure 12: Workability of Loose Surface Course Mixtures ................................................................................. 32 Figure 13: Workability of Loose Binder Course Mixtures................................................................................... 32 Figure 14: Bulk Densities of Surface Course Mixtures ...................................................................................... 33 Figure 15: Bulk Densities of Binder Course Mixtures ........................................................................................ 33 Figure 16: Maximum Densities of Surface Course Mixtures .............................................................................. 34 Figure 17: Maximum Densities of Binder Course Mixtures ................................................................................ 34 Figure 18: Stiffness Test Arrangement ........................................................................................................... 35 Figure 19: Surface Course Stiffness Assessment ............................................................................................. 35 Figure 20: Binder Course Stiffness Assessment ............................................................................................... 36 Figure 21: Tensile Strength Test Arrangement ................................................................................................ 36 Figure 22: Surface Course Tensile Strength .................................................................................................... 37 Figure 22: Binder Course Tensile Strength...................................................................................................... 37 Figure 23: Wheel Tracking Test Arrangement ................................................................................................. 40 Figure 24: Wheel Tracking Data..................................................................................................................... 40 Figure 25: Crack Resistance Test Arrangement ............................................................................................... 41 Figure 26: Crack Resistance Data................................................................................................................... 42 Figure 27: Fatigue Test Arrangement ............................................................................................................. 43 Figure 28: Resistance to Fatigue Cracking ...................................................................................................... 43

Tables Table 1 : Sample Identification ....................................................................................................................... 8 Table 2: Properties of Base Bitumen................................................................................................................ 8 Table 3: Grading of Reclaimed Tyre Rubber..................................................................................................... 9 Table 4: Aggregate Properties ........................................................................................................................ 9 Table 5: Results from Samples produced using Blending Protocol 1..................................................................10 Table 6: Results from Samples produced using Blending Protocol 2..................................................................11 Table 7: Results from Samples produced using Blending Protocol 3..................................................................12 Table 8: Test Matrix ......................................................................................................................................13 Table 9: Penetration (dmm) at 25oC...............................................................................................................19 Table 10: Ring and Ball Softening Point (oC) ...................................................................................................19 Table 11: Penetration and Softening Point Test Results ...................................................................................24 Table 12: Comprehensive Test Results ...........................................................................................................26 Table 13: EN 14023 Test Results ...................................................................................................................27 Table 14: Sample References.........................................................................................................................30 Table 15: Workability Test Parameter.............................................................................................................31 Table 16: Mixtures for Mechanical Assessment................................................................................................38 Table 17: Summary of Mixture Volumetrics.....................................................................................................39 Table 18: Failure Cycles at 200 Microstrain .....................................................................................................44 Table 19: Summary of Stiffness Ratios of Water Sensitivity Data......................................................................45 Table 20: Binder Drainage Test Results ..........................................................................................................46 Table 21: Particle Loss Test Results ...............................................................................................................46


Appendices Appendix A Workability Data Summary Appendix B Bulk Density Data Appendix C Maximum Density Data Appendix D Mixture Stiffness Appendix E Mixture Tensile Strength Appendix F Wheel Track Test Data Appendix G Crack Resistance Data Appendix H Fatigue Resistance Data Appendix I Retained Stiffness Appendix J Binder Drainage Appendix K Durability Test Data


1.0 Introduction

1.1 Background WRAP (The Waste & Resources Action Programme) is a not-for-profit private company backed by funding from the Department for Environment, Food and Rural Affairs, the Scottish Executive, the Welsh Assembly Government and Northern Ireland Executive. WRAP works in partnership to encourage and enable businesses and consumers to be more efficient in their use of materials and recycle more things more often. This helps to minimise landfill, reduce carbon emissions and improve our environment. Rubberised asphalt (RA) using the wet process is widely used within road construction projects in the United States, South Africa and parts of mainland Europe and Australia. A stakeholder seminar aimed to assess the feasibility of using the rubberised asphalt material in the UK was held by WRAP in November 2005. From the seminar feedback, it was decided that more information was required and therefore WRAP agreed to undertake a demonstration trial within an actual road construction project that would include importation of the specialist blending equipment required to mix tyre rubber with bitumen. In preparation for this trial, it became apparent that the selection of appropriate bitumen and rubber crumbs could be a critical factor. This trial however did not take place and subsequently a new contract, TYR032, was let to Scott Wilson (SW) in October 2007. SW was to undertake a desk study and a suite of laboratory testing using a wide range of UK materials. The laboratory testing had two functions:

To provide information to potential users of rubberised binder on the properties that can be

expected from various sources/types of crumb rubber available in the UK, when blended with

typical UK bitumens. This will assist potential users with mixture design;

To demonstrate that asphalt mixtures suitable for surface and binder courses can be produced and

to determine the relative performance of these.

To secure prompt delivery of the work program, the laboratory testing was carried out in partnership with the Nottingham Transportation Engineering Centre (NTEC), with SW in the leading role.

1.2 Scope of Work

Works carried out were divided into a number of tasks as detailed in WRAP’s contract document TYR032. They are reproduced here as follows:

TASK 1: PROJECT MANAGEMENT

A senior member of SW staff was responsible for the daily management of the project. This was seen as vital to the success of the project since substantial co-ordination and communication was required to deliver within the given timescales. The Project Manager was supported by a Project Director who was responsible for the Technical Review to ensure that the project met the highest standards of testing and reporting. Works carried out at NTEC were also managed by an equivalent Project Manager and Project Director.

TASK 2: MEETINGS

The project timetable stressed the importance of project participant meetings to the delivery of the programme; hence these meetings were identified as a separate task within the project plan. These were organised under the direction of the SW Project Manager. SW prepared notes from these meetings, including actions required, for circulation to all parties. These notes were used to support the project management process, and to provide a record of activities.

TASK 3: DESK STUDY

The study researched the use of RA and communication with relevant industry experts led to identification and provision of information on the practical issues associated with RA. This included receiving technical advice from experts within SW and NTEC, consultation with UK suppliers, and discussions with those using RA overseas. The desk study is presented in a separate report.


TASK 4: PRELIMINARY ‘SCREENING’ TESTS

The screening tests as detailed in the contract comprised manufacturing a total of 48 rubberised bitumen blends whose detailed composition was agreed with WRAP. This was followed by measurement of viscosity, penetration and softening point of each blend.

TASK 5: COMPREHENSIVE TESTING OF THE MOST SUITABLE BLENDS

Comprehensive testing was carried out on three of the most suitable rubberised bitumen blends (from TASK 4 and TASK 6) which were selected based on the viscosity test results. This included discussion with WRAP about the proposed blends before moving on to the comprehensive laboratory assessment.

TASK 6: ASSESSMENT OF THE EFFECT OF RUBBER PARTICLE SIZE

The three most suitable rubberised bitumen blends (confirmed in TASK 4) were further assessed for the effect of rubber particle size on the blend properties. At the end of the above suite of assessment, the three best performing rubberised bitumen blends were selected, and taken forward to TASK 5 and TASK 7.

TASK 7: EN 14023 TESTS ON THE THREE BEST PERFORMING BLENDS

As required by the tender invitation document, the three preferred rubberised bitumen blends, selected at the end of TASK 6, were further assessed in accordance with EN 14023 test requirements but excluding the elastic recovery tests (EN 13398).

TASK 8: ANALYSIS AND INTERIM REPORT

Data generated in the Stage 1 assessment (TASKS 4 – 7) and the respective analysis was compiled and presented in an interim report. This short report summarised the laboratory testing results and recommended which rubberised bitumen blends to be taking forward to rubberised asphalt testing (TASK 9).

TASK 9: ASSESSMENT OF RUBBERISED ASPHALT MIXTURES

A screening assessment on twelve number RA mixtures (six surface course and six binder course materials) and four sets of control asphalt mixtures was carried out on mixtures prepared in the laboratory by using bespoke workability tests, developed by SW. A total of eight ‘best performing’ RA and two control asphalt mixtures were selected from the screening assessment and subjected to further mechanical tests. RA and control samples were manufactured by a laboratory roller compactor to the target air voids specified for surface and binder courses respectively, followed by removal of cores with dimensions appropriate for testing.

TASK 10: FINAL REPORTING

The task comprised a comprehensive report which included all laboratory testing procedures and results, lesson learnt and recommendations for larger scale trials. This report forms the main technical output of the project.


2.0 Materials for Testing

2.1 Samples for Test The following are the constituent materials received from suppliers and used during the testing. They were each assigned SW laboratory reference numbers as shown in Table 1 below:

Table 1 : Sample Identification SWPE Reference

Description Supplier

07/058AA Cryogenic Car Tyre Rubber 0.4-0.63 mm TyreGenics 07/058AA Cryogenic Car Tyre Rubber 0.63-1.4 mm TyreGenics 07/058AA Cryogenic Car Tyre Rubber 1.4-2.0 mm TyreGenics 07/140A Venezuelan Bitumen 40/60 Nynas 07/140B Venezuelan Bitumen 100/150 Nynas 07/140C Middle Eastern Bitumen 40/60 UK Bitumen 07/140D Middle Eastern Bitumen 100/150 UK Bitumen 07/140E Ambient Truck Tyre Rubber Size 12 J. Allcock and Sons 07/140F Ambient Truck Tyre Rubber Size 30 J. Allcock and Sons 07/140G Ambient Truck Tyre Rubber 4-6mm J. Allcock and Sons 07/140H Ambient Car Tyre Rubber 0.5-2mm (Coarse) Monckton Rubber Technology 07/140I Ambient Car Tyre Rubber 0-0.5mm (Fine) Monckton Rubber Technology 07/140J Limestone aggregate Aggregate Industries, Ivonbrook Quarry 07/140P Granite aggregate Aggregate Industries, Croft Quarry 07/140V Gritstone aggregate Tarmac Dolyhir Quarry 07/140T Hydrated Lime Singleton Birch, Melton Ross Quarries

2.2 Properties of Base Bitumen The base bitumen used to manufacture the rubberised bitumen blends comprised four different binder types (two hard and two soft binders) that were produced from Venezuelan and Middle Eastern crudes. Results of initial tests on these binders have been summarised in the table below:

Table 2: Properties of Base Bitumen

SARA Analysis Description

Penetration at 25oC (dmm)

Softening Point (oC)

Viscosity at 177oC (cP)

Saturates (%)

Aromatics (%)

Resins (%)

Asphaltenes (%)

Venezuelan Bitumen 40/60

54 52.2 109 7.0 55.9 20.4 16.7

Venezuelan Bitumen 100/150

135 42.0 72 8.5 55.1 18.9 17.6

Middle Eastern Bitumen 40/60

50 52.5 80 3.4 60.9 21.2 14.4

Middle Eastern Bitumen 100/150

124 41.0 75 4.4 64.8 19.9 10.9

2.3 Grading of Reclaimed Tyre Rubber


The rubber reclaimed from either car or truck tyres, from either ambient or cryogenic production method, were supplied in sizes available as “stock products”; their particle size distributions (gradings) were subsequently determined. The results are summarised below (“?” denotes possibility for the grading falling within the stated category):

Table 3: Grading of Reclaimed Tyre Rubber Type Cryogenic Car Ambient Car Ambient Truck

Rubber Size

Target Grading

0.4-0.63 mm

0.63-1.4 mm

1.4-2.0 mm

Fine Coarse Truck No.30

Truck No.12

Truck 4-6mm

Sieve Size (mm)

Cumulative % Passing

5 - 100 100 100 100 100 100 100 100

3.35 - 100 100 100 100 100 100 100 92

2 100 100 100 100 100 80 100 100 61

1.18 65 – 100 100 85 12 100 9 100 77 12

0.600 20 – 100 100 23 6 88 0 87 41 2

0.300 0 – 45 71 5 2 29 0 7 34 2

0.063 0 – 5 2 0 0 1 0 0 0 2

PAS 107 Category

- Fine - Granulate Fine? Granulate Fine? - Granulate?

According to Publicly Available Specification (PAS) 107, rubber particles having sizes of 1 – 10mm, 0 – 1mm, and 0 – 0.5mm are classified as granulate, powder and fine powder respectively. These categories are also shown in the above table, for cross-reference purposes. Some of the “stock products” presented in the above table do not comply with the PAS 107 categories. This is an issue that needs to be addressed by the recycled rubber industry, and it is hoped that they can agree with standardised classes for rubber particle size (such as that proposed in PAS 107) in the future. Nevertheless, it is expected that a blend of two rubber sizes may be required in order to meet the target grading for use in RA.

2.4 Aggregate Properties

The table below summarises the outcome of tests to determine the mechanical and physical properties of the aggregates.

Table 4: Aggregate Properties

MECHANICAL/PHYSICAL BS Ref. Limestone Gritstone Granite PROPERTIES Coarse Fine Coarse Fine

Apparent Relative Density BS EN 1097-6 2.71 2.62 2.77 2.69 2.67

S.S.D. Relative Density BS EN 1097-6 2.68 2.64 2.74 2.61 2.65

Oven Dry Relative Density BS EN 1097-6 2.65 2.62 2.72 2.57 2.64

Water Absorption (%) BS EN 1097-6 0.9 0.6 0.7 1.5 2.1

Aggregate Abrasion Value BS EN 1097-8 4.9 3.4

Polished Stone Value BS EN 1097-8 66 58

Micro Deval Coefficient BS EN 1097-1 23 22 14

Los Angles Coefficient BS EN 1097-2 33 15 21

Loose Bulk Density (Mg/m3) 1.5

Magnesium Sulphate Soundness Value

BS EN 1367-2 5

7 9


3.0 Preliminary Work Since the laboratory testing of rubber/bitumen binders was carried out in parallel by two different laboratories, it was considered important that a reproducibility exercise was carried out to minimise any variations that might be produced by these laboratories. A harmonised testing protocol was also established.

3.1 Reproducibility Exercise A blending protocol exercise was initially carried out to assess the reproducibility of results between the NTEC and SW laboratories. Throughout this assessment process, regular communication and discussion took place between NTEC, SW, WRAP and WRAP’s Technical Advisor. The material used in this exercise was grade 100/150 bitumen ex-Middle Eastern crude plus 18.5% ambient truck tyre rubber.

3.1.1 Blending Protocol 1 Initially, blending was carried out to the protocol detailed in the Contract Specification; however, it appeared that, for the adopted blending equipment, the specified shear rate of 700 rpm was not high enough to create a vortex within the binder sample and the rubber remained floating on top of the bitumen. Both laboratories used exactly the same high shear Silverson L4RT mixer with mixing head normally used for blending polymer. Consequently, the two laboratories improvised the blending speed independently in order to force the rubber to mix into the bitumen; the results are summarised in Table 5.

Table 5: Results from Samples produced using Blending Protocol 1

Date 30/10/2007 NTEC

30/10/2007 SWPE

Viscometer Brookfield Bohlin Spindle size 34 25 Tested by NTEC SWPE Blending protocol 1 1 Viscosity [cP] after time (minutes) 30 9070 2900 60 4700 1060 120 1070 850 180 830 - Penetration (dmm) after 180 minutes 121 - Softening Point (oC) after 180 minutes 53.8 -

Table 5 shows a significant difference between NTEC and SW results, with both showing viscosity values of the blend reducing with increased blending time. Several factors were suspected:

Non-uniform blending due to inadequate blending speed over time;

Rubber was fed in over too short a time period (less than 5 minutes);

Consequently, the blending procedure was modified to enable the rubber to mix in (Blending Protocol 2) and to extend the period for the feeding of rubber to within the first 10 minutes of blending. 3.1.2 Blending Protocol 2 The second blending was carried out in the following sequence: 1 Apply high shear mixing at 6000 rpm for 10 minutes and feed in rubber during this period; 2 Reduce the mixing speed to 1500 rpm for 10 minutes; 3 Increase the mixing speed to 6000 rpm for 10 minutes, then take a sample at 30 minutes and test;


4 Resume mixing at 1500 rpm for 20 minutes; 5 Increase the mixing speed to 6000 rpm for 10 minutes, then take a sample at 60 minutes and test; 6 Resume mixing at 1500 rpm for 50 minutes; 7 Increase the mixing speed to 6000 rpm for 10 minutes, then take a sample at 120 minutes and test; 8 Resume mixing at 1500 rpm for 50 minutes; 9 Increase the mixing speed to 6000 rpm for 10 minutes, and then take a sample at 180 minutes and test.

The results are summarised in Table 6 below.



31/10/2007 SWPE

01/11/2007 NTEC

Viscometer Brookfield Bohlin Brookfield Spindle size 34 25 27 Blending protocol 2 2 2 Viscosity (cP) after time (minutes) 30 12110 2500 7720 60 2560 800 5490 120 1790 600 4370 180 960 400 2600 Penetration (dmm) after 180 minutes 118 108 87 Softening Point (oC) after 180 minutes 54.2 49.2 61

The results produced by the two laboratories as presented in Table 6 show some variations in the two samples produced by NTEC and a significant difference between NTEC and SW results. The viscosity values of the blend were reducing with increased blending time. Several factors were suspected:

Non-uniform blending due to inadequate blending speed over time;

Different viscometer used in the two laboratories: Brookfield (NTEC) and Bohlin (SW)

viscometers;

Different size of spindle used in these viscometers;

Different containers used during blending:

o NTEC used a 1 litre tin over a hotplate

o SW used a 5 litre mixing bowl placed in a heated mantle.

Consequently, a further modification to the test protocol was adopted by:

modifying the blending procedure (Blending Protocol 3)

using the same viscometer (Brookfield) but with a spindle size No. 27

using the same blending container (1 litre tin) over a hotplate followed by a further

discussion of the results.

The spindle size No. 27 was the same size as that used by Consulpav, who were advisers to the original (abandoned) demonstration trial. From a close look at the available results, it was also suspected that the blend was sensitive to applied shear rate (as opposed to rpm value). 3.1.3 Blending Protocol 3 Blending was carried out in the following sequence: 1 Apply high shear mixing at 6000 rpm for 30 minutes and feed in rubber during the first 10 minute period; 2 Stop and take a sample at 30 minutes and test; 3 Resume mixing at 1500 rpm for the remaining minutes. Samples taken and tested at 60, 120 and 180

minutes. The results are presented in Table 7.




02/11/2007 SWPE

Viscometer Brookfield Brookfield Spindle size 27 27 Blending protocol 3 3 Viscosity (cP) after time (minutes) 30 7240 8000 60 10900 19800 120 19800 19600 180 1200 out of range Pen after 180 minutes 111 n/a SP (oC) after 180 minutes 56.2 n/a

The results produced by the two laboratories are presented in Table 7, showing:

more comparable test results produced by NTEC and SW;

both laboratories reported viscosity values increasing with increased blending time (apart

from the 180 minute blending);

very high viscometer reading during 60-120 minute blending recorded by NTEC and SW.

3.2 Adopted Test Method

Whilst the results obtained by both laboratories from Protocol 3 appear to be more comparable, there was a concern about the effect of prolonged high shear blending. It was subsequently agreed between NTEC, SW, WRAP and Technical Advisor to WRAP that the blending protocol to be adopted for the main test program would involve the following:

Apply high shear mixing at 6000 rpm for 15 minutes and reduce the mixing speed to 1500

rpm for the remaining time. Feed in rubber during the first 10 minute period;

Stop and take a sample at the 30th minute and test;

Resume mixing at 1500 rpm for the remaining minutes. Samples to be taken and tested at

60, 120 and 180 minutes.

4.0 Task 4: Preliminary ‘Screening’ Test According to the ASTM D8 definition, rubberised bitumen is a blend of bitumen, reclaimed tyre rubber and certain additives in which the rubber component is at least 15 percent by weight of the total blend and has reacted in the hot bitumen sufficiently to cause swelling of the rubber particle. Consequently, therefore, rubberised bitumen is not a homogenous blend; traces of swelled rubber particles are still visible in the finished blend. Experience in the US, Australia, South Africa and parts of Europe indicate that rubber is typically added at a rate between 18 and 22% by weight of the total binder. Consequently, therefore, rubber additions of 15, 18.5 and 22 percent by weight of the total binder has been adopted for this study, to cover the above range of minimum and maximum rates of rubber addition. As mentioned previously, four different sources of rubber (one of which is a blend of two sources) and four different base bitumens have been selected for this project. Combination of these variables resulted in a total of 48 rubberised bitumen blends. Each of the 48 rubberised bitumen blends was manufactured using the blending procedure detailed in Paragraph 3.2, specifically: by blending/mixing in a Silverson L4RT high shear apparatus, at 177oC and 6000 RPM for 15 minutes and 1500 RPM for 165 minutes, resulting in 180 minutes (3 hours) of blending. The specification for


rubberised binder adopted in this study required the viscosity test results (at 177oC and 20 RPM) of these blends to be between 1500cP – 5000cP. Subsequently viscosity measurement on the 48 rubberised bitumen blends and 4 unmodified bitumen control (i.e. base bitumen) samples was carried out. The viscosity testing was initially carried out to AASHTO T316 in accordance with the WRAP Contract Document TYR032; however, during the course of testing, it was found that another method, ASTM D6114 is more universally adopted for testing rubberised bitumen. Consequently, the latter was adopted for the remainder of the tests. Each laboratory carried out the test using a Brookfield Viscometer with a No. 27 spindle. In addition, a parallel test was also carried out by SW using a Bohlin viscometer. For consistency, however, the main analysis was based on the results obtained from the Brookfield Viscometer testing. Testing was performed on samples taken at 30, 60, 120 and 180 minutes during blending/mixing. In addition, penetration (EN 1426) and softening point (EN1427) tests were carried out on each blend at the end of the mixing period (180 minutes).

4.1 Test Matrix NTEC tested Middle Eastern bitumen while SW tested Venezuelan bitumen as shown in the Test Matrix in Table 8. Each laboratory tested 24 bitumen-rubber blends plus 2 control (base) bitumen samples.

Table 8: Test Matrix

15% Rubber/85%Bitumen 18.5% Rubber/81.5%Bitumen

22% Rubber/78%Bitumen

TESTING 1

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

Venezuelan 40/60

Venezuelan 100/150

Middle East 40/60

Middle East 100/150

Tests performed by SW Tests performed by NTEC

4.2 Grading of Crumb Rubbers A target grading (Tender Invitation Document (TID)) envelope, based upon that successfully used in the US, was specified in the Contract Documents TYR032. A 50/50 blend of fine and coarse car tyre rubber was required in order to meet the grading requirements. Figure 1 below shows the outcome of the grading exercise.


Figure 1: Grading of Crumb Rubbers (Standard Size)

0

10

20

30

40

50

60

70

80

90

100

0.01 0.1 1 10

Sieve Size (mm)

%Pa

ssin

g

100% Car Tyre [50/50 blend fine/coarse]

100% Truck Tyre [No.12]

50%Car + 50%Truck Tyres

100% Cryogenic [0.63-1.4mm]

Upper TID Limit

Lower TID Limit

All the supplied rubber materials were found to lie within the permitted limits as shown in the figure above, although it should be noted that they were generally closer to the lower (coarser) limit. The figures that follow show a summary of the viscosity values obtained during the testing exercise; only data generated by Brookfield Viscometer are presented in this section. Results are tabulated and summarised according to the binder type and rubber percentage, for example VE22 stands for Venezuelan bitumen mixed with 22% rubber and ME22 for Middle Eastern bitumen mixed with 22% rubber.


4.3 Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre

Figure 2: Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre

Venezuelan 100/150 BITUMEN

Blending Time (minutes)

30 60 120 180

VE22 * * * * VE18.5 1790 4870 16332 16785 VE15 362 868 812 884

100

1000

10000

100000

0 50 100 150 200

Blending Time (mins)

Visc

osity

(cP)

VE18.5

VE15



30 60 120 180

VE22 * * * * VE18.5 7055 11680 39226 * VE15 1148 1194 2218 3340

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

VE18.5

VE15

Middle East 100/150 BITUMEN


30 60 120 180

ME22 * * 24600 7100 ME18.5 2210 2580 4677 10190 ME15 1420 2180 1360 1000

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP) ME22

ME18.5

ME15



30 60 120 180

ME22 22700 4800 3200 1500 ME18.5 30970 5390 1950 1340 ME15 7280 4030 3180 1790

1000

10000

100000

30 60 120 180


Visc

osity

(cP) ME22

ME18.5

ME15

Note: * results are not available due to either very high viscosity or out of measurement range


4.4 Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre

Figure 3: Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre

Venezuelan 100/150 BITUMEN Blending Time (minutes)

30 60 120 180

VE22 * * * * VE18.5 4750 7563 11570 13680 VE15 1240 2235 4232 5170

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

VE18.5

VE15


30 60 120 180

VE22 8495 4150 6510 * VE18.5 2140 2341 3827 4630 VE15 1089 1945 2240 1880

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

VE22

VE18.5

VE15

Middle East 100/150 BITUMEN Blending Time (minutes)

30 60 120 180

ME22 * * * * ME18.5 17000 10700 4800 1200 ME15 100 200 200 200

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

ME18.5

ME15


30 60 120 180

ME22 4188 19840 41760 * ME18.5 1310 3300 1500 3200 ME15 600 480 380 410

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP) ME22

ME18.5

ME15



4.5 Viscosity of Rubberised Bitumen Blends with 50% Ambient Car/50% Ambient

Truck Tyres

Figure 4: Viscosity of Rubberised Bitumen Blends with 50% Ambient Car and 50% Ambient Truck Tyres


30 60 120 180

VE22 1156 4480 4996 * VE18.5 3300 2950 7500 1350 VE15 715 1032 1152 1194

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP) VE18.5

VE15

VE22


30 60 120 180

VE22 * * * * VE18.5 2240 2320 9167 10450 VE15 889.5 1120 1690 1723

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(min

s)VE18.5

VE15


30 60 120 180

ME22 7000 65000 30000 35000 ME18.5 7310 8890 730 590 ME15 450 280 280 240

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP) ME18.5

ME15

ME22


30 60 120 180

ME22 1000 15300 4600 9100 ME18.5 540 570 950 930 ME15 830 1640 890 710

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

ME22

ME18.5

ME15



4.6 Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre

Figure 5: Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre


30 60 120 180

VE22 825 1633 7838 9238 VE18.5 2200 1500 2200 2000 VE15 270 340 320 350

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP) VE22

VE18.5

VE15


30 60 120 180

VE22 4288 5770 10627 10850 VE18.5 1485 4190 4480 4510 VE15 858 1108 1581 1790

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP) VE22

VE18.5

VE15


30 60 120 180

ME22 8600 28500 25900 22100 ME18.5 1750 1620 1520 1320 ME15 260 460 460 270

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

ME22

ME18.5

ME15


30 60 120 180

ME22 14700 29400 31000 7900 ME18.5 2100 2240 3815 5370 ME15 240 530 530 390

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

ME22

ME18.5

ME15


4.7 Penetration and Softening Point Values Results from the penetration and softening point tests, carried out on samples taken at the end of 180 minutes blending, are summarised below.

Table 9: Penetration (dmm) at 25oC



TESTING 1 ca

r

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

Venezuelan 40/60 36 32 54 39 30 34 31 30 - - - -

Venezuelan 100/150 69 54 67 61 - 41 52 66 - - 41 52

Middle East 40/60 44 41 44 36 35 70 47 46 - 81 43 41

Middle East 100/150 63 77 91 71 74 124 80 91 - 69 57 56

Table 10: Ring and Ball Softening Point (oC)



TESTING 1

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

car

tru

ck

50

:50

car

/tru

ck

cryo

gen

ic

Venezuelan 40/60 67.5 67.3 52.2 64.5 72 76.5 72 70.2 - - - -

Venezuelan 100/150 59 59 59.8 57 - 71 61 63 - - 74 65.4

Middle East 40/60 60.4 68.4 64 63 67.6 62.8 64 63.8 - 63 68.6 70

Middle East 100/150 54.8 58.2 51 55 55.6 52.4 55.8 51.4 - 65.2 64 61.6

Note: - indicates test not carried out since viscosity criteria were not achieved

4.8 The Best Blends Originally, three ‘best’ blends were to be selected from the 48 rubberised bitumen blends; however, it was subsequently agreed with WRAP and their consultant that six blends would be selected. The viscosity test results were compared against target viscosity values between 1500cP – 5000cP. In addition one would expect the viscosity to rise as the rubber dissolved in the bitumen and would stabilise when solution was essentially complete. Where this did not occur in the required time, the viscosity fell with time, or was outside the permitted range, this bitumen/rubber combination was not considered favourably. It should be noted that the sample size of the test specimen and lack of complete homogeneity in mixing could lead to outlying results. Subsequently, based upon the viscosity criterion and these other considerations, the following blends were selected for further assessment in TASK 6:

ME 40/60 + 18.5% ambient car

VE 40/60 + 18.5% ambient car

VE 40/60 + 15% ambient truck

ME 40/60 + 18.5% cryogenic car

VE 100/150 + 18.5% cryogenic car

VE 40/60 + 18.5% cryogenic car.


5.0 Task 6: Assessment of Effect of Rubber Particle Size Task 6 involved a further assessment of the effect of rubber particle size on the blend properties of the six most suitable blends identified at the end of Task 4. These were:







Since the adopted “standard” rubber particle size was already at the coarser side of the grading envelope specified in the WRAP TID, this task focussed on the effect of using a finer rubber particle size. A similar suite of testing (as in Task 4) was carried out on a finer rubber particle size (see Figure 6); viscosity measurement of each rubberised bitumen blend at 177oC and 20 RPM was carried out by using both Brookfield and Bohlin viscometers on samples taken at 30, 60, 120 and 180 minutes during blending/mixing. In this exercise, both maximum (peak-ASTM D6114) and “stabilised” (AASHTO T316) viscosity values were recorded as well as penetration test (EN 1426) and softening point test (EN 1427) results.

Figure 6: Grading of Crumb Rubbers (Finer Size)

0

10

20

30

40

50

60

70

80

90

100

0.01 0.1 1 10

Sieve Size (mm)

%Pa

ssin

g

100% Car Tyre [Fine]

100% Truck Tyre [No.30]

100% Cryogenic [25/75 Blend0.4-0.63mm/0.63-1.4mm]

Upper TID Limit

Lower TID Limit

Results from this suite of assessment are presented in the sections below; the viscosity results for blends using “standard” size rubber are also reproduced in this section to ease the comparison.


5.1 Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre (Fine)

Figure 7: Viscosity of Rubberised Bitumen Blends with 100% Ambient Car Tyre


Blending Time (minutes) 30 60 120 180

VE18.5-standard 2140 2341 3827 4630

VE18.5-fine-peak 9560 11590 10340 840

VE18.5-fine-stabilised 4658 5933 6102 6750

Brookfield Viscometer

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

VE18.5-standardVE18.5-fine-peakVE18.5-fine-stab



VE18.5-standard * 3396 4932 5025

VE18.5-fine-peak 4120 4868 4260 3280

VE18.5-fine- stabilised 1945 2343 1921 2126

BohlinViscometer

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)




ME18.5-standard 1310 3300 1500 3200

ME18.5-fine-peak 7950 9300 3800 1900

ME18.5-fine- stabilised 2700 3200 2250 1500

ME18.5s-(repeat) 1292 3557 5140 7098

ME16-standard 3210 3201 1350 2547


100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

ME18.5-standardME18.5-fine-peakME18.5-fine-stabME18.5-standard reME16-standard



ME18.5-standard * * * *

ME18.5-fine-peak 2750 2910 2250 1450

ME18.5-fine- stabilised

1910 2120 1450 800

BohlinViscometer

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

ME18.5-f ine-peak

ME18.5-f ine-stab



5.2 Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre (Fine)

Figure 8: Viscosity of Rubberised Bitumen Blends with 100% Ambient Truck Tyre Venezuelan 40/60 BITUMEN


30 60 120 180

VE15-standard

1148 1194 2218 3340

VE15-fine-peak

7800 6300 19500 3420

VE15-fine- stabilised

2500 3500 11500 4128

VE15r-standard (repeat)

2112 10120 37690 *

VE15.5-standard

6350 * * *


100

1000

10000

100000

0 50 100 150 200Blending Time (mins)

Visc

osity

(cP)

VE15-fine-peak

VE15-fine-stab

VE15-standard

VE15-standard rep



30 60 120 180

VE15-standard 1551 1686 2445 2478

VE15-fine-peak 2300 2000 3000 2750

VE15-fine- stabilised 1400 1500 1750 950

BohlinViscometer

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

VE15-standard

VE15-fine-peak

VE15-fine-stab



5.3 Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre (Fine)

Figure 9: Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre


BROOKFIELD 30 60 120 180 VE18.5-standard 2200 1500 2200 2000

VE18.5-fine-peak 2150 8000 4350 630



100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)



BOHLIN 30 60 120 180 VE18.5-standard 1800 1550 1700 1840

VE18.5-fine-peak 2100 3800 2800 1550


BohlinViscometer

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)

VE18.5-standardVE18.5-f ine-peakVE18.5-f ine-stab


BROOKFIELD 30 60 120 180 VE18.5-standard 1485 4190 4480 4510

VE18.5-fine-peak 3850 7652 10263 15340



100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)



BOHLIN 30 60 120 180 VE18.5-standard 2118 2445 2954 3958

VE18.5-fine-peak 2143 3235 5265 4420


BohlinViscometer

100

1000

10000

100000

0 50 100 150 200


Visc

osity

(cP)



Middle East 40/60 BITUMEN BROOKFIELD 30 60 120 180 ME18.5n-standard 3590 2380 1180 680

ME18.5s-standard (repeat)

2100 2240 3815 5370

ME18.5-fine-peak 1167 4745 7570 11720



100

1000

10000

100000


Visc

osity

(cP)

ME18.5n-standard

ME18.5-f ine-peak

ME18.5-f ine-stab

ME18.5s-standard

Middle East 40/60 BITUMEN BOHLIN 30 60 120 180 ME18.5s-standard (repeat)

2050 2650 2966 3123

ME18.5-fine-peak 1191 3043 4320 4492


BohlinViscometer

100

1000

10000

100000


Visc

osity

(cP)

ME18.5n-standard

ME18.5-f ine-peak

ME18.5-f ine-stab

ME18.5s-standard

Figure 9 (Cont.): Viscosity of Rubberised Bitumen Blends with 100% Cryogenic Car Tyre

5.4 Penetration and Softening Point Values Results from the penetration and softening point tests, carried out on blends containing “fine” rubber taken at the end of 180 minutes blending, are summarised below. The results for blends using “standard” size rubber are also reproduced in this section to ease the comparison.

Table 11: Penetration and Softening Point Test Results

Penetration (dmm) at 25oC Softening Point (oC) Sample Fine Standard Fine Standard

VE 40/60 + 18.5% ambient car 36 30 68 72 ME 40/60 + 18.5% ambient car 36 35 67 67.6 VE 40/60 + 15% ambient truck 58 32 64.5 67.3 ME 40/60 + 18.5% cryogenic car 35 46 67 63.8 VE 100/150 + 18.5% cryogenic car 82 66 51 63 VE 40/60 + 18.5% cryogenic car 38 30 64 70.2

5.5 Selected Blends

Finer rubber particle size, having larger surface area than the ‘standard’ size, may yield greater reaction and interaction with base bitumen and ultimately lead to the increased viscosity observed in the above test results. Parallel testing using both viscometers also highlighted an important finding, that different testing equipment can produce different results. For consistency, the analysis and decision making were subsequently based on the viscosity results obtained by Brookfield viscometer only. All blends incorporating a finer rubber particle size exceeded the maximum viscosity values of 5000 cP; consequently none of them was considered suitable for adoption in the next stage of assessment. Therefore, the


options were limited to those incorporating the “standard” size rubber as used in Task 4; specifically, the following blends were considered to have the most desirable properties:

VE 40/60 + 18.5% ambient car;


The third blend was a choice between VE 40/60 + 15% ambient truck and ME 40/60 + 18.5% ambient car. However, there was speculation as to whether the content of truck tyre rubber in VE 40/60 bitumen could be increased slightly and/or whether the ambient car tyre rubber in ME 40/60 would yield reproducible results. Consequently, upon a request from WRAP, three further blends were manufactured and tested:

VE 40/60 + 15.5% ambient truck (i.e. a slight increase in rubber content);

ME 40/60 + 18.5% ambient car (i.e. repeat testing);

ME 40/60 + 16% ambient car (i.e. a reduction in rubber content)

As evidenced from the results presented in Figures 7 and 8, the VE 40/60 + 15.5% ambient truck immediately showed viscosity greater than 5000 cP within 30 minutes blending, and the ME 40/60 + 18.5% ambient car exceeded the 5000 cP viscosity at 180 minutes blending time. The viscosity test result for the blend containing ME 40/60 + 16% ambient car remained within the 1500 – 5000 cP range. Following a further discussion with WRAP, the following three best blends (all incorporating the “standard” size rubber as used in Task 4), were selected:

ME 40/60 + 16% ambient car



VE 40/60 and ME 40/60 denote Venezuelan and Middle Eastern bitumens grade 40/60 respectively. The above blends were subsequently subjected to a further suite of testing in TASK 5 and TASK 7. 6.0 Task 5: Comprehensive Testing of the Most Suitable Blends The following three rubberised bitumen blends were selected at the end of TASK 6:




The above three blends were subsequently subjected to the following comprehensive tests:

determination of Ageing Index (BBA HAPAS SG4 protocol);

penetration testing at 4oC (EN 1426);

softening point testing (EN 1427);

rotational viscosity at 177oC (ASTM D6114); and,

resilience at 25oC (EN 13880-3).

Results obtained from this assessment are detailed in Table 12 below.


Table 12: Comprehensive Test Results

Binder Sample

Tests

ME 40/60 Base Bitumen

VE 40/60 Base Bitumen

16% Ambient Car - ME40/60

18.5% Ambient Car - VE40/60

18.5% Cryogenic Car - VE40/60

Recommended values for rubberised bitumen with 50/70 base bitumen

Ageing Index 5.28 7.31 3.16 2.16 2.18 < 15

G* at 25oC (Unaged), Pa 6.59E+05 3.98E+05 6.64E+05 8.33E+05 7.67E+05 - G* at 25oC (After RTFOT and HiPAT), Pa 3.48E+06 2.91E+05 2.10E+06 1.80E+06 1.67E+06 -

Penetration (4oC, 200g, 60s), dmm - - 29 20 19 > 15

Softening Point, oC 51 50 63 72 70.2 > 54 Rotational Viscosity (177oC), cP 80 109 2550 4630 4510 1500-5000

Resilience (25oC), % 0 5 13 29 26 >25

The above results show that ageing the bitumen progressively increased the G* value, which is a complex shear (or stiffness) modulus at a given temperature and loading rate; hence indicating a stiffened or hardened binder. A road paving bitumen typically used in the UK is expected to have an Ageing Index (the ratio between G*HiPAT and G*Unaged at 0.4 Hz) of less than 15; generally the lower the value, the less susceptible the binder would be expected to be to ageing. The level of increase in softening point, viscosity and resilience, and the reduced penetration values, can be linked to the increase in rubber content in these blends. With respect to the typical target reference values for rubberised bitumen incorporating 50/70pen base bitumen, the above test results indicate that:

The selected blends satisfied the requirements for ageing index, softening point and

viscosity;

Both rubberised blends with Venezuelan bitumen met the criterion for resilience, however,

the Middle Eastern blend fell short of the recommended resilience value although remained

higher than that of the base bitumen.



7.0 Task 7: EN 14023 Tests on the Three Best Performing Blends The three preferred rubberised bitumen blends selected at the end of TASK 6 were further assessed in accordance with EN 14023 test requirements, excluding the elastic recovery tests. Test results are presented below.

Table 13: EN 14023 Test Results

Binder Sample

Tests Methods 16% Car - ME40/60

18.5% Car - VE40/60

18.5% Cryogenic - VE40/60

Penetration (25oC, 100g, 5s), dmm EN 1426 35 30 30 Softening Point, oC EN 1427 63 72 70.2 Cohesion - Vialit pendulum, J/cm2 SHW Clause 939 1.9 1.4 1.5 Temperature at Peak Cohesion, oC 50 50 53 Temperature range at cohesion above 0.5 J/cm2, oC >55 50 50 RTFOT, Change of mass, % EN 12607-1 -0.16 -0.23 -0.21 RTFOT, Penetration (25oC, 100g, 5s), dmm EN 1426 24 24 27 RTFOT, retained penetration, % - 69% 80% 90% RTFOT, softening point, oC EN 1427 65 79.5 73 RTFOT, increase in softening point, oC - 2 7.5 2.8 Flash Point, oC EN ISO 2592 >300 298 >300 Fraass breaking point, oC EN 12593 -12 -8 -11 Plasticity range, oC Sub-clause 5.1.9 75 80 81.2

Storage stability, penetration at the top, dmm EN 13399, EN 1426 39 35 30

Storage stability, penetration at the bottom, dmm EN 13399, EN 1426 41 33 32

Storage stability, difference in penetration, dmm - 2 2 2

Storage stability, softening point at the top, oC EN 13399, EN 1427 56.5 71.5 63

Storage stability, softening point at the bottom, oC EN 13399, EN 1427 67 76 77

Storage stability, difference in softening point, oC - 10.5 4.5 14 HiPAT, Penetration (25oC, 100g, 5s), dmm EN 1426 22 18 24 HiPAT, retained penetration, % - 63% 60% 80% HiPAT, softening point, oC EN 1427 71 90 83.5 HiPAT, increase in softening point, oC - 8 18 13.3

The increased softening point and the reduced penetration values after the short-term ageing test (RTFOT) can be linked to the increased rubber content of these blends. A further reduction in penetration and increase in softening point of the residual binders after the long-term ageing test (HiPAT) is also found, consistent with the trend observed on the RTFOT samples. These results generally suggest improved properties of the rubberised bitumens over those expected for 40/60pen bitumen. Cohesion is commonly accepted as an indicator of a bitumen’s ability to resist tensile stress under rapid impact loading, such as due to skidding or vehicle braking, and hence the ability to retain aggregate intact within an asphalt material (resist fretting). The higher the cohesion value, the better the aggregate retention. Another related parameter, the temperature range over which the cohesion is at or above 0.5 J/cm2 (TR) is commonly considered to be an indicator of service temperature range; the higher the value, the wider the service temperature range of a bituminous binder. The peak cohesion values and the temperature range over which cohesion values are at or above 0.5 J/cm2 for the rubberised bitumen blends are much higher than (at least double) those of typical 40/60pen bitumen, indicating significantly improved cohesion characteristics for these blends, implying an improved ability to retain aggregate. Fraass breaking point (FBP) is the temperature at which thin bitumen films start to show cracks under a small deflection. The lower the FBP value, the better the low temperature performance of the binder. The test results show that the rubberised bitumen blends had FBP values not greater than -8oC. This is similar to that measured for a typical UK 40/60 penetration binder (around -7oC), suggesting that the addition of rubber does not have a detrimental effect on the low temperature properties of the binder. The plasticity range is the numerical difference between the softening point and Fraass breaking point, which is sometimes considered as the range of service temperatures that a binder may be expected to perform satisfactorily over. The plasticity ranges of the rubberised bitumen blends (i.e. 75oC or higher) will meet the requirements for EN14023 polymer modified binder Classes 2, 3 and 4, and are considered wider than those found typically for a 40/60pen bitumen (i.e. around 60oC). This suggests the rubberised bitumen blends can be expected to have a relatively wide service temperature range. As with any other bituminous material, rubberised bitumen will release combustible fumes when heated to sufficiently high temperatures. The flash point provides an indication of the temperature at which a heated bituminous sample will instantaneously flash in the presence of an open flame. The flash point values of the rubberised bitumen blends comply with those generally specified for polymer modified binders, i.e. higher than 250oC. Thus the risk that these blends could cause a fire during material production can be expected to be no greater than that for polymer modified bitumens. However, the blends were observed to release visibly thick white fume with strong odour during laboratory testing, specifically when the test temperature was increased above 200oC. It is recommended that the significance of the presence of this fume is assessed in more detail. The results on the stability of the blends during hot storage suggest only a marginal difference in the penetration value of the top and bottom part of the storage container, whilst there is up to a 25% difference in the softening point values. Whilst the latter difference is significantly greater than that expected from 40/60pen bitumen or some polymer modified bitumens, the results are considered not unusual and consistent with those reported in the USA for similar materials. Rubberised bitumen blends are known to be susceptible to phase separation and/or devulcanization (thinning due to rubber degradation) after prolonged hot storage.


8.0 Workability Assessment A screening assessment on twelve number RA mixtures (six surface course and six binder course materials) and four sets of control asphalt mixtures was carried out on mixtures prepared in the laboratory using bespoke workability tests, developed by SW. This workability assessment aimed to establish mixture characteristics and performance; the results are presented in the sections below. Porous Asphalt and Stone Mastic Asphalt were selected for use as the control surface course and binder course mixtures, respectively, whereas the rubberised asphalts adopted the composition (grading and binder content) normally used for open and gap graded rubberised asphalts, respectively. Due to the limited time available to complete this study, a comprehensive mixture design exercise was not carried out, for example determination of VMA (voids in mineral aggregate); instead, designs typically adopted for these asphalt mixtures were used. The gradings are presented in Figure 10 and the respective binder contents adopted are presented in Table 15. It should be noted, however, that the rubberised asphalt gradings exclude the grading of filler additives (around 2%).

Figure 10: Asphalt Mixture Gradations

0

10

20

30

40

50

60

70

80

90

100

0.01 0.1 1 10 100

Sieve Size (mm)

Cum

ulat

ive

% P

assi

ng

Control PAARSC-GS/GR

0

10

20

30

40

50

60

70

80

90

100

0.01 0.1 1 10 100Sieve Size (mm)

Cum

ulat

ive

%Pa

ssin

g Control SMAARBC-GRARBC-LS

Control PA BS Sieve Size (mm) Range Target

ARSC - GS/GR

14 100 100 100 10 90 – 100 95 95 4 - - 20 2 5 – 25 9 15 0.063 2 – 10 4 3

Control SMA ARBC BS Sieve Size (mm) Range Target GR LS 20 100 100 100 100 14 90 – 100 92 93 95 10 35 – 60 42 63 66 6.3 23 – 35 31 - - 2 17 – 29 25 17 14 0.063 8 – 13 10 4 4

The four control asphalt mixtures are referenced by Asphalt type-Aggregate type-Binder penetration, for example; PA-GS-125pen indicates porous asphalt (PA) with Gritstone aggregate (GS) and a binder with a 125 penetration (i.e. 100/150); and SMA-GR-50pen indicates a Stone Mastic Asphalt (SMA) with Granite aggregate (GR) and a binder with a penetration of 50 (i.e. 40/60). The twelve rubberised asphalt mixtures are referenced by applicable pavement layer-Aggregate type-Blend type. For example ARSC-GR-1 indicates an Asphalt rubber in the surface course layer (ARSC) containing Granite aggregate (GR), all mixed according to Blend type 1. ARBC-LS-2 indicates an Asphalt rubber in the binder course layer (ARBC) containing Limestone aggregate (LS), all mixed according to Blend type 2. The three rubberised bitumen blends were used, which were those selected at the end of TASK 6:

VE 40/60 + 18.5% ambient car (Blend 1)

VE 40/60 + 18.5% cryogenic car (Blend 2)

ME 40/60 + 16% ambient car (Blend 3)


Specifically the blend types are presented in more detail below:

Blend 1: 81.5% VEN40/60 + 9.25% ambient car (dust 07/140I) + 9.25% ambient car

(coarse 07/140H)

Blend 2: 81.5% VEN40/60 + 9.25% cryogenic car 07/058AA (dust) + 9.25% cryogenic car

(coarse)

Blend 3: 84% ME40/60 + 8% ambient car (dust 07/140I) + 8% ambient car (coarse

07/140H)

The respective references for the four control and twelve rubberised asphalt mixtures are presented in Table 14.

Table 14: Sample References

SURFACE COURSE

Control Bitumen

Rubberised Bitumen Blend 1


Rubberised Bitumen Blend 3 Aggregate

- VE 40/60 + 18.5% ambient car



Gritstone PA-GS-125pen ARSC-GS-1 ARSC-GS-2 ARSC-GS-3

Granite PA-GR-125pen ARSC-GR-1 ARSC-GR-2 ARSC-GR-3

BINDER COURSE

Limestone SMA-LS-50pen ARBC- LS-1 ARBC- LS-2 ARBC- LS-3

Granite SMA-GR-50pen ARBC- GR-1 ARBC- GR-2 ARBC- GR-3

8.1 Workability of Loose Mixture The “viscosity” of loose coated RA mixtures was assessed over a period of 10 minutes of mixing and represented as mixture resistance to the mixing torque applied (measured and recorded during high temperature mixing) per unit weight. A suitably instrumented asphalt mixer (pictured below) was used throughout the testing. The workability test data and the results are summarised in Table 15 and Figures 12 and 13.


Figure 11: Asphalt Mixer

Table 15: Workability Test Parameter

No Type Binder content Fibre Target Mixing

Temperature Actual Mixing Temperature

1 PA-GS-125pen* 4.5% 0.3% 140 +/- 5 137.0 2 PA-GR-125pen* 4.5% 0.3% 140 +/- 5 140.0 3 SMA-LS-50pen* 5.8% 0.3% 165 +/- 5 160.0 4 SMA-GR-50pen* 5.8% 0.3% 165 +/- 5 170.0 5 ARSC-GS-1 9.0% - 165 +/- 5 169.6 6 ARSC-GS-2 9.0% - 165 +/- 5 168.2 7 ARSC-GS-3 9.0% - 165 +/- 5 160.5 8 ARSC-GR-1 9.0% - 165 +/- 5 175.0 9 ARSC-GR-2 9.0% - 165 +/- 5 168.0 10 ARSC-GR-3 9.0% - 165 +/- 5 170.7 11 ARBC- LS-1 8.7% - 165 +/- 5 171.0 12 ARBC- LS-2 8.7% - 165 +/- 5 166.0 13 ARBC- LS-3 9.0% - 165 +/- 5 171.5 14 ARBC-GR-1 8.7% - 165 +/- 5 168.0 15 ARBC-GR-2 8.7% - 165 +/- 5 167.0 16 ARBC-GR-3 9.0% - 165 +/- 5 167.5

Note: *Control asphalt mixtures Figures 12 and 13 below give a graphical representation of the required mixing torque for both surface course and binder course mixtures.


Figure 12: Workability of Loose Surface Course Mixtures

Surface Course

0.00

0.10

0.20

0.30

0.40

0.50

0.60

PA-GS-125pen

ARSC-GS-1

ARSC-GS-2

ARSC-GS-3

PA-GR-125pen

ARSC-GR-1

ARSC-GR-2

ARSC-GR-3

Asphalt Type

Nor

mal

ised

Tor

que

(Nm

/kg)

Less workable

More workable

Lower resistance to mixing torque per unit weight is considered to be an indication of better workability. From the graph above, it can be seen that all gritstone and granite ARSC mixtures require a lower mixing torque compared to their respective control (PA) mixtures. This suggests that the mixtures have improved workability compared with their respective control mixtures. Although rubberised bitumen is more viscous than the control bitumen, in the respective mixtures it has been found, in this case, to have improved workability, probably because of the higher mixing temperature and higher binder content. The graph above also shows that in the case of gritstone mixes, workability gradually increases from Blend 1 through to Blend 3 whereas the workability of the granite mixes is roughly the same.

Figure 13: Workability of Loose Binder Course Mixtures

Binder Course

0.000

0.100

0.200

0.300

0.400

0.500

0.600

SMA-LS-50pen

ARBC-LS-1

ARBC-LS-2

ARBC-LS-3

SMA-GR-50pen

ARBC-GR-1

ARBC-GR-2

ARBC-GR-3

Asphalt Type

Nor

mal

ised

Tor

que

(Nm

/kg)

Less workable

More workable

Figure 13 above shows that the binder course mixtures ARBC-LS-1 and ARBC-LS-2 require a higher mixing torque than the control SMA mixture and hence have a lower workability; in all other cases, the rubber asphalt mixtures were more workable then the respective controls. Limestone Blend 1 had the least workability followed by Blend 2 and Blend 3 respectively, while granite Blends 1 and 2 had the same level of workability but lower than that of Blend 3. Following the completion of the mixing process (i.e. after 10 minutes), the loose samples were “held” at the mixing temperature for a prolonged duration up to 120 minutes, in order to simulate the possible effects of mix storage (e.g. in the haul truck, during the delivery period from the mixing plant to site). Samples were taken after 10, 30, 60, 90 and 120 minutes from the start of testing, compacted using a Marshall hammer with 50 blows per


face and then subjected to an assessment of mixture volumetrics (bulk and maximum density), stiffness and strength over the holding periods.

8.2 Bulk Densities The bulk densities (EN 12697-6) of the RA mixtures as well as the control mixtures were determined for samples compacted after 10, 30, 60, 90 and 120 minutes holding time at the respective mixing temperatures. For the surface courses, the bulk density was determined by dimensions, whilst that for the binder courses was determined by the sealed method (using self-adhesive aluminium foil). The results of these tests have been summarised in the figures below.

Figure 14: Bulk Densities of Surface Course Mixtures

Surface Course (Gritstone)

1.95

2.00

2.05

2.10

2.15

2.20

2.25

0 20 40 60 80 100 120 140

Mixing/Holding Time (min)

Bulk

Den

sity

(Mg/

m3 )

PA-GS-125pen ARSC-GS-1ARSC-GS-2 ARSC-GS-3

Surface Course (Granite)

1.95

2.00

2.05

2.10

2.15

2.20

2.25

0 20 40 60 80 100 120 140


Bulk

Den

sity

(Mg/

m3 )

PA-GR-125pen ARSC-GR-1ARSC-GR-2 ARSC-GR-3

The results show that although there are slight variations of bulk density with mixing/holding time, the average bulk density of each mix generally remains constant throughout the period of “holding” at high temperature. The surface course control mixtures had the lowest bulk densities for each type of aggregate. The gritstone mixtures showed an initial decrease in bulk density value with holding time for Blend types 1 and 2, with Blend 1 having the highest bulk density and Blend 3 the lowest. The granite mixtures did not exhibit a clear variation of bulk density with blend type although Blends 1 and 3 showed an initial decrease in bulk density and Blend 3 generally had the highest average bulk density and was more consistent.

Figure 15: Bulk Densities of Binder Course Mixtures

Binder Course (Limestone)

2.10

2.15

2.20

2.25

2.30

2.35

2.40

0 20 40 60 80 100 120 140

Mixing/Holding Time (Min)

Bul

k D

ensi

ty (M

g/m3 )

SMA-LS-50pen ARBC- LS-1

ARBC- LS-2 ARBC- LS-3

Binder Course (Granite)

2.10

2.15

2.20

2.25

2.30

2.35

2.40

0 20 40 60 80 100 120 140

Mixing/Holding Time (Min)

Bul

k D

ensi

ty (M

g/m3 )

SMA-GR-50pen ARBC-GR-1

ARBC-GR-2 ARBC-GR-3


With regard to the binder course, all the mixtures (except limestone Blend 3) showed a slight decrease in the value of bulk density initially. The ARBC mixtures showed a general increase in bulk density with mixture blend type i.e. Blend 1 had the lowest bulk density value while Blend 3 had the highest bulk density value for both limestone and granite mixtures.

8.3 Maximum Densities

Maximum density measurements (EN 12697-5) of the RA mixtures and the control mixtures were carried out on samples manufactured after 10 minutes and 120 minutes of mixing/holding time. Results of these tests have been summarised in the figures below.

Figure 16: Maximum Densities of Surface Course Mixtures

Surface Course

2.15

2.20

2.25

2.30

2.35

2.40

2.45

2.50

2.55

PA-GS-125pen

ARSC-GS-1

ARSC-GS-2

ARSC-GS-3

PA-GR-125pen

ARSC-GR-1

ARSC-GR-2

ARSC-GR-3

Asphalt Type

Max

imum

Den

sity

(Mg/

m3 )

10 Minutes120 Minutes

Figure 17: Maximum Densities of Binder Course Mixtures

Binder Course

2.15

2.20

2.25

2.30

2.35

2.40

2.45

2.50

2.55

SMA-LS-

50pen

ARBC-LS-1

ARBC-LS-2

ARBC-LS-3

SMA-GR-

50pen

ARBC-GR-1

ARBC-GR-2

ARBC-GR-3

Asphalt Type

Max

imum

Den

sity

(Mg/

m3 )

10 Minutes120 Minutes

Maximum densities of the mixtures after 10 minutes of mixing/holding were not significantly different from those after 120 minutes of mixing/holding.


8.4 Stiffness

An assessment of the effect of holding times on mixture stiffness at 20oC was carried out, using the Indirect Tensile Stiffness Modulus (ITSM) test procedure in the Nottingham Asphalt Tester (NAT), to BS EN 12697-26 Annex C. The test arrangement is illustrated in Figure 18, and the results of this assessment are presented in Figures 19 and 20 below.

Figure 18: Stiffness Test Arrangement

Figure 19: Surface Course Stiffness Assessment

Stiffness Vs Mixing/Holding Time (Gritstone)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 20 40 60 80 100 120 14


Stif

fnes

s at

20o C

(MPa

)

0


Stiffness Vs Mixing/Holding Time (Granite)

0

500

1000

1500

2000

2500

3000

3500

4000

4500

5000

0 20 40 60 80 100 120 140


Stif

fnes

s at

20o C

(MPa

)


Surface course mixtures generally showed an increase in mixture stiffness with increase in mixing (holding) time. The control mixtures showed significantly lower values of stiffness in comparison with the rubberised asphalt mixtures. This indicates improved load spreading ability of the RA mixtures compared with that of the porous asphalt control samples.


Figure 20: Binder Course Stiffness Assessment

Stiffness Vs Mixing/Holding Time (Limestone)

0

1000

2000

3000

4000

5000

6000

7000

0 20 40 60 80 100 120 14


Stiff

ness

at 2

0o C (M

Pa)

0

SMA-LS-50pen ARBC- LS-1

ARBC- LS-2 ARBC- LS-3

Stiffness Vs Mixing/Holding Time (Granite)

0

1000

2000

3000

4000

5000

6000

7000

0 20 40 60 80 100 120 14


Stiff

ness

at 2

0o C (M

Pa)

0

SMA-GR-50pen ARBC-GR-1ARBC-GR-2 ARBC-GR-3

Binder course mixtures also showed a small increase in mixture stiffness with increase in mixing (holding) time. Apart from ARBC-LS-1, mixture stiffness values were generally similar for all the Stone mastic asphalt (SMA). Overall, the holding time does not lead to a reduction in stiffness. It should be noted that all RA mixtures exceeded the recommended minimum stiffness value of 1700 MPa.

8.5 Tensile Strength An assessment of the effect of holding times on mixture tensile strength at 25oC was carried out using the Indirect Tensile Test at a loading rate of 50mm/minute, to EN 12697-23. The test arrangement is shown in Figure 21 and the results of this assessment are presented in Figures 22 and 23.

Figure 21: Tensile Strength Test Arrangement


Figure 22: Surface Course Tensile Strength

Tensile Strength Vs Mixing/Holding Time (Gritstone)

0

200

400

600

800

1000

1200

1400

0 20 40 60 80 100 120 140


Tens

ile S

treng

th a

t 25

o C (k

Pa)


Tensile Strength Vs Mixing/Holding Time (Granite)

0

200

400

600

800

1000

1200

1400

0 20 40 60 80 100 120 140


Tens

ile S

tren

gth

at 2

5o C

(kP

a)


Figure 22: Binder Course Tensile Strength

Tensile Strength Vs Mixing/Holding Time (Limestone)

0

500

1000

1500

2000

2500

0 20 40 60 80 100 120 140


Tens

ile S

tren

gth

at 2

5o C

(kP

a)

SMA-LS-50pen ARBC- LS-1ARBC- LS-2 ARBC- LS-3

Tensile Strength Vs Mixing/Holding Time (Granite)

0

500

1000

1500

2000

2500

0 20 40 60 80 100 120 140


Tens

ile S

tren

gth

at 2

5o C

(kPa

)

SMA-GR-50pen ARBC-GR-1

ARBC-GR-2 ARBC-GR-3

Apart from samples manufactured using granite aggregate, surface course and binder course mixtures did not show a significant change in tensile strength with increase in mixing (holding) time. The surface course control mixtures showed significantly lower values of tensile strength; those of the RA mixtures were 2 – 3 times higher. The tensile strength of gritstone Blends 2 and 3 initially decreases then either stabilises or increases. Blend 1 shows the most consistent results of the gritstone blends. Regarding the binder course, limestone Blend 2 had tensile strength values similar to the control mixtures’ tensile strength values, with those of Blend 3 being slightly lower and those of Blend 1 more significantly so. The low tensile strength values of limestone Blend 1 are consistent with its low values of stiffness. Research by De Beer et al (1999) on tyre/pavement contact stresses reported horizontal stresses due to the tyres of B-747 aircraft between 260 and 500 kPa. The above laboratory test results from rubberised asphalt exceeded 600 kPa, indicating sufficiently good quality material.


9.0 Mechanical Assessment Following discussion of the results from the Task 8 testing with WRAP, the RA mixtures to be subjected to mechanical assessment have been summarised in Table 16 below. The description of these mixtures is consistent with that presented in Chapter 8.

Table 16: Mixtures for Mechanical Assessment

SURFACE COURSE

Control Bitumen



Rubberised Bitumen Blend 3 Aggregate

- VE 40/60 + 18.5% ambient car



Gritstone PA-GS-125pen ARSC-GS-1 ARSC-GS-2 ARSC-GS-3

Granite - ARSC-GR-1 - -

BINDER COURSE

Limestone SMA-LS-50pen ARBC- LS-1 ARBC- LS-2 ARBC- LS-3

Granite - - ARBC- GR-2 -

9.1 Specimen Manufacturing Five asphalt slabs (305 mm x 405 mm x 50 mm) of each mixture type were manufactured using a laboratory roller compactor. From those slabs, a total of three 200 mm diameter, fifteen 100 mm diameter and five 150 mm diameter cores were obtained for each mixture type. Previous work has highlighted the laboratory roller compactor as the method of laboratory compaction best suited to emulate site compaction. A 50 mm thickness was chosen as being representative of the thickness likely to be used in practice.

9.2 Test Methodology The following suite of mechanical assessment was carried out on the selected RA mixtures (four surface course and four binder course materials) and two sets of control asphalt mixtures (a control binder course and a control surface course material):

Voids at refusal to EN 13108-20;

Mixture volumetrics: bulk density to EN 12697-6, maximum density to EN 12697-5 and air

void contents to EN 12697-8,

Deformation resistance to BS 598-110 (small scale wheel tracker);

Crack resistance by semi circular bending test to prEN 12697-44;

Fatigue resistance to BBA HAPAS SG3/08/254 Annex A.13 (identical with DD ABF);

Retained stiffness (water sensitivity) to BBA HAPAS SG3/05/234 Annex A.2 – on surface

course RA samples only;

Binder drainage (basket method) to EN 12697-18;

Durability (Cantabro) test to EN 12697-17.

9.3 Mixture Volumetrics

Prior to carrying out the mechanical tests, an assessment of mixture volumetrics was carried out. This provided, to some extent, an early indication of any potential problems associated with workability. The bulk density of surface course mixtures was determined by dimensions and that of binder course mixtures by the sealed method.


The air void contents of each core subjected to the Percentage Refusal Density (PRD) test procedure was also determined. The initial and refusal air void contents were calculated using the initial dried bulk density and the refusal density respectively, determined in accordance with BS 598-104. A summary of the mixture volumetrics of the selected cores removed from the slabs are presented in Table 17 as follows.

Table 17: Summary of Mixture Volumetrics

Bulk Density Maximum Density Air Voids

(Mg/m3) (Mg/m3) (%) At Refusal*

Mixture Type

By Dimension Sealed By

Dimension Sealed Density (Mg/m3)

Voids (%)

PA-GS-125pen 2.073 2.530 18.1 2.426 4.1

ARSC-GS-1 2.223 2.414 7.9 2.363 2.1

ARSC-GS-2 2.187 2.408 9.2 2.374 1.4

ARSC-GS-3 2.093 2.450 14.6 2.354 3.9

ARSC-GR-1 2.145 2.315 7.3 2.221 4.1

SMA-LS-50pen 2.362 2.450 3.6 2.418 1.3

ARBC- LS-1 2.246 2.359 4.8 2.312 1.9

ARBC- LS-2 2.256 2.359 4.4 2.317 1.8

ARBC- LS-3 2.225 2.359 5.7 2.322 1.6

ARBC- GR-2 2.252 2.330 3.3 2.304 1.1

Note: Figures above are an average of four results except maximum densities which are an average of three results * ARSC and ARBC samples were tested at 177oC; PA samples were tested at 140oC and SMA samples were tested at 165oC. Refusal density testing is normally adopted to assess the risk of secondary compaction of dense asphalt mixtures. In this test, samples of field compacted asphalt are subjected to further compaction to refusal, in the laboratory, by means of a vibrating hammer. For dense asphalt mixtures, air voids at refusal above 1% is typically the minimum specified value. Very low air voids at refusal (e.g less than 1%) are generally considered to be an indication of mixture susceptibility to secondary compaction, leading to low resistance to rutting. In this case, all dense asphalt samples (SMA and ARBC) showed refusal air void contents greater than 1%, which was satisfactory. For porous and open graded mixtures (e.g. PA and ARSC), where mixture strength mostly relies on stone-to-stone contact, the voids at refusal would be expected to be higher. It is possible that aggregate degradation during refusal density testing may have taken place in ARSC-GS-1 and ARSC-GS-2 which led to air voids at refusal of around 1 to 2%.

9.4 Deformation Resistance

The Wheel Tracking Test (WTT) was carried out to BS 598-110, under the standard test conditions of a wheel load of magnitude 520 N, which moves backwards and forwards in simple harmonic motion at 42 passes per minute (21 cycles per minute). The test arrangement is illustrated in Figure 23.


Figure 23: Wheel Tracking Test Arrangement

In this test, wheel tracking is continued until 45 minutes has elapsed, or a 15mm rut has developed, and the permanent deformation is recorded at 5 minute intervals. The WTT was evaluated in accordance with SHW 942 Class 3 for the surface course and PD 6691 Class 2 for the binder course, i.e. at a test temperature of 60oC. Details of the WTT results are presented in Appendix F; a summary is shown in Figure 24 below.

Figure 24: Wheel Tracking Data

Wheel Tracking Test (Surface Course)

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

PA-GS-125pen ARSC-GS-1 ARSC-GS-2 ARSC-GS-3 ARSC-GR-1

Asphalt Type

Rut

Dep

th (m

m) /

Rut

Rat

e (m

m/h

r)

WTT Rut Depth (mm)

WTT Rut Rate (mm/hr)



Figure 25: Crack Resistance Test Arrangement

Asphalt mixtures with very high stiffness may exhibit poorer resistance to crack propagation and/or thermal cracking. The thermal crack resistance was assessed by semi circular bending tests at a loading rate of 5mm/minute (0.085 mm/s) at 0oC, in accordance with prEN 12697-44; see Figure 25. Four samples of each of the mixtures were tested; the results are summarised in Figure 26.

From Figure 24 above, the surface course RA samples have a lower rut depth and rut rate compared to the control samples i.e. PA-GS-125pen. This indicates that the RA samples have better resistance to deformation. Both the control and surface course RA specimens meet the requirements for surfacing material for application on heavily stressed areas (SHW Clause 942 Class 3), i.e. having a rut depth and rut rate at 60oC testing of less than 7mm and 5mm/h respectively. For the binder course, only ARBC-GR-2 had better resistance to deformation compared to that of the control sample SMA-LS-50pen. It should be noted, however, that all binder course samples met the requirements of PD 6691 Class 2.

9.5 Crack Resistance

Wheel Tracking Test (Binder Course)

0.0

1.0

2.0

3.0

4.0

5.0

6.0

SMA-LS-50pen ARBC- LS-1 ARBC- LS-2 ARBC- LS-3 ARBC- GR-2

Asphalt Type

Rut

Dep

th (m

m) /

Rut

Rat

e (m

m/h

r)

WTT Rut Depth (mm)

WTT Rut Rate (mm/hr)

Note: Values shown are an average produced from 3 results

Figure 26: Crack Resistance Data

Surface Course

0

20

40

60

80

100

120

140

PA-GS-125pen ARSC-GS-1 ARSC-GS-2 ARSC-GS-3 ARSC-GR-1

Asphalt Type

Ave

rage

frac

ture

to

ughn

ess

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

1.60%

1.80%

Aver

age

stra

in a

t m

axim

um fo

rce

Average fracture toughness Kfc (N/mm3/2)Average Strain at maximum force εmax

Binder Course

0

20

40

60

80

100

120

140

160

SMA-LS-50pen ARBC- LS-1 ARBC- LS-2 ARBC- LS-3 ARBC- GR-2

Asphalt Type

Aver

age

fract

ure

toug

hnes

s

0.00%

0.20%

0.40%

0.60%

0.80%

1.00%

1.20%

1.40%

Aver

age

stra

in a

t m

axim

um fo

rce

Average fracture toughness Kfc (N/mm3/2)Average Strain at maximum force εmax

Except for ARSC-GR-1, the surface course RA samples had a higher average fracture toughness and average strain at failure than the PA-GS-125pen control sample. This implies that the ARSC samples have better flexibility and better resistance to low temperature cracking. Results for ARBC samples, however, show that the RA samples have lower average fracture toughness and strain at failure when compared to the control sample SMA-LS-50pen. The fracture toughness values of all samples tested are at least twice those reported by van Rooijen et al (2005) for a number of airfield asphalt pavement materials (incorporating different bitumen grades and polymer modifier); it should be noted, however, that the values reported by van Rooijen were obtained at 5oC.

9.6 Fatigue Resistance

Fatigue resistance was assessed in the NAT by using a controlled load Indirect Tensile Fatigue Test (ITFT) in accordance with the BBA HAPAS Document SG3/08/254 Annex A.13 (which is identical with DD ABF (1995), “Method for the determination of the fatigue characteristics of bituminous mixtures using indirect tensile fatigue”); the test arrangement is shown in Figure 27. The applied level of loading (hence, horizontal stress) was adjusted for individual specimens to generate lives to failure generally between 10 and 100,000 cycles. Due to limited time available, each set of data comprised ITFT results from only five 100 mm diameter core samples; the standard protocol (full procedure) requires twelve samples per test set.


Figure 27: Fatigue Test Arrangement

For the analysis of the ITFT data, initial tensile strains were determined from an analysis of the stress distribution within the sample during testing. A log-strain to log-life relationship was then plotted, forming a single, linear relationship. Figure 28 below shows the fatigue resistance of the specimen (both Control and RA materials). The characteristics of the corresponding fatigue regression lines are presented in Equations 1 – 10.

Figure 28: Resistance to Fatigue Cracking

Surface Course

100

1000

10000

10 100 1000 10000 100000Number of Cycles to failure

Tens

ile M

icro

stra

in

ARSC-GS-1

PA-GS-125pen

ARSC-GS-2

ARSC-GS-3

ARSC-GR-1

Control

Eq 1: PA-GS-125pen: ε = 3260 * Nf-0.345 R2 = 0.97

Eq 2: ARSC-GS-1: ε = 1857.3 * Nf-0.2329 R2 = 0.98

Eq 3: ARSC-GS-2: ε = 1517.5 * Nf-0.1981 R2 = 0.98

Eq 4: ARSC-GS-3: ε = 4399.3 * Nf-0.3914 R2 = 0.95

Eq 5: ARSC-GR-1: ε = 2363.5 * Nf-0.2369 R2 = 0.97


Binder Course

100

1000

10000

10 100 1000 10000 100000Number of Cycles to failure

Tens

ile M

icro

stra

in

SMA-LS-50pen

ARBC- LS-2

ARBC- LS-1

ARBC- LS-3

ARBC- GR-2

Control

Eq 6: SMA-LS- 50pen: ε = 1255.3 * Nf

-0.1841 R2 = 0.77

Eq 7: ARBC-LS-1: ε = 2527.7 * Nf-0.2439 R2 = 0.97

Eq 8: ARBC-LS-2: ε = 20190 * Nf-0.5222 R2 = 0.98

Eq 9: ARBC-LS-3: ε = 3448.3 * Nf-0.3049 R2 = 0.91

Eq 10: ARBC-GR-2: ε = 3104.1 * Nf-0.2705 R2 = 0.96

Note: ε = tensile microstrain (με) and Nf = number of cycles to failure. For the surface course samples, ARSC-GS-3, which has the lowest rubber content, has a fatigue resistance comparable to that of the control sample PA-GS-125pen. The rest of the surface course RA samples exhibit a better fatigue resistance in comparison.

Table 18: Failure Cycles at 200 Microstrain

Mix Type Cycles to Failure (Nf) at 200 microstrain

PA-GS-125pen 3,263

ARSC-GS-1 14,311

ARSC-GS-2 27,714

ARSC-GS-3 2,689

ARSC-GR-1 33,677

SMA-LS-50pen 21,531

ARBC- LS-1 32,885

ARBC- LS-2 6,884

ARBC- LS-3 11,368

ARBC- GR-2 25,270


A minimum value of 10,000 cycles to failure at a microstrain of 200 was adopted as a ‘pass/fail’ criterion for this study. Table 18 above shows that PA-GS-125pen, ARSC-GS-3 and ARBC-LS-2 do not meet this requirement. The longest fatigue lives at this level of microstrain are given by ARSC-GR-1 and ARBC-LS-1, for surface and binder courses respectively.

9.7 Retained Stiffness

The resistance to moisture damage of surface coarse samples only was also assessed, using retained stiffness as the criterion, in addition to the durability assessment presented later. A threshold value of retained “strength” of 70% has been suggested by some authorities, below which a mixture is deemed to be sensitive to water, but for the purposes of this research, a threshold value of 80% retained stiffness after a soaking regime was adopted as the criterion. The soaking regime was carried out by using the following procedure:

Each test specimen was initially pre-saturated with water by applying a partial vacuum of

(510 ± 25) mm Hg for (30 ± 1) minutes. Subsequently the specimen was subjected to 3

cycles of conditioning in water.

Each conditioning cycle comprised conditioning the specimen in a hot water bath at (60 ±

1)oC for (6 ± 1) hours, then transfering the specimen from the hot water bath and

immediately placing it in a cold water bath at (5±1) oC for (16 ± 1) hours, and then

removing the specimen from the cold water bath and immediately placing it in a water bath

at (20±0.5) oC for at least 2 hours.

At the end of each conditioning cycle, the stiffness value (at 20oC) of the specimen was

determined in accordance to BS EN 12697-26 Annex C.

Details of the retained stiffness results are presented in Appendix I. A summary is shown in Table 19 as follows.

Table 19: Summary of Stiffness Ratios of Water Sensitivity Data

Stiffness Ratio, Mean of 6 Cycle 1 Cycle 2 Cycle 3 Mix Type Mean Range Mean Range Mean Range

PA-GS-125pen 1.13 1.00 - 1.27 0.82 0.75 - 0.86 0.89 0.82 - 0.95

ARSC-GS-1 1.21 1.05 - 1.31 1.13 0.97 - 1.19 1.13 0.96 - 1.27

ARSC-GS-2 1.14 0.79 - 1.33 1.10 0.91 - 1.22 1.09 0.98 - 1.15

ARSC-GS-3 1.14 1.00 - 1.21 1.08 0.94 - 1.18 1.13 1.02 - 1.35

ARSC-GR-1 1.02 0.90 - 1.11 0.89 0.79 - 0.94 1.06 0.97 - 1.35

Notes: Stiffness ratio (Cyclei) = Conditioned ITSMi ÷ Unconditioned ITSMu; where i = 1, 2 or 3 (ITSM1, ITSM2, ITSM3, denotes Indirect Tensile Stiffness Modulus (ITSM) after conditioning cycle 1, 2 or 3, respectively).

An examination of the data in Table 19 and Appendix I show that the mean cycle 3 stiffness ratios for the RA samples are all above 1 and consistently greater than those of the control sample PA-GS-125pen. This indicates that the surface course RA blends would not be deemed to be susceptible to water related durability problems in service.

9.8 Binder drainage

The binder drainage test was carried out by using the basket method in accordance with the requirements of BS EN 12697-18. Both control and RA specimen were subjected to the binder drainage test. For each specimen, the binder drainage test was carried out in triplicate; the results are summarised in Table 20. A particular specimen would be regarded as having low risk of binder drainage when the test results showed less than 0.3% drainage.


Table 20: Binder Drainage Test Results

Mix Type Binder Drainage (%) Range (%)

Risk of Binder Drainage Low PA-GS-125pen 0.14 0.05 – 0.20 Low ARSC-GS-1 0.15 0.02 – 0.24 Low ARSC-GS-2 0.06 0.00 – 0.17 Low ARSC-GS-3 0.02 0.00 – 0.03

ARSC-GR-1 0.01 0.00 – 0.02 Low

SMA-LS-50pen 0.21 0.11 – 0.26 Low

ARBC- LS-1 0.02 0.02 – 0.02 Low

ARBC- LS-2 0.03 0.00 – 0.05 Low

ARBC- LS-3 0.40 0.39 – 0.41 “Some”

ARBC- GR-2 0.01 0.00 – 0.02 Low

Table 20 suggests that apart from ARBC-LS-3, all RA samples exhibited minimal percentages of binder loss (i.e. less than 0.3%) and therefore have a very low risk of binder drainage.

9.9 Durability The durability or resistance to particle loss was assessed by the loss of mass of compacted mixture subjected to 300 revolutions in the Los Angeles (LA) machine, in accordance with EN 12697-17. This process is commonly referred to as Cantabro testing. It enables an estimation of the resistance to abrasion of laboratory compacted specimens to be made. A threshold value of a maximum particle loss of 26% was specified for this study. Details of the Cantabro test results are given in Appendix K and are summarised in Table 21 below.

Table 21: Particle Loss Test Results SURFACE COURSE

Mix Type PA-GS-125pen ARSC-GS-1 ARSC-GS-2 ARSC-GS-3 ARSC-GR-1

Average Particle Loss %

19 (16 – 22)

8 (6 – 10)

7 (5 – 9)

9 (5 – 14)

4 (3 – 5)

BINDER COURSE

Mix Type SMA-LS-50pen ARBC- LS-1 ARBC- LS-2 ARBC- LS-3 ARBC-GR-2

Average Particle Loss %

15 (14 – 18)

5 (3 – 7)

6 (5 – 6)

3 (2 – 5)

4 (3 – 7)

Note: Values are an average of 4 specimens; figures in parenthesis denote range of results. Table 21 above suggests that the control mixtures have significantly higher percentages of particle loss compared to the RA mixtures. This implies that the RA mixtures will be expected to have better durability (resistance to particle loss or abrasion) than the control mixtures. All samples, however, show material loss lower than that specified for this study, indicating good resistance to particle loss of the tested samples.


10.0 Conclusions and Recommendations A preliminary assessment carried out prior to commencing Task 4 highlighted the sensitivity of rubberised bitumen to blending arrangements and test conditions (specifically viscosity testing). Factors that may affect the test results include:

Blending protocol (duration, shear rate)

Viscometer (type/make, shear rate, spindle size)

Viscosity data interpretation (peak value/ASTM 6114, stabilised value/AASHTO T316)

Subsequently, harmonised blending and test conditions were adopted, i.e. high shear blending at 6000 rpm for 15 minutes, with rubber being fed in during the first 10 minutes, followed by reducing the shear rate to 1500 rpm for the rest of the remaining blending time of 180 minutes. A Brookfield viscometer with spindle size No. 27 was used in the main test program.

During Task 4, 48 rubberised bitumen blends of previously agreed composition were manufactured and subjected to viscosity (ASTM 6114), penetration (EN 1426) and softening point (EN 1427) testing. The viscosity test results were compared against target viscosity values between 1500cP – 5000cP and, based upon the viscosity criterion, the ‘best’ six blends were selected for further assessment in TASK 6. These blends comprised:







The six rubberised bitumen blends from Task 4 were further assessed for the effect of rubber particle size on the blend properties. During this exercise, both maximum (peak-ASTM D6114) and “stabilised” (AASHTO T316) viscosity values were recorded as well as penetration (EN 1426) and softening point (EN 1427) test data. It was found that finer particle sizes, which have a larger surface area, led to greater reaction and interaction with base bitumen and hence a higher viscosity of the rubberised bitumen. The three best performing rubberised bitumen blends were then selected (all containing “standard” size rubber), and taken forward to Task 5 and Task 7. These comprised:




The three rubberised bitumen blends from Task 6 were subjected to more comprehensive testing under Task 5 and Task 7 with the objective of gathering information about the properties and performance characteristics of the blends, and to provide information as required by BS EN 14023 (“Framework specification for polymer modified bitumens”). Results from Task 5 showed that the selected blends satisfied the requirements for ageing index, softening point and viscosity and that both rubberised blends with Venezuelan bitumen met the criterion for resilience. The Middle Eastern blend fell short of the recommended resilience value, although its performance was better than that of unmodified binder. After assessment in accordance with EN 14023 test requirements in Task 7, it was found that:

Short-term and long-term ageing test results generally suggest improved properties of the

rubberised bitumens over those expected for 40/60pen bitumen.

The peak cohesion values and the temperature range over which cohesion values are at or

above 0.5 J/cm2 for the rubberised bitumen blends are much higher than those for typical

40/60pen bitumen, indicating significantly improved cohesion characteristics for these

blends and hence improved ability to retain aggregate.


Fraass breaking point values for the rubberised bitumen were at least similar to or better

than those for 40/60pen bitumen.

The plasticity ranges of the rubberised bitumen blends met the requirements for EN14023

polymer modified binder Classes 2, 3 and 4, and were wider than those found typically for

40/60pen bitumen. This suggests that the rubberised bitumen blends can be expected to

have a relatively wide service temperature range.

The flash point values of the rubberised bitumen blends comply with those specified for

polymer modified binders, although the blends were observed to release visibly thick white

fume with strong odour during testing, specifically when the test temperature was increased

above 200oC. It is recommended that the significance of the presence of this fume is

assessed in more detail.

The results on the stability of the blends during hot storage suggested only a marginal

difference in the penetration value of the top and bottom part of the storage container and

a difference of up to 25% in the softening point values. This was not considered unusual

and is consistent with results reported in the USA for similar materials.

Task 9 involved a two step process, the first being a screening assessment (workability assessment) of twelve rubberised asphalt (RA) mixtures and four control mixtures, and the second being mechanical assessment of the eight best performing blends from the screening assessment and two control mixtures. Due to the limited time available to complete this study, a comprehensive mixture design exercise was not carried out; instead, designs typically adopted for these asphalt mixtures were used. During the screening assessment, it was found that:

Apart from ARBC-LS-2, all RA mixtures required a lower mixing torque compared to their

respective control mixtures. This indicates that the RA mixtures have improved workability

compared with the control mixtures, probably as a result of the higher mixing temperature

and higher binder content, despite rubberised bitumen being more viscous than control

(unmodified) bitumen.

Bulk density test results showed that there are only slight variations of bulk density with

mixing/holding time and that the average bulk density of each mix generally remains

constant throughout the period of “holding” at high temperature.

The maximum densities of the mixtures after 10 minutes of mixing/holding were not

significantly different from those after 120 minutes of mixing/holding.

RA mixtures generally showed an increase in mixture stiffness with increase in mixing

(holding) time and these values were greater than the recommended 1700 MPa. The

control mixtures showed significantly lower values of stiffness in comparison with the RA

mixtures. This indicates improved load spreading ability of the RA mixtures over that of the

control mixtures. Overall, the holding time did not lead to a reduction in stiffness.

Apart from samples manufactured using granite aggregate, surface course and binder

course mixtures did not show a significant change in tensile strength with increase in mixing

(holding) time. All RA surface courses showed tensile strength higher than that of the

control, whilst the RA binder course showed less consistent performance. In all cases, the

tensile strength values of the RA mixtures were greater than 600 kPa, indicating good

quality material.


The findings from the mechanical assessment are summarised below:

Air voids at refusal were all above the 1% minimum allowable for dense asphalt mixtures.

On this basis ARBC mixtures are expected to have adequate resistance to secondary

compaction; secondary compaction may lead to rutting.

Surface course RA samples have lower rut depth and rut rates compared to PA control

samples, and on this basis have better resistance to deformation. All surface course samples

met the requirements for surfacing material for heavily stressed areas (SHW Clause 942

Class 3). Similarly, the binder course samples also showed good resistance to deformation

and compliance with the requirements of PD 6691 Class 2.

Crack resistance data revealed that surface course RA samples had better flexibility and

better resistance to low temperature cracking than the control PA. However, the binder

course RA samples appeared to have poorer crack resistance than the control SMA

ARSC-GS-3 had fatigue resistance comparable to that of the control PA sample, with both

samples being the only surface course samples below the minimum value of 10,000 cycles

to failure for a microstrain of 200 adopted for this research project. The rest of the surface

course samples had better fatigue resistance. The fatigue resistance of the binder course is

less clear cut, with some RA samples showing improved performance at high strain but not

at low strain, compared with that of the control, and vice versa.

The retained stiffness ratios of all the surface course RA samples were above 1 and

consistently greater than those for the PA control samples, indicating that the RA blends

would not be deemed susceptible to water related durability problems while in service.

Test results showed that all mixtures, apart from ARBC-LS-3, had a low risk of binder

drainage.

Surface course RA mixtures would be expected to have better durability (resistance to

particle loss) than the control mixtures, since particle (Cantabro) loss test results showed

them to have significantly lower percentages of particle loss in comparison. In practice,

these results can be translated as indicating good resistance to fretting.

A performance ranking summary for the RA mixtures is presented below:

Surface Course Parameter PA-GS-

125pen ARSC-GS-1

ARSC-GS-2

ARSC-GS-3

ARSC-GR-1

Workability 3 2 2 2 1 Stiffness 3 1 1 2 2 Tensile strength 3 1 2 2 2 Voids at refusal 1 3 3 1 1 Deformation resistance 2 2 2 2 2 Crack resistance 3 1 1 1 3 Fatigue resistance 3 2 1 3 1 Resistance to Moisture Damage 2 1 1 1 1 Binder drainage 1 1 1 1 1 Durability 3 1 1 2 1 Total 24 15 15 17 15


Binder Course Parameter SMA-LS-

50pen ARBC- LS-1 ARBC- LS-2 ARBC- LS-3 ARBC- GR-2

Workability 1 2 2 1 1 Stiffness 2 3 2 1 1 Tensile strength 3 1 1 1 1 Voids at refusal 3 1 1 1 3 Deformation resistance 2 2 2 3 1 Crack resistance 3 2 2 2 2 Fatigue resistance 1 1 3 2 1 Binder drainage 2 1 1 3 1 Durability 2 1 1 1 1 Total 19 14 15 15 12

Note 1= best performance overall; 3= worst performance overall The smaller the total score the higher is the overall performance The sensitivity of the ranking performance was further assessed by normalising the performance (i.e. applying a weighting factor) to the parameters most appropriate to the material’s application. In this case, the key (very important) performance parameters are tensile strength, crack resistance, fatigue resistance and durability. For the surface course, the mixture resistance to moisture damage is also considered important. Other parameters, considered as part of good mixture design, have not been included in the weighting exercise. This further analysis is summarised below.

Surface Course Parameter Weighting

Factor* PA-GS-125pen

ARSC-GS-1

ARSC-GS-2

ARSC-GS-3

ARSC-GR-1

Workability 0 0 0 0 0 0

Stiffness 0 0 0 0 0 0

Tensile strength 1 3 1 2 2 2

Voids at refusal 0 0 0 0 0 0

Deformation resistance 1 2 2 2 2 2

Crack resistance 1 3 1 1 1 3

Fatigue resistance 1 3 2 1 3 1 Resistance to Moisture Damage 0.5 1 0.5 0.5 0.5 0.5

Binder drainage 0 0 0 0 0 0

Durability 1 3 1 1 2 1

Total 15 7.5 7.5 10.5 9.5

Binder Course Parameter Weighting

Factor* SMA-LS-50pen

ARBC- LS-1

ARBC- LS-2

ARBC- LS-3

ARBC- GR-2

Workability 0 0 0 0 0 0

Stiffness 1 2 3 2 1 1

Tensile strength 1 3 1 1 1 1

Voids at refusal 0 0 0 0 0 0


Deformation resistance 1 2 2 2 3 1

Crack resistance 1 3 2 2 2 2

Fatigue resistance 1 1 1 3 2 1

Binder drainage 0 0 0 0 0 0

Durability 1 2 1 1 1 1

Total 13 10 11 10 7 Note: *Weighting factor: 0 = not included, 0.5 = important, and 1 = very important to pavement performance. The smaller the total score the higher is the overall performance. The key ‘weighting factor’s can be tailored to the specific performance level that may be promoted depending upon the rubberised asphalt application. Similarly, the type of control sample can be replaced by other proprietary asphalt surfacing products and comparison can be made in the same way by including the relevant performance information of the elected product into the above tables. The above analysis indicates that surface course rubberised asphalt particularly would be expected to have superior overall performance compared with the respective control. However, design optimisation is necessary for any rubberised asphalt, before adopting any of these mixtures in practice for larger scale applications; this will ensure a greater level of enhanced performance can be obtained. Good quality control is always essential to ensure good quality rubberised asphalt can be manufactured and laid, and ultimately deliver the target level of performance. Work was limited to one source (processing plant) for each rubber type. Thus variations in rubberised bitumen properties due to use of rubber other than those from these sources remains unknown and should be the subject of further research. 11.0 Acknowledgements This study was commissioned and funded by the WRAP Tyre Programme. Aggregates, bitumens and rubbers that have been used on this study were mostly provided in-kind by a number of UK industries, including (in no particular order): Aggregate Industries, Tarmac, Nynas, UK Bitumen, Moncton Rubber Technology, Tyregenics, Allcock and Sons, and Singleton Birch. Streams of advice were also received from the Technical Advisor to WRAP, Mr Ian Walsh. Supports and contributions from the above organisations are gratefully acknowledged.

12.0 References

De Beer M, Fisher C and Jooste, F (1999). “Determination of Pneumatic Tyre/Pavement Interface Contact Stresses under Moving Loads and some Effects on Pavements with Thin Asphalt Surfacing Layers”. Transportation Research Board Meeting, National Academy of Science, Washington.

Van Rooijen, R.C., and De Bondt, A.H. (2005). “A Step towards Improved Functional Specifications for Asphalt Mixture and Bitumen Properties for Airfield Applications”, 1st European Airfield Pavement Workshop, Amsterdam.


www.wrap.org.uk/tyres