bookchapter_fka07

First Edition 2008 © HASANAN MOHD NOR, MOHD ROSLI HAININ & HARYATI YAACOB 2008

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopy, recording, or any information storage and retrieval system, without permission in writing from Universiti Teknologi Malaysia, Skudai, 81310 Johor Darul Tak'zim, Malaysia.

Perpustakaan Negara Malaysia Cataloguing-in-Publication Data

Highway materials and construction / editors Hasanan Mohd Nor, Mohd Rosli Hainin, Haryati Yaacob. ISBN 978-983-52-0574-3 1. Pavements, Asphalt concrete. 2. Asphalt concrete. 3. Highway engineering--Malaysia. I. Hasanan Mohd Nor. II. Mohd. Rosli Hainin, 1961-. III. Haryati Yaacob. IV. Universiti Teknologi Malaysia. Fakulti Kejuruteraan Sivil. 625.85

Pereka Kulit: MOHD. NAZIR MD. BASRI

Diatur huruf oleh / Typeset by HASANAN MOHD NOR & RAKAN-RAKAN

Fakulti Kejuruteraan Awam Universiti Teknologi Malaysia

81310 Skudai Johor Darul Ta'zim, MALAYSIA

Diterbitkan di Malaysia oleh / Published in Malaysia by PENERBIT

UNIVERSITI TEKNOLOGI MALAYSIA 34 – 38, Jalan Kebudayaan 1, Taman Universiti,

81300 Skudai, Johor Darul Ta'zim, MALAYSIA.

(PENERBIT UTM anggota PERSATUAN PENERBIT BUKU MALAYSIA/ MALAYSIAN BOOK PUBLISHERS ASSOCIATION dengan no. keahlian 9101)

Dicetak di Malaysia oleh / Printed in Malaysia by UNIVISION PRESS

Lot 47 & 48, Jalan SR 1/9, Seksyen 9 Jln. Serdang Raya, Tmn Serdang Raya

43300 Seri Kembangan, Selangor Darul Ehsan MALAYSIA

CONTENTS

Preface

vii

Chapter 1 Speciality Mixes in the New Malaysian Public Work Department Road Specification Azmi bin Hassan, Zulakmal Hj. Suffian

1

Chapter 2 Providing a Safe Economic and Durable Highway Surface Alan Woodside

18

Chapter 3 Determination of Periodic Pavement Rehabilitation Treatment using Pavement Condition Assessment (PCA) Ahmad Kamel bin Abdul Malik

37

Chapter 4 Evaluation of Malaysian Asphaltic Concrete Mixture using Superpave and Marshall Mix Design Method Juraidah Ahmad, Mohd Yusof Abdul Rahman, Mohd Rosli Hainin, Mustaque Hosasain

49

Chapter 5 Crack Progression Models for Flexible Pavements Sugeng Wiyono,Othman Che Puan, Mohd Rosli Hainin

63

Chapter 6 Investigation on Tyre/Road Noise using Ulster Load Tyre Road Haryati Yaacob, Mohd Rosli Hainin, Alan Woodside, David Woodward

72

Chapter 7 Performance of Hot Mix Asphalt using Fine Crumber Norhidayah Abdul Hassan, Mohd Rosli Hainin, Haryati Yaacob

93

Chapter 8 The Effect of Rainfall on Asphalt Surfacing Materials Nursetiawan, David Woodward, Alan Strong

107

Chapter 9 Research into Sustainable Asphalt Surfacing Mixes in the UK David Woodward, Alan Woodside, Alan Strong, Paul Phillips, Bob Allen

119

Chapter 10 Extent of Pavement Disintegration associated with Low Cost Road Failure J. Ben-Edigbe

133

PREFACE

Highway materials and construction is a vast field covering a wide range of technical areas that are rapidly evolving. This book contains ten chapters, covering on some outstanding research on highway materials and construction conducted over the years. This books aims to provide the latest information, findings and innovation in highway engineering areas. It will be found useful not only by researcher and students of civil engineering in universities or other centers of higher education, but also to the engineers and industrial people who may be involved in the application of new technology. We sincerely hope that this book will facilitate and serve as a reference to all highway engineering lovers. Hasanan Mohd Nor Mohd Rosli Hainin Haryati Yaacob Faculty of Civil Engineering Universiti Teknologi Malaysia 2008

240 Advances in Manufacturing and Industrial Engineering (2008)

1

SPECIALTY MIXES IN THE NEW MALAYSIAN PUBLIC WORK

DEPARTMENT ROAD SPECIFICATION

Azmi Hassan Dato’ Ir. Dr. Director, Roads Branch, Public Work Department,

Malaysia

Zulakmal Hj Sufian Ir., Principal Assistant Director, Road Maintenance Branch, Public Work

Department, Malaysia

1. INTRODUCTION The flexible pavement structure in Malaysia consists typically of bituminous surfacing, granular road base, drainage sub-base and the formation subgrade. The pavement structure is designed in accordance to the Arahan Teknik (Jalan) 5/85 (1) which is adapted from the AASHO (American Association Of State Highway Officials) Road Test.

Asphaltic concrete mix has been introduced as the bituminous surfacing in the 80’s, both as the wearing and binder course. Designed using the Marshall Method (2), asphaltic concrete was preferred over the Bituminous Macadam surfacing, a recipe mix, as the latter exhibited several unacceptable distress patterns such as raveling and early development of cracks. Surface dressing was

2 Highway Materials And Construction

another common surfacing used mainly for FELDA and FELCRA settlement roads. Surface dressing is normally designed using the Road Note No. 3 of the Transport Research Laboratory, United Kingdom. (3).

All bituminous mixes play a similar and very important role in road pavement construction, i.e. providing a hard and impermeable layer to the road pavement. The “hard” layer prevents undue deformation in the unbound layers when subject to traffic loading. The impermeable asphalt bound layer also prevents water from reaching the lower layers of the pavement structure thereby weakening the layers.

2. REQUIREMENTS FROM BITUMINOUS MIXES

The basic requirements from bituminous mixes are as follows: a. Stability Resist permanent deformation b. Flexibility Resistant to cracking or fracture due to:

• Subgrade weakness • Repeated flexure • Volume changes of underlying material

c. Durability Resistant to: • Water • Ageing of bitumen

d. Workability We want the mix to spread and compact to the required density without difficulty

e. Safety Good skidding resistance with no fretting or ravelling

f. Impermeable Prevent water from reaching underlying layers g. Stiffness We want high stiffness to distribute the traffic

stresses i.e. “The higher the modulus, the lower the subgrade stress”

What gives us the requirements?

Highway Materials And Construction 3

a. Stability Stone contact:

Stone must have high crushing strength Aggregate must have a mechanically

stable grading Aggregate must have a rough texture Crushed cubical rocks better than gravel

Mortar: Stiffness must be kept high using hard

bitumen High filler contents to help stability Sharp angular rough textures sands

b. Flexibility Long term movement (subsidence) To avoid fracture use softer bitumens,

higher binder contents and open aggregate grading

Fatigue fracture Where thick carpets are used, require

dense mixes with high bitumen contents When thin carpets (e.g. low cost roads),

use soft binders to allow flow and healing

c. Durability Improved by dense impervious mixes, preventing ingress of air and water. Obtained by dense graded aggregates and high bitumen contents

d. Safety Skid resistance: Require adequate surface texture depth and good polishing characteristics of the stone. No free bitumen to surface (bleeding)

e. Impermeable Low voids (even these may be interconnected allowing ingress). Use dense mixes and high bitumen contents

f. Stiffness The higher the modulus the better. Use dense gradings and hard bitumens.


3. MODES OF PAVEMENT DISTRESS IN MALAYSIA

Cracking

Research work by the Public Works Institute Malaysia (IKRAM) has shown that asphaltic concrete (ACWC 20 or 14) roads failed predominantly through cracking. For new roads, the asphaltic concrete normally failed through top-down and fatigue cracking while reflection crack normally occurs on roads where resurfacing was done directly on top of a previously cracked road.

Top- down cracking Bituminous surfacings in Malaysia use primarily the conventional 80/100 pen grade bitumen as binder. For the production of asphaltic concrete, the 80/100 bitumen is normally blended with hot aggregate at a mixing temperature of 150 - 170 0C. Research work by the JKR and other research institution overseas have shown that bitumen tends to harden while at the early stage of handling, in storage, during mixing and in service. Chemical reaction between bitumen and oxygen has been determined to be the main cause of bitumen hardening during these stages.(4) It is an accepted fact that 80/100 pen grade bitumen loses at least a grade after undergoing hot mixing process at the asphalt producing plant. During service, exposure of the bitumen film to ultra-violet radiation exacerbated the oxidation of the bituminous binder. The thin bitumen film in the top layer of the road surfacing soon becomes brittle and causes an increase in the stiffness moduli of asphaltic mixes, but at the same time reduces the strain required to induce cracking.

Field cores retrieved from asphaltic concrete road surface of different ages revealed that bitumen coating in the top layer of the road surface becomes hardened typically after 4 to 5 years in service. Chemical reaction between the bitumen and the atmospheric oxygen is the primary cause for hardening of the bitumen. Brittled bitumen tends to crack due to shrinkage stresses


caused by the diurnal temperature change, or by traffic induced stresses, or by a combination of both. Surface crack allows the hardening process to take place deeper into the body of the mix resulting in the propagation of surface crack downwards, a phenomenon referred to as top-down cracking.

Top-down cracking occurs prematurely compared to the normal design life of road pavement (seven to ten years). This is due to the thin bitumen film thickness (typically 5-10 micron) prevalent in the mix. The hardening process due to oxidation is exacerbated by the presence of high voids content (5 - 8 %) in the mix due to inadequate compaction or improper mix design. Exposure to the ultra violet radiation and moisture accelerate further the deterioration rate of the asphaltic concrete mix.

Reflection crack Research by Mutalif et.al. (5) on the effectiveness of 40 mm thick asphaltic concrete overlays, showed that cracks in the existing surface reflected through the new asphaltic concrete overlay in a relatively short period. The rate at which the cracks reflect depended on the type of cracking and the magnitude of the pavement deflection prior to overlay, and the cumulative flow of commercial vehicles after the construction. This study shows that the use of 40 mm asphaltic concrete overlays to rehabilitate roads with interconnected cracks is ineffective.

The asphaltic concrete is stiff mix due to its close gradation and inclusion of 2% cement. Opening and closing of cracks in the original layer due to traffic loading create stress tips in the bottom layer of the overlay above. These movements, the extent of which are determined by the structural capacity of the road pavement, crack block size and traffic loading, cumulate strains in the bottom layer of the overlay. The asphaltic concrete overlay will typically crack when the cumulative strains exceed about 1%. Subsequent traffic loading will cause these cracks to propagate upwards.


Rutting The introduction of asphaltic concrete in Malaysia has brought with it the problem of the extensive quality control testing that is required to produce such mixes to the required Marshall tolerances. Even when these materials are produced to specification, they are often inappropriate in areas of high traffic stresses, such as climbing lanes and main junctions, as is evident by the rapid permanent deformation along the wheel paths throughout the country. This type of surface distress is hazardous to the road users as it allows surface water to accumulate, thus increasing the risk of skidding and aquaplaning. Water pondings also increase the risk of water infiltration into the pavement structure. The deformation makes driving dangerous especially when vehicles change lanes. If it is allowed to deteriorate, it can lead to longitudinal cracking of the surface and subsequently permits the ingress of water. As the deformation causes a reduction in driving comfort and affects safety, the road has to be inevitably rehabilitated now and then.

Rehabilitation of this nature always present technical problems that are not easily solved and inflicts a considerable maintenance cost, whether it is decided to remove the surfacing under distress and replace them or whether it is only a matter of providing regulating overlays. This type of maintenance, nonetheless, proved to be ineffective and uneconomical.

Lack of macrotexture

Besides cracking and rutting, asphaltic concrete road surfacings in Malaysia lack the desirable macrotexture for skidding resistance at high speeds (>50 km/hr). Johari et. al. (6) observed that the average texture depth (sand patch) on asphaltic concrete surfacings is 0.35 mm. Research work by the TRL suggested that (7) that the risk to accidents increases when the macrotexture falls below 0.7 mm.

Skidding resistance is the term used to describe friction between the road and the tyre. It is generally considered that a dry road


normally provides adequate grip, but in wet conditions (about 25 percent of the time in Malaysia), friction falls and there is a greatly increased risk of skidding. For this reason, the term skidding resistance is normally assumed (and is always used here) to refer to wet roads. Although tyre characteristics are important, the main contribution to skidding resistance comes from the road itself (8)

Two components determine the skidding resistance of a road surface: microtexture and macrotexture. Microtexture depends on the surface roughness of aggregate and varies with aggregate types. It is frequently described as either rough or smooth. Macrotexture is formed by the aggregate arrangement at the road surface. Macrotexture can be obtained by aggregate protrusion at the surface (e.g. Hot Rolled Asphalt with chippings rolled on top) or aggregate inversion (in the case of porous asphalt). At speeds greater than 50 km/hr, coarse macrotexture is required for channelisation of water trapped between tyres and road surface thereby evading aquaplaning. In Malaysia, the average yearly rainfall ranges between 1700 mm and 2000 mm hence the importance of macrotexture cannot be overemphasized.

4. NEW BITUMINOUS MIXTURE IN MALAYSIA

As an option to prolong the life and enhance the performance of bituminous surfacing, the PWD has introduced many new mixes in its projects and maintenance programs since early 90’s. The experiences gained through years of monitoring the performance of the mixes culminate with the introduction of the 2008 Road Specifications. Starting this year, the PWD enforces the use of the 2008 Road Specifications in all of its new road projects and maintenance programs. This paper highlights three of the new mixes included in the specifications, namely Polymer Modified Asphalt, Stone Mastic Asphalt and Porous Asphalt. Due to their


specialised features and purposes, the PWD has termed the new mixes as Specialty Mixes.

Polymer Modified Asphalt

It is well established that although binder forms a small proportion of an asphaltic mix, it is primarily responsible for the visco-elastic response of the mix, and hence many aspects of road performance. The main objective of using modified binders in asphaltic mix is to provide a cost effective solution in improving the resistance to permanent deformation of the surfacing materials at high temperature and under extreme loading conditions (9). This is achieved by stiffening the binder so that the viscous response of the asphaltic mix is reduced resulting in a corresponding reduction in permanent strain. Alternatively, the elastic component of the binder is increased, thereby reducing the viscous component, which again results in a reduction in permanent strain.

Secondary benefits in terms of resistance to fatigue cracking, better load spreading ability and resistance to aging may also be gained with some of the additives.

There are many types of polymer additives used in binder modification and they fall into two categories, elastomers and plastomers. The elastomers resist permanent deformation when stretched and their strength increases with level of elongation. Natural and synthetic rubbers and lattices of Styrene Butadine Styrene (SBS) are common elastomeric polymers used in modifying bitumen to make the binder more resilient and flexible. Plastomers form a rigid and tough structure that resists deformation. Copolymer of Ethylene Vinyl Acetate (EVA) and polyethylene (PE) are common plastomeric polymer to produce a binder with increased stiffness. The degree of modification required, and hence the cost, will depend upon the needs of the site to be treated.

Construction of polymer modified asphalts (PMAs) in Malaysia began in the late eighties, initially as full scale trials and


subsequently in major rehabilitation projects. Various polymers had been used and these include the Polybilt 101, XCS 503, Chemcrete, Gilsonite and synthetic rubber. Proprietary polymer modified bitumens (PMBs) have also been used in asphalt construction and these include Multigrade, Caribit-Plus, Cariphalte-DM, Asphapol 2000, Novophalt, Sealoflex and Flexipave.

In general, the performance of PMAs has been a mixed one. In an experiment comparing the relative performance of a control asphaltic concrete with similar material modified with Polybilt polymers, Hizam (10) reported that Polybilt modified asphalt deformed to a level of 25% that of the control sample when tested using the Immersion Wheel Tracking Test. The laboratory results were subsequently confirmed by findings from a full scale trial.

In another trial involving Chemcrete, and Caribit-Plus modifiers, the PMAs showed marked improvements in rutting resistance. However, cracks started to develop 6 months after construction in the Caribit-Plus and Chemcrete sections. Visual inspection of cores taken from the affected areas showed that the cracks originated from the top and had propagated down by no more than 20 mm. The control section for this trial had unexpectedly performed better with minimal rutting and no cracks after approximately 16 months. Subsequent tests on the mix composition showed that the binder content was 5.27%, much less than the design value of 5.95%. The improved performance could be due to the reduction of the binder content (11).

The above observations indicate that improved resistance to rutting is possible with most PMAs, however other modes of distress namely cracking are likely to dominate with some PMAs. Improvement to rutting resistance is equally achievable with the conventional mixes given proper design and quality control. Additionally, the PMB improves adhesiveness between binder and aggregate. The improved property has mitigated delamination problems on PMA surfacing during the country’s recent flood encounters.


The inclusion of modifiers will incur an increase in cost of the binder in the range of 75% to 500%. This normally hikes the cost of a PMA by about 20 to 150 percent compared with the conventional asphalt. For a cost effective investment in PMAs, the tangible improvements in the modified mix should at least be on par with the increase in cost due to modification. This is not normally achieved by a single additive as it is effective only in improving certain parameters of the asphaltic mix. An elastomer, for example, may improves the fatigue resistance of an asphaltic mix but has less influence on the elastic stiffness of the mix. The design of polymer modified asphalts has therefore to consider job specific needs.

Nonetheless, the PWD are convinced that use of PMAs is quite justified in specific areas of applications where the conventional asphalts have failed. Such areas include the airport pavements, highly stressed roads and specialty mixes like the porous asphalts and stone mastic asphalts.

In Malaysia, the construction of Kuala Lumpur International Airport runway and taxiway pavement in 1996 saw the largest application polymer modified asphalt. The design of bituminous binder was done using the then newly launched Strategic Highway Research Program’s (SHRP) performance grading. The SHRP grading was adopted since research work has shown that the conventional quality control tests like the penetration @ 25 0C, softening point etc. do not relate well to field performance. The SHRP PG-grading also takes into account aging characteristics of the binder after production as well as in service. Based on the PG-grading, a PG 76 was specified for the airport runway and taxiway. W In the new PWD road specifications, a PG-76 polymer modified bitumen is specified and the PMA is also subject to additional test procedures namely the Indirect tension test for resilient modulus (ASTM D 4123) and Dynamic, unconfined, compressive creep test (EN 12697-25).


Porous Asphalt

Porous asphalt is an innovative bituminous mixture for use as a friction course. It differs from the conventional continuously graded asphaltic concrete in that it is produced using open-graded aggregate and contains a relatively large air voids after compaction. It is a non-structural layer and required to be laid on an impervious pavement so that rainwater entering into it is rapidly drained to the road shoulders.

Porous asphalt evolved in the United States from experimentation with a seal coat treatment. Owing to its nature of application whereby aggregate were spread and rolled lightly into a layer of bitumen, the seal coat was relatively short-lived under high speed and heavy traffic. In the 1970s, this surface treatment was improved by mixing chip seal type aggregate with relatively high bitumen content and laying the mix using a conventional paver. This resulted in a surfacing of an open surface texture and enhanced friction.

At about the same time, porous asphalt was developed in Europe as a special-purpose friction course that drains rain water from the pavement surface and reduces rolling tyre noise levels. Even though the improvement of friction between the pavement surface and vehicle tyres is known to be an advantage of this friction course, the porosity of the mixture is its dominant feature.

Other benefits that are usually associated with the use of porous asphalt as a friction course include;

i. Improvement of skid resistance at high speeds, especially

during wet weather. ii. Reduction of splash and spray. iii. Suppression of hydroplaning effects.


iv. Reduction of headlight reflection and glare on wet pavement surface.

v. Improvement of visibility at night.

Porous asphalt was introduced in Malaysia only in 1991 when the Public Works Department undertook a project to resurface Federal Route 1 between Cheras and Beranang using the Skid Resistant Silent Surfacing (SRSS), a porous asphalt using binder modified with polymer additive Polybilt XCS 503. Since then, different designs of porous asphalt had been experimented with at various locations throughout the country. These include:

1. Porous asphalt with Gilsonite

FT 1, Jalan Beranang - Seremban

2. Porous asphalt with natural rubber

FT 1, Tampin and Route 2, Kuantan

3. Porous asphalt with Europrene Sol-T

FT 2, Federal Highway

4. Porous asphalt with Cariphalte

FT 3, Jalan Tebrau and FT 2, Federal Highway

Performance of Porous Asphalt

Research work by PWD has shown that, in general, Porous Asphalt laid at the above locations performed satisfactorily. For example, porous asphalt laid on Jalan Tebrau lasted for 5 years before it was milled off. Similar performance was observed for the porous asphalt on Federal Highway.

The Porous Asphalt surfacing reduces vehicle spray, headlight glare and surface ponding of water during downpour. Compared with the conventional asphalt, porous asphalt in all test locations achieved higher macrotexture and recorded lower rutting.


With the exception of Gilsonite modified Porous Asphalt, all other Porous Asphalts showed good drainability. The lack of drainability in the Gilsonite modified Porous Asphalt was however found to be due to the excessive compaction effort applied during laying.

It was also observed that, the drainability of Porous Asphalts diminishes with time. Two main reasons for this are rearrangement of the aggregate skeleton due to trafficking and inclusion of debris in the mix. On the Federal Highway, the reduction in drainability is more apparent in the slow lane. In the fast lanes, air pressure from high speeding traffic creates partial vacuuming effect and sucks the debris out of the mix.

Due to its high voids contents, Porous Asphalt was prone to damage due to diesel spillage. On the Federal Highway (Route 2), Porous Asphalt constructed in bus lay-bys developed potholes very early after their construction and was milled off and replaced by the semi rigid surfacing after less than one year.

Based on the above sites, the PWD have identified the benefits and limitations of Porous Asphalt. Porous Asphalt is most suitable to be constructed on high speed trunk roads with minimal road side development to minimize early clogging of the voids. Proper road surface cambering of at least 2.5% improves the performance of Porous Asphalt road by facilitating drainage of water through the body of the mix to the roadside drain or shoulder. Porous asphalt is also able to reduce the tyre and surface contact noise by as much as 2- 4dBa, thus it is an appropriate surfacing for highways or trunk roads situated besides residential areas. The use of porous asphalt on gradients higher than 10% is not recommended because water tends to surface as it flows down the gradient. Porous asphalt is also not recommended at traffic junctions, tight curves, areas with slow moving traffic and for structurally deficient pavement. Raveling is the main mode of failure on porous asphalt and soon it develops into potholes. The authorities should specify a porous mix to patch potholes created by raveling.


Stone Mastic Asphalt

Stone Mastic Asphalt (SMA) was originally developed in the Europe. The USA started using the SMA in 1990. Essentially, SMA is a polymer modified hot bituminous mixture with a large proportion of coarse aggregate and rich bitumen-filler mastic. Although some countries had use SMA without the polymer modified bitumen, new SMA constructions usually use polymer modified bitumens due to their many benefits. Generally, SMA comprises approximately over 65% coarse aggregate and a minimum of 8% filler content. The coarse aggregate, through point to point contact, forms a high skeleton with good internal friction and aggregate interlock to resist load-induced shear. It provides durable surface that is resistant to cracking and rutting. In 1996, a study on SMA in the USA has shown that 90% of projects had 0.4 mm rutting and 25% no rutting. Advantages of SMA include;

i. Expected longer life than dense graded mixtures ii. Fewer disruption to traffic due to less maintenance

activities, iii. Saving in additional resurfacing, iv. Higher resistance to rutting, v. Reduce tyre splash, vi. Reduce noise

In 1999, the PWD constructed a test section on SMA on Route 1, Templer Park that included SMA, with and without fibre, and ACWC20 as a control. Constructed on a climbing lane, SMA performed significantly better than the ACWC with minimum rutting (< 4mm ) after two years while the ACWC developed rutting between 10 – 15mm within the same period. The inclusion of cellulose fibre in SMA` is designed to allow rich binder content (> 6%) in SMA, making it more durable. Such high amount of binder content in the conventional ACWC mix will result in bleeding and shoving of the mix under traffic loading. Zulakmal


(12) found that SMA with high PMB content (6 – 6.5%) performed significantly better than ACWC20 at optimum PMB content (5.4%) in creep test, indicating that the stone arrangement, rather than the PMB, is more critical in determining the creep (rutting) performance. Since the trial, the PWD has constructed several stretches of SMA projects such as on Jalan Tebrau, Johor, Jalan Bukit Putus, Negeri Sembilan and Dabong to Kemubu, Kelantan. The performances of SMA at these locations are still good and being monitored.

Currently, SMA costs about 50% more than conventional dense graded mixtures due to inclusion of fibres, PMBs and higher binder content. However, it is also possible to reduce the thickness of SMA surfacing which has the effect of offsetting the additional costs.

5. CONCLUDING REMARKS Of the many types of mixes in the new PWD specification, the above three specialty mixes seems to be the most significant in mitigating pavement distresses.

The PWD has constructed and monitored the performance of many other specialty mixes such as the Gap Graded Asphalt, Chip Seal, Microsurfacing, Semirigid, and Coloured surfacing and successfully transformed the observations into the 2008 Road Specifications. In drafting the specifications, the PWD has researched and embodied experience from overseas about similar mixes. It is hoped that the new specifications will have a significant positive impact on the road construction industry.


REFERENCES 1. Public Works Department. Arahan Teknik (Jalan) 5/85.

Standards Unit, Roads Branch, PWD Malaysia.

2. Asphalt Institute.The Asphalt Handbook, Manual Series No. 4. Asphalt Institute, 1989.

3. Transport Research Laboratory. A Guide To Surface Dressing In Tropical And Sub-Tropical Countries. Overseas Unit, Transport Research Laboratory, Crowthorne, UK. 1982.

4. Traxler R N. Durabiity of asphalt cement, Proc Assoc Asph Pav tech, vol 32, pp 44-63, 1963. Association of Asphalt Paving Technologies, Seatle.

5. Ab Mutalif Ab Hameed. Pavement resurfacing:How soon will it recrack!!. Advisory Note No:IKRAM P-1. Institut Kerja Raya Malaysia.

6. H J Kwang, G Morosiuk and J Emby. An assessment of the skid resistance and macrotexture of bituminous road surfacings in Malaysia. Overseas Centre, TRL, UK.

7. P G Roe, D C Webster and G West. The relation between the surface texture of roads and accidents. TRL Research Report 296.

8. Lupton G N and T Williams. Wet road skidding resistance-the relative contribution of the tyre and of road surface texture. Department of the environment, TRRL Report SR86UC, TRRL, Crowthorne, UK.

9. Valkering C P and Vonk W. Thermoplastic rubbers for the modification of bitumens: Improved elastic recovery for high deformation resistance asphalt mixes. 15th. ARRB Conference, Darwin. 1990.

10. Mohd Hizam Harun. Summary of Program 4: Bitumen and Bitumen Additives. Public Works Institute Malaysia 1995.


11. Fernando M J and Guirguis H R. Natural rubber for improved surfacing. Vol 12, Part 2, Australian Road Research Board, 1984.

12. Zulakmal Hj Sufian. Deformation Characteristics of Binder Rich Mixes. M.Sc. dissertation, University of Birmingham, 1996.


2

PROVIDING A SAFE ECONOMIC AND DURABLE HIGHWAY SURFACE

Alan Woodside

Director of Transport and Road Assessment Centre, School of the Built Environment, University of Ulster

1. INTRODUCTION

The economic growth of a nation can be directly related to the infrastructure development. However, much of the wealth of a nation can be lost due to either congestion on the roads or collisions by traffic. In the U.K. alone collisions cost the government in excess of £3000M/per annum. Statistics show that the cost of a life is greater than £1.2M, consequently a multiple pile-up on any of the primary routes may result in, not only many deaths, but a considerable cost to the country.

In Northern Ireland alone there are approximately 4 fatalities per week on the roads due to traffic collisions. This exceeds the number of people killed by the “troubles”. When one considers the fact that there are >50,000 deaths each year on the roads of EUROPE one must address the problem from humanitarian point but also from a financial point of view.

2. THE PAST Looking back through the pages of history and taking time to study the evolution of the highway infrastructure of the world it becomes


very obvious that our forefathers made some "great steps forward for humanity". Others, one might say, made fundamental errors in their design but, in spite of their mistakes, the highways were developed and much of the infrastructure of the world constructed.

The Panathenian Way, built by the ancient Greeks in 400 BC as the approach to the Acropolis, was constructed of large blocks of local limestone to provide a clean, even surface for transporting the sacrifices up past Mars Hill to the Parthenon for temple worship. The ensuing Romans developed their highway network throughout Western Europe, not for religious purposes, but for military purposes. Again, using the local materials, they formed an even running surface of local slate or sandstone placed upon a lime bound gravel cement road-base, thus making best use of the natural resources. Unfortunately, the finished running surface did not lend itself to "hard wear" necessary for chariot wheels nor the more recent effect of the foot of man, which has caused excessive polishing.

As one continues on down through the centuries little progress has taken place to improve communicational facilities in UK in highway engineering until the network development in the 18th century by Thomas Telford and John Louden MacAdam, two eminent Scottish engineers. Telford assisted greatly with the design and construction of bridges but showed limited knowledge in pavement design with the provision of hand-pitched sub-bases - the vertical pitched stone causing 'punching-shear' in the sub-grade. However, MacAdam was more astute in his design, stipulating -"no stone greater than that which will fit into a man's mouth shall be used in road construction" - thus making more effective use of the natural materials and developing the first dry-bound Macadam, a forerunner to our present day bituminous macadam.

The running surface developed by the Scottish engineers consisted primarily of dry-bound macadam or in certain cases, wet mix macadam thus making use of local materials by simply adjusting


the grading. Some of these gradings are used to this present day on low volume roads such as forestry roads etc.

1863 saw the arrival of asphalt in Paris at Rue de Rivoli and the event is recorded by Charles Dickens. The City of Belfast followed some three years later when a section of the city centre was surfaced with Rock Asphalt only to be followed three years later in 1869 by the City of London, thus the introduction of this phenomenal material onto the streets of the capital - the street being Threadneedle Street in London. Unfortunately the original section in Paris has been recently resurfaced with porous asphalt. But Paris 1863 was one of the principal milestones in the evolution of our highway infrastructure and the use of innovative materials.

Since this early work in Paris many types of surfacing has been developed involving the use of much of our natural resources. From Monument Valley in USA to the M25 around London, materials have been developed and implemented, modified and improved, engineers endeavouring to provide the most ideal construction in their attempt to form "a safe, economic and durable" road surface.

Multi-millions of pounds and dollars have been poured into research projects throughout the world by governments and oil companies, contractors and suppliers, all seeking the ultimate in road design. However, one thing is common to all, that is the finite nature of the world's natural resources. These God given resources have a very short lifespan if one is to use them in an irresponsible manner. The author believes that all natural resources, ie aggregates (including sands, rock, etc) and bitumens, which are used in the construction of the highway, should be used with:

(a) Engineering sensitivity

(b) Environmental consideration

(c) Economical restraints.

One would require a very strong argument to obtain planning permission for any additional quarries to be developed from "green


field" sites in the UK or in many parts of Western Europe due to the over-supply from some active sources. Consequently, the author would recommend the more selective use of materials by providing "the correct materials in the correct place". In order to meet this criteria one must know and understand the materials - their ability to perform, their ability to meet a required need. Too often materials are over specified by clients. Even worse than this is when excessively good (high specification) materials are used in a low specification locality within the pavement structure.

3. WHOLE LIFE TESTING

The new European Standards state that "any aggregate may be used if it has a proven record". This is an excellent theory but it propagates the use of over-design and the excess use of superior materials and the consequential costly disposal of "unwanted" material. The solution lies in the performance related testing as a means of simulating the whole life testing (WLT) of a material or a matrix.

How often a "failure to comply with the specification" leaves the road engineer in an unenviable position of not knowing how to proceed - remove "unsuitable" recipe material which, if efficiently assessed, could meet all of the demands placed upon it during its life time. The author would propose a means of full-scale WLT of the material in the laboratory. This could possibly be developed with a variable scale of reduced payments.

Too often the Highway Authority is restricted by the use of outdated, irrelevant methods of specifying materials which may in turn prohibit the use of innovative or new materials -WLT could be the answer. Nowhere is this more applicable than in the assessment of marginal or waste materials, ie those materials which fail to meet the criteria laid down by British Standard or DIN (German) or other such National levels of acceptability. As a


Nation and a Community or a Union one must develop levels of acceptability based on performance rather than a recipe.

4. MARGINAL MATERIALS

Work carried out by Woodside and Woodward on unwanted, unacceptable (interims of BS Specification) laterite material which was found in the interbasaltic beds of County Antrim, N. Ireland, has shown that this material could be calcined at 1450oC to form a super high grade skid-resistant roadstone. Thus a waste product can be enhanced to make it perform satisfactorily and enable better use of natural resources - calcinisation enabled exploitation.

Another more common form of marginal material is that of construction waste. This is most acceptable environmentally. However, it still remains unpopular in United Kingdom with only 2% of all aggregates coming from a 'recycled' source. This is particularly low when compared with FRG who use 40% recycled aggregate and 80% of all aggregates used in Netherlands are recycled. Consequently, the Building Research Establishment in the United Kingdom are carrying out a major research project to assess the possible potential for such material as a suitable replacement for virgin material - thus conserving natural resources by either recycling (changing) or reuse.

How can one assess whether a material will perform satisfactorily? It is the author's opinion that one can only assess whether or not a material will perform satisfactorily by whole life testing. This has been used by the author in the assessment of Type 3 sub-base material. Where failure to meet BS tests or individual particle size would normally condemn the material, with the use of a segmental cell and a dynamic loading system it was possible to study the performance of the material at various moisture contents, followed by simulative wheel tracking.

A laboratory tenacity test was devised to measure and study the effect of dust on roadstone chippings. By determining the


minimum pull-off force necessary to remove chippings from binder under several loading and environmental conditions it was possible to set acceptable levels of dust content on chippings.

5. AGGREGATE CHARACTERISTICS

The shape of the stone was shown to play an important role in the strength of a mix, results being obtained for equivalent standard tests, ie LAAV, AIV and ACV using both high and low flakiness index samples. The author would suggest that this research emphasises how "improved results" may be obtained by simply changing the shape of the aggregate.

The removal of flaky stone can be achieved in one of three ways: (a) flakiness, (b) sieving or (c) crushing. The author recommends the latter and his research has shown that the tertiary crusher was capable of reducing the Flakiness Index to single figures. The material which failed to shape could then be used in other products such as concrete or asphaltic mixes where the aggregate is enveloped in a matrix. Sometimes, however, it may be necessary to think of using the product in another area - such as rock armour or railway ballast. A general rule is to "use the poorer (weaker) material lower in the road structure" remembering that "there is no such thing as a bad aggregate, only one which will perform better than others".

6. RECYCLING OF HOT ROLLED ASPHALT

The recycling of an existing road structure would appear to be an excellent means of exploiting natural resources - here is a quarry, a source of stone and bitumen already on site. This area of research should be developed more fully as it conserves the natural resources and reduces the need for disposal of waste. However, it


is not without its problems, as experienced by the author, viz the particle size of the planed-off material, the penetration grade evaluation and the environmental problem when reheating are three such problems.

However, recent research at UUJ has shown that Penetration Values is related to chemical composition and this may be directly determined using GCMS system. This has enabled rapid analysis to be carried out on samples of one gram.

Furthermore, the author would recommend the use of additives to bitumen to enhance the performance of a mix - usual additives would include SBS, EVA which may change the characteristics of the binder or Fibres which can enhance the performance of the mix by inhibiting the binder drainage and a thicker binder film thickness. Each of these materials have different characteristics which can optimise the performance of a binder, eg SBS will enable the chippings to be held with greater bond, thus deterring loose chippings. Whereas fibres can enhance the overall structural integrity of a mix, particularly when used in a Stone Mastic Asphalt or a Porous Asphalt mix.

7. THE REQUIREMENTS FOR ROAD PAVEMENT MATERIALS

Walsh of Kent County Council in UK stated recently at the CSS Conference that

“........ road pavement material has to fulfil a number of not always compatible requirements, as follows:.......”

(a) resist permanent deformation

(b) resist fatigue cracking

(c) be impermeable (if necessary)


(d) contribute to the strength of the pavement structure

(e) not be excessively brittle

(f) be workable during construction

In addition a wearing course must:

1. Provide a skid resistant surface in itself or by supporting and holding surface applied chippings and maintain adequate texture for the speed of traffic.

2. Resist abrasion.

3. Have acceptable riding quality.

4. Satisfy environmental considerations: noise; spray; glare; light reflectance; colour fuel use.

Whilst (a) demands a hard binder with low binder content and (b) and (f) a soft binder with high binder content there is usually a design ‘window’ which permits an appropriate mix to be designed. With the use by the HA of thick pavements, (large volumes of materials which justify the use of heavy compaction plant, requirements (b) and (f) are significantly less important than for LA designers whose design task is therefore more onerous.

In the case of wearing courses the additional requirements make an even smaller window. In particular satisfying requirement (1) by the application of pre-coats to HRA in winter, can make the achievement of (a) almost impossible. One solution to this problem is by specifying, or permitting as an alternative, a surfacing material such as Stone Mastic Asphalt (SMA) which does not require such an application.

8. DEFORMATION RESISTANCE Bitumen is a visco-elastic solid which will flow readily under static/slowly applied loads or at high temperatures whilst behaving as an elastic solid at high speeds or low temperatures hence resisting large shock loadings without deformation. Higher


penetration bitumens, ie 100 pen as against 35 pen will have poorer deformation resistance. These properties can be adjusted by the incorporation of a polymer from the vast range available.

Aggregates are not temperature susceptible and under normal circumstances are unaffected by loading.

Mixtures of bitumen and aggregates will behave appropriately depending upon whether the bitumen or the aggregate skeleton is the dominant feature. In HRA binder properties predominate, in macadam mixtures (including SMA thin surfacings and porous asphalt) it is the aggregate unless the mixture is overfilled. Mixtures can also suffer deformation by consolidation of the material under traffic if the initial compaction was insufficient.

Resistance to deformation of base materials can be measured in the Nottingham Asphalt Tester by the relatively new Repeat Load Axial test (RLAT) [BS DD 226.1996] and for wearing courses by the long standing dry Wheel Tracking test. [BS 598 Pt 110.1996].

The deformation of the total pavement is measured routinely by Deflectograph or, for special investigations, Falling Weight Deflectometer. Both these devices need knowledge of pavement layer constituents and thickness for interpretation of their results. Changes in results over time are an indicator of structural performance.

Recent work has shown that where the road surface has rutted, it is necessary to determine whether it is the wearing course and/or the base material which has deformed. This can only be done by cutting a slot across the carriageway. In many cases where the HRA wearing course has been properly designed using the Marshall Stability method but laid on a DBM 100 pen base, it has been the latter which has caused the problem.

The rut resistance of HRA mixtures is very sensitive to binder and filler volumes and bitumen penetration (this may be improved by the addition of polymers), the maximum temperature of the layer when trafficked and the speed of that traffic.


9. FATIGUE This is the resistance of the road materials to cracking as result of repeated loads.

Cracking is caused by tensile forces in the relevant layer. These forces are generated at the top and bottom of the structural pavement by the action of the passing wheel. They can also occur because of warping forces caused by temperature gradients and reflected cracks from an underlying cracked or jointed layer, eg CBM.

Aggregates cannot resist tensile forces. Bitumen, particularly polymer modified material is quite good, at least in its early life. To resist fatigue, therefore, the more bitumen present the better and asphalts are significantly better choices than macadams. Bitumen, however, embrittles and its penetration value decreases with age, becoming less and less effective even if its stiffness and hence load spreading capabilities increases. This effect can be reduced by sealing the pavement and/or reducing the voids to prevent ingress of air.

For thin overlays on minor roads, which may exhibit significant deflection under a relatively few HGV axles per day as they were never formally designed, or for pavements with less than 250 mm of blacktop, ie carrying less than approx. 5 msa, fatigue resistance should not be ignored.

Fatigue resistance can be measured in the Nottingham Asphalt Tester by the relatively new Indirect Tensile Fatigue test (June 1995) and in addition for Wearing Courses by the Yield Strain test [TRL procedure Feb 1995].


10. PERMEABILITY Ideally the tyres of the vehicle should be in contact with the matrix and thus avoid the possibility of aquaplaning, to achieve this aim one must ensure that the water is dispelled either into the mix or to the side of the tyres.

The ability of the mixture to let water and air pass through.is known as permeability It is a combination of the volume of voids and whether or not they are interconnected.

Air embrittles bitumen, reducing its fatigue resistance, its cohesion and its adhesion to the aggregates. Once adhesion has been lost, water rapidly strips the binder from the aggregate especially those with poor initial affinity, eg Flint Gravels.

Some wearing course mixtures, by a combination of good aggregate/bitumen compatibility, thick binder films and a high starting penetration can resist the effect of being permeable sufficient to have a useful life, eg SMA or thin surfacing or porous asphalt. Such mixtures are valuable as they take the water from the surface reducing spray and to an extent absorbing tyre generated noise.

Base mixtures should be designed to have low permeability to prolong their lives. As there is no standard permeability test currently available, this is achieved by a recipe specification of ‘dense’ gradings, ie low theoretical voids in the dry aggregate mass (including filler) and enough binder to just fill them. Alternatively recipes from BS 4987 or 594 may be selected but these are unlikely to optimise the other parameters. In addition good compaction is necessary; experience shows that this is only achieved by a minimum voids specification this can be easily assessed by either PRD or NDT methods.


11. STRUCTURAL STRENGTH The strength of the pavement structure is determined primarily by the stiffness or load spreading capabilities of the layer. It is a combination of aggregate interlock and the stiffness of the binder. One simple way of improving the stiffness of the binder is by the addition of fibres.

The stiffer the layer, the thinner it can be. This principle underlies all analytical pavement design.

Material stiffness can be measured in the Nottingham Asphalt Tester by the Stiffness Modulus test (ITSM) [BS DD 213. 1993, under review Draft 4 current]. It has units GPa. A doubling of GPa from say 2 to 4 leads to a theoretical pavement thickness reduction of 16%, ie about 30 mm. Analytical pavement models take no account of embrittlement.

Macadam mixtures are generally stiffer than Asphalt. Reducing the penetration of the binder increases the stiffness of both macadam and asphalt. Recipe mixtures to BS 4987 contain more binder than is necessary for maximum stiffness in the laboratory, though making compaction easier, lowering voids, improving aggregate interlock, hence increasing stiffness in-situ. This is particularly relevant for small scale roadwork.

Mixtures designed for a minimum stiffness in the laboratory must be accompanied by specification clauses that control in-situ: minimum voids (to prevent overfilling and poor deformation resistance); maximum voids (for durability and to achieve the predicted aggregate interlock); minimum binder volume (for durability) and tight control on binder and grading tolerances to ensure a consistent layer performance. It is possible using BS 4987 tolerances to cover a range of 3.3 GPa to 9.0 GPa.

Voids can be calculated using measurements from a calibrated Nuclear Density Meter (speedy but not very accurate) or from cores, together with the maximum theoretical density measured by


the Rice density test [BS DD 228.1996] recently introduced from ASTM standards.

12. WORKABILITY For most large scale applications, workability is not a serious consideration. unless being laid in adverse weather conditions Most mixtures can be laid satisfactorily for most of the year in the UK. however other parts of the world may not be so fortunate. Thin surfacings and surface dressings are obvious exceptions.

Softer binders, bitumen emulsion and foam bitumen cold mixtures enable sufficient time to be gained to permit storage and compaction. Softer binder, however, has a significant effect on stiffness and deformation resistance. Cold mixtures do not resist abrasion and so are only suitable for bases. mixtures, however some companies are developing products which they hope to use as surfacing materials.

13. SKID RESISTANCE AND TEXTURE This may be defined as the stopping forces generated by the road/tyre interface. Although it was researched in the 1950’s, the application of the UK skidding resistance policy is the primary reason for the excellent accident record in the UK compared to the rest of the world. It is a combination of microtexture, the roughness of the aggregate particles, and the macrotexture, also known as rugocity or overall surface roughness.

Microtexture is measured in the laboratory by the Polished Stone Value test, macrotexture by the texture depth. Texture depth can be measured by the patch test [BS 598 Pt 105], which is the standard, the mini texture meter or the high speed texture meter. These devices have been correlated for HRA but not yet for thin surfacings. This is because these newer surfacing materials have negative texture, interconnected sockets in a plane surface, unlike HRA and surface dressing where the aggregate protrudes above the


surrounding surface known as positive texture. The patch test may overstate the texture the tyre can experience, the laser may not properly go down the sockets.

Microtexture reduces with time as the aggregates polish. Harder aggregates polish more than soft aggregates which wear. The latter may therefore not perform as well in service as their Polished stone value result predicts. Microtexture may be non-existent on new surfaces as a result of the binder coating on the aggregate which will wear off with time SMA benefits from the application of clean grit to speed up this process. It may happen faster on pre-coats because of the greater point contact pressure.

Macrotexture also reduces with time as aggregate embeds, binder flushes up or particles reorientate (until fretting or wear occurs through embrittlement). Softer binders will be more prone to loss of macrotexture. Roads of all speeds need some macrotexture but if minimum tyre noise generation is necessary, both new and retained texture should be controlled.

Overall skid resistance is measured routinely by SCRIM, individual sites are often measured by Griptester and small areas can use the Pendulum. The last has been standardised, the two others are in the process of standardisation. A proper correlation of results needs to be done.

14. ABRASION RESISTANCE With aggregates normally used by UK suppliers this is not a problem on highways. However, it may be necessary to specify minimum abrasion resistance for surface dressing on heavily trafficked concrete roads or a minimum Los Angeles Abrasion Value (maximum 10% fines value) as a sensitive surrogate, to ensure adequate wear ensures an effective microtexture is retained.

15. NOISE


Noise is generated by engine, transmission, wind and tyres. The road surface only affects the last of these. It increases as speed increases but is perceived as significant, both to bystanders and vehicle occupants at all speeds.

Generally the greater the macrotexture the greater the tyre noise emission. However, the shape of the surface profile can significantly affect this. Negatively textured surfaces, for a similar texture depth, are significantly quieter than for example surface dressings or HRA and 20 mm pre-coats - up to 6dB(A) for bystanders and 7dB(A) in the car.

In a recent road trial carried out in Doncaster UK it was found that SMA and Hitex out performed all others in skid resistance and noise emissions.

A thicker layer of open textured material, eg porous asphalt, has been found to lower tyre noise at the road edge. However, the noise reducing properties of this material falls to similar values as thin surfacings after 3 yrs as the pores clog with detritus.

16. SPRAY Spray is not perceived as such a great problem in the UK as in Europe because our more rugous surfaces generate larger droplets that do not generate a fog. Porous Asphalt was created primarily to reduce spray and may still have limited application for this use.

17. FUEL SAVINGS The use of Stone Mastic Asphalts or Porous Asphalt can be beneficial as the Negative textured surfaces offer less rolling resistance and hence use less fuel. There is insufficient data in the UK to quantify this currently.


Table 1 Relative performance of wearing courses (more * the better) Surfacing type

Surface dressing

HRA Porous asphalt

Safepave ULM SMA

Structural nil **** ** * ** *** Rutting nil *** **** **** ***** ****

* Cracking * ***** *** ** *** *** Ride nil *** **** ***** **** **** Suitable texture

** *** ***** ***** **** ****

Skidding - Early - Normal

***** *****

**** ****

*** ****

*** ****

*** ****

*** ****

Noise dB(A)

82 80 74 76 76 76

Spray ** *** ***** **** *** *** Contract risk

*** **** ** *** *** ****

Durability ** ***** **** **** **** *****

Speed ***** ** *** **** *** *** Cost £1.00 £3.50 £6.75 £3.50 £3.80# £3.80

# # Contractors have confirmed that the total cost of a new construction will not exceed that using HRA. The most important decision must be made by the designer, ie "the correct materials in the correct place" and he should consider the overall performance of the ideal mix. By enhancing the product one can have achievement


18. THE FUTURE

Tom Paxton once wrote: It's a long and dusty road, It's a hot and heavy load, And the folks I meet ain't always kind. Some are bad and some are good, Some have done the best they could. Some have tried to ease my troubled mind But I can't help but wonder where I'm bound.

Where is highway engineering going in the future? The author has attempted to show how the road infrastructure has developed down through the ages. Hopefully one can now appreciate that materials change, loadings change, environmental conditions change, consequently, the pavement designs must change. Our method of assessing and predicting performance might also need to change. The NRA in Ireland have developed their 2020 Vision for National Roads: • Reduce Congestion

Emissions reduced by 50% Economic sustainability Air, noise, access improvement in towns/villages Improved safety Dual C’way 6 times safer than single

C’Way Sustainable Construction –

o materials type and source o energy

Integrated Land Use Planning & Transportation o Sustainable plans o Enforce the plans

"Old men shall dream dreams and your young men see visions". My vision for the future of highways in Europe is for new surfacings


which will offer better “value for money” in terms of whole-life costing, surfacings which will be durable, economic, safe and environmentally friendly. However, it will always be necessary to maintain the infrastructure. Maintenance, like the poor, will always be with us. It will be part of the engineer's responsibility in the future to assess materials and predict their performance and estimate the whole life costing. Simulated Whole Life Testing (SWLT) should enable one to achieve better value for money. Maintenance provides the engineer with a wonderful opportunity to use crack retarding fabrics, anti-skid non-spray, quieter surfacings and these are now available in the latest mixes such as Stone Mastic Asphalts. Furthermore, the unacceptable accident statistics must be reduced either by Education, Enforcement or Engineering improvements. As highway engineers we should never be content with an "acceptable level of fatalities". The cost of life is more precious than that expressed by statistics in terms of pounds sterling. 19. CONCLUSIONS What does the future hold? What problems will the highway engineering profession face in the optimisation and exploitation of our natural resources? The author believes that there will be a greater need to achieve value for money. Secondly, the shortage of natural resources will mean that recycling must be considered as a viable option, and finally, environmental effect on the behaviour of materials must be considered and likewise the effect of extraction on the environment must also be considered. European Standards and BS recipe mixtures will satisfy most applications. Where one or more of the above requirements needs enhancement, good design of pavement structure and materials will be necessary.


Good design can save money in the short term and long term yet enhance performance. Designers need to ensure that their intentions will be realised by proper specifications, Quality Assurance Systems and contract monitoring processes and/or performance measurements after a certain period of time.


3

DETERMINATION OF PERIODIC PAVEMENT REHABILITATION

TREATMENT USING PAVEMENT CONDITION ASSESSMENT (PCA) - A

CASE STUDY

Ahmad Kamel Bin Abdul Malik

Senior Manager, Selia Selenggara Engineering Sdn. Bhd.,

1. INTRODUCTION A Pavement Condition Assessment was carried out along FT03 – Jalan Johor Bahru - Endau from Sec 2 (Wadi Hana in JB) to Sec 24 (boundary JB-KT) in June 2004 to assess the structural and functional integrity of the existing pavement and the adequacy of the existing drainage and provision of furniture along the road. This road serves as the main access to the east coast of Johor to Kota Tinggi and to Endau at the border of Johor and Pahang. According to traffic census carried out by JKR in April 2004, the road records high daily traffics volume at 17,176 average number of vehicles per day in both directions with 10.8 percent are heavy vehicles. The road was previously under the maintenance of Johor State JKR. Selia Selenggara Sdn Bhd (now Selia Selenggara Selatan Sdn Bhd) took over the road maintenance through the Privatisation of Federal Road Maintenance Concession in 2001.


Under the Concession, the scope of work of the road maintenance includes:- a. Periodic Maintenance for pavement and non-pavement

b. Routine Maintenance

c. Emergency works

Problem Statement and Objectives

existing carriageway is of flexible pavement construction and made up of combination of 3-lanes dual carriageway from Sec 0 – 15, 2-lanes dual carriageway from Sec 15 – 22 and reducing to single carriageway from Sec 22 onwards. Traffic lights access junctions and interchanges were prominent along the road. Widenings are found at mostly exit and entry lanes from and into the main carriageway.

The road is experiencing various forms of deterioration at stretches along the road length evaluated, affecting its structural and functional integrity and also the drivers’ riding comfort. These defects occurred both before and after the Concession’s takeover date. Selia Selenggara has undertaken several routine and periodic repairs to up-keep the functionality of the road to an acceptable level. However, the maintenance budget approved in the yearly periodic maintenance so far was not able to implement major repair, which seemed to be needed along the road. In order to assess the structural and functional integrity of the road, hence a systematic recommendation for the pavement repair, Selia Selenggara undertook a pavement evaluation programme throughout the road length as follows:-

a. The surface and functional evaluation using Selia Selenggara’s High Speed Network Survey Vehicle to determine surface roughness, rutting and crossfall profiles, surface texture and surface visible defects along the pavement surfaces on all lanes.


b. The structural assessment using Selia Selenggara’s Dynatest Falling Weight Deflectometer System (FWD) at 500m intervals on all lanes in both directions.

c. Coring and Dynamic Cone Penetrometer at representative chainages to determine the thickness of the construction layers, layer materials and CBR of the underlaying sub-grade.

An overall assessment was then made based on the output of these testing programmes and a recommendation was made for the rehabilitation of FT003.

Scope of Case Study

The site Pavement Condition Assessment programme consists of the following methods of testing:-

1. High Speed Network Survey Vehicle to capture as many surface information as possible.

2. HWD testing using Dynatest’s HWD system for the pavement layers structural performance.

3. Coring and Dynamic Cone Penetrometer tests to determine layer thicknesses of the bound and unbound layers respectively.

This case study examines the Pavement Condition Assessment testing programme and test results analysis in determining the pavement rehabilitation treatment.

2. BACKGROUND

PCA methods of functional and structural evaluation involved the used of non destructive and destructive methods of testing. The testing involved the used of specialized equipments such as Network Survey Vehicle and Falling Weight Deflectometer.Selia


Selenggara Engineering Sdn. Bhd. a subsidiary of Selia Selenggara Sdn. Bhd. acquired these specialised equipments in 2002 and has since carried out PCA along the Federal Routes, MHA highways, airports and overseas.

In November 2007, Selia Selenggara Engineering has upgraded its PCA systems to the latest version which significantly improved data collection and post processing efficiency. The system upgrade includes the enhancement of the Global Positioning Systems to an accuracy of sub one meter.

3. METHODOLOGY

A Pavement Condition Assessment was carried out along FT03 – Jalan Johor Bahru - Endau from Sec 2 (Wadi Hana in JB) to Sec 24 (boundary JB-KT) in June 2004 to assess the structural and functional integrity of the existing pavement and the adequacy of the existing drainage and provision of furniture along the road. This road serves as the main access to the east coast of Johor to Kota Tinggi and to Endau at the border of Johor and Pahang. According to traffic census carried out by JKR in April 2004, the road records high daily traffics volume at 17,176 average number of vehicles per day in both directions with 10.8 percent are heavy vehicles.

The site evaluation programme consists of the following methods of testing:-

1. High Speed Network Survey Vehicle to capture as many surface information as possible.

2. HWD testing using Dynatest’s HWD system for the pavement layers structural performance.

3. Coring and Dynamic Cone Penetrometer tests to determine layer thicknesses of the bound and unbound layers respectively.


The existing carriageway is of flexible pavement construction and made up of combination of 3-lanes dual carriageway from Sec 0 – 15, 2-lanes dual carriageway from Sec 15 – 22 and reducing to single carriageway from Sec 22 onwards. Traffic lights access junctions and interchanges were prominent along the road. Widenings are found at mostly exit and entry lanes from and into the main carriageway. The existing pavement surface which are showing distresses are majority in the form of rut, surface cracks and uneven surface. Excessive surface failures in the form of rut at wheel paths appear at most junctions and at slow lanes.

Network Survey Vehicle Survey

High Speed Road Condition Survey was carried out on all lanes in both bound. Roughness (IRI m/km) and rutting derived from the survey showed that the road is in poor to severe conditions especially at stretches where no periodic maintenance were carried out in the past three years. Where maintenance have been carried out in 2001-2004, the surface conditions were mostly acceptable. The average lane roughness value (International Roughness Index, IRI in m/km) ranges from 2.5 – 47.6 m/km (medium to severe) on slow lanes and to 1.5 – 17.79 m/km (low to medium) on fast lanes. The outbound to Kota Tinggi is showing higher values at slow lanes, while rutting are consistently high at slow and middle lanes. The measured average rut depths ranges from 2.00 – 87 mm on Slow Lanes compared to up to 36.9 mm on the middle lanes and up to 31mm on the fast lanes. The excessive rut deemed to cause permanent deformation to pavement structure which the distress is due to excessive loading on the pavement causing permanent deformation of the pavement structure. High maximum values are observed on slow lanes


Falling Weight Deflectometer Tests

A total 88 FWD tests were carried out on both bound at about 500m intervals on all lanes. Back analysis of FWD deflections were carried out using Dynatest’s software showed that the foundation (sub grade) is generally of medium to good condition, with the layer elastic modulus between 49 - 653MPa (up bound) and 55 – 658MPa (JB bound). The corresponding CBR from the DCP are 18 – 53% and 32 – 51% on up and JB bound respectively. The base layer showed consistently low values at average of 487MPa and 434MPa on up and JB-bound respectively. The low modulus of the base layers are reflected in the remaining life assessment, which showed lower than 5-years at low modulus values of the base layers. The bituminous layer modulus did not show good results with a marginally average modulus of 2330MPa and 2700MPa at up and JB-bound respectively. A generally accepted good performance of the bituminous layer is when the modulus greater than 3000MPa. The corresponding remaining life the road showed that about 20.5% of the FWD test points have life lower than 5 years. From the FWD analysis, both JB-bound and Kota Tinggi bound showed poor conditions at places and require immediate attention especially up to the base layer.

Dynamic Cone Penetrometer tests

14 cores were extracted and DCP were carried out at the core holes. The layer structure showed an acceptable thickness in both bituminous and base layers. The bituminous layer thickness ranges from 150 – 317mm while the base layer ranges from 93 – 584mm. However, the core logging showed about 50% of the samples was detached at the layers and Core no.14 crumbled at edges. These are indicative of poor bonding at layer inter-phases and poor bituminous materials.


4. RESULTS

High Speed Road Vehicle Survey

A total of 112 lane-km (or about 461,055.5m2) of paved surface was surveyed by the High Speed Network Survey Vehicle. All data was processed using ARRB NSV software. Outputs from the High Speed Network Survey Vehicle are:-

i) Surface Roughness in International Roughness Index, IRI, m/km

ii) Rut Depth in mm

iii) Sensor Measured Texture Depths (SMTD) at the wheel paths, given in mm

iv) Digital Images captured to evaluate visual condition of road surface and presented in surface defects such as cracks, bleeding etc

The processed data for Roughness (m/km), rutting (mm) and texture depths, calculated at 100m blocks are given in Appendix C. The data interpretations are described in the following sections.

i) Roughness

The roughness values were calculated for all lanes in both directions. The profiles and distribution of roughness along the road were plotted and are shown in Figure 8. Summary of IRI is given in Table 6. Figure 6 shows indicative IRI indicator for pavement.

From Appendix C and Figure 8, the average lane roughness is between 2.05 – 2.95 measured in IRI m/km for all lanes on both bound. However local high values in IRI occur along Sec 7.9 – 14 on the slow lanes of both bound and at Sec 17 – 23 in all lanes of


both bound. As shown in Figure 8, except on slow lanes, it was indicated that the roughness improved very drastically at sections where maintenance have been carried out in 2001 – 2004. Generally these indicate that the overall road surface is experiencing medium to severe roughness when compared to IRI guide in Table 6 and in Figure 6.

ii) Rutting

From Figure 8 and Table 6, it was indicated that rutting measured by the Multi Laser profiler are generally Low to Medium category i.e less than 25mm depth. However, localized sections were observed to experience high and severe rut. These sections are Sec 2 – 7 (slow lanes to JB), Sec 12 – 21 (slow lanes to JB), Sec 2 – 7 (Middle lane to Kota Tinggi), Sec 10 – 18 (middle lane to Kota Tinggi) and Sec 14 – 23 (fast lanes both to JB and KT. These locations agree with the corresponding low FWD back calculated stiffnesses on base and bituminous layers as described in Section 4.2.Detailed calculated rut depths are given in Appendix C.

iii) Texture Depths

The calculated SMTD showed a generally good to medium. An average value of 0.2 – 0.33 mm were observed for all lanes in both bound. However, some locations with lower than 0.2mm SMTD were also found. Table 6 showed the summary of SMTD for all lanes. Detailed calculated SMTD are given in Appendix C.

iv) Other Surface Defects

Other surface defects were also recorded during the High Speed Survey. Table 7 shows the distribution of other surface defects sampled. In general, crocodile cracks and loss of surface aggregates are major contribution to surface defects.

Falling Weight Deflectometer Results


A total 88 FWD tests were carried out on both bound at about 500m intervals on all lanes. Due to variations in pavement response, the contact pressure applied by the FWD varied slightly from test to test. In order to compare test points, the contact pressure was normalized to 707kN/m2.

The output raw data, tabulated deflections, calculated layer moduli, remaining life and overlay requirements are given in Appendix D. The profiles of measured deflections, layer moduli, remaining life and overlay requirements is shown in Figure 9. Table 8 shows summary of HWD survey results.

Back analysis of FWD deflections was carried out using Dynatest’s ELMOD5 software. The results showed that the foundation (sub grade) is generally of medium to good condition, with the layer elastic modulus between 49 - 653MPa (up bound) and 55 – 658MPa (JB bound). The corresponding CBR from the DCP are 18 – 53% and 32 – 51% on up and JB bound respectively. The base layer showed consistently low values at average of 487MPa and 434MPa on up and JB-bound respectively. The low modulus of the base layers are reflected in the remaining life assessment, which showed lower than 5-years at low modulus values of the base layers. The bituminous layer modulus did not show good results with a marginally average modulus of 2330MPa and 2700MPa at up and JB-bound respectively. A generally accepted good performance of the bituminous layer is when the modulus greater than 3000MPa. The corresponding remaining life the road showed that about 20.5% of the FWD test points have life lower than 5 years.

From Figure 9, it is generally observed that Sec 10 – 23 to Kota Tinggi showed good bituminous layer over less performing base layer. Where these occur, the corresponding life is consistently low. On Johor Bahru bound, similar pattern was observed along Sec 23– 10. Immediate attention will be required along these sections


Coring and Dynamic Cone Penetrometer (DCP)

14 cores were extracted during the pavement assessment. Core logging was carried out for each core to determine the thickness of bound layers and the construction material. Core logging report is presented in Appendix E. After coring was completed, DCP tests were carried out at each core hole to determine the thickness of the unbound layers. DCP penetration records, DCP analysis, No. of Blows, Penetration Rate (mm/blow) and CBR% are plotted against the depths of penetration are given in Appendix F. Summary of layer thickness is given in Table 9.

From Table 9 and Appendices E and F, it was found that the layer structure showed an acceptable thickness in both bituminous and base layers. The bituminous layer thickness ranges from 150 – 317mm while the base layer ranges from 93 – 584mm. However, the core logging showed about 50% of the samples was detached at the layers and Core #14 crumbled at the edges. These are indicative of poor bonding at layer inter-phases and poor bituminous materials. It most cases, CBR of the sub-grade showed relatively good material.

5. CONCLUSION AND RECOMMENDATION

Assessment of Remaining Life

Two main criteria normally considered in controlling the structural life of flexible pavement are the fatigue Iife the bituminous material and deformation in the sub grade. Fatigue crackings in the wheel paths and pavement surfaces are related to the tensile stress and strain generated in the bound layer material when loaded by


traffic. Deformation failure in the sub grade is related to the vertical strain generated at the top of the pavement.

The calculated remaining life of the pavement for each lane is tabulated in Appendix D and is summarized in Table 8. As described in Section 4.2 above, about 20.5% of the tested location with HWD has remaining life less than 5 years. These locations are Sec 21- 19, Sec 16 – 14, Sec 7 – 6.5, and Sec 1.5 – 0 of Johor Bahru Bound. Immediate repair shall be needed at these stretches.

Findings

1. Low remaining life of less than 5-years is found at Johor Bahru Bound. The locations are Sec 21- 19, Sec 16 – 14, Sec 7 – 6.5, and Sec 1.5 – 0.

2. In most locations, the foundation (determined from the FWD back-calculation on sub grade modulus) is in good condition having elastic moduli of greater than 50Mpa. However, bituminous layers and the granular road base were of lower values than expected. Treatment will be focused to strengthen these layers. The sections likely to face problems in due time would be Sec 10 – 23 in both bound. Detail demarcation would be needed to determine the area involved.

3. Rutting occur in most slow lanes. These locations shall need attention : Sec 2 – 7 (to KT and JB), Sec 12 – 21 (to JB), Sec 10 – 18 (to KT) and Sec 14 – 23 (to KT and JB);

Recommendation

1. Based on the findings above it is recommended that treatment to enhance the functional properties of the wearing course such as rut resistance, texture and improve roughness value to less than IRI


2.5 Alternative material grading or surface dressing technologies, such as Chip Seal or Stone Mastic Asphalt (SMA), may be used with careful specifications and high quality control. Detailed design on materials and method shall be required.

2. To replace normal bituminous materials surfacing with Stone Mastic Asphalt (SMA) at the main traffic light junctions seemed to be the best option to arrest rutting and diesel spillage at stoppages.

REFERENCES

1. American Association of State Highway And Transportation (AASHTO).

2. Kementerian Kerja Raya – Highway Planning Unit.

3. Jabatan Kerja Raya SPJ 1988.

4. Lembaga Lebuhraya Malaysia (LLM)

5. JKR Arahan Teknik 5/85


4

EVALUATION OF MALAYSIAN ASPHALTIC CONCRETE MIXTURES USING SUPERPAVE AND MARSHALL

MIX DESIGN METHOD

Juraidah Ahmad 1), Mohd Yusof Abdul Rahman 2) , Univeristi Teknologi Mara, UiTM Shah Alam, Malaysia

Mohd Rosli Hainin Universiti Teknologi Malaysia, Skudai, Malaysia

Mustaque Hossain Professor, Kansas State University, Manhattan, Kansas, USA

1. INTRODUCTION In a mix design, regardless of the method used to produce the asphaltic mixtures, the most important element is to develop a well performed pavement with regards to durability, stability, flexibility and resistance to skidding. However, the traditional mix design methods that are empirical in nature do not simulate the actual condition of the pavement and not all roads performed as expected and shows some signs of distress even after early stage of construction due to traffic loading and environmental conditions


(Asi, 2004). These roads have shorter pavement life and developed distress like raveling, undulations, rutting, bleeding and potholes on the surface.

Most developing countries are still working with conventional mixes and at present it is timely for a more thorough and comprehensive technology to replace the conventional method. In a Marshall mix design, there exists a limitation of accuracy in determining the full effects of variations in environmental and loading conditions, as well as material properties and types of pavement performance. However, the effect of environment with respect to temperature is considered not critical in Malaysia because the temperature is consistent throughout the country. Instead, traffic loading and mix design material properties are of much concern and should be thoroughly addressed to develop better performance pavements in Malaysia.

The Superpave mix design method was developed in 1987 with a major research effort to increase resistance to permanent deformation to reduce asphalt pavement rehabilitation and maintenance costs significantly by simulating compaction effort in realtion to expected traffic and addressing all elements of mix design. Sousa et al. (1991) found that the gyratory compaction adopted in the Superpave system is capable of producing laboratory specimens whose volumetric and engineering properties adequately simulate those of field specimens from a wide variety of pavements. This was done by incorporating four elements which are aggregate selection, asphalt binder selection based on climate, compaction based on traffic, and selection of an aggregate skeleton and asphalt binder content based on volumetric properties.

In Superpave mix design system, volumetric properties are used as key indicators of mix quality. According to Kennedy et al., 1994 and McGennis et al., 1995 the major steps in volumetric testing and analysis process are selection of materials, selection of design aggregate structure, selection of design asphalt content and evaluation of moisture sensitivity of the design mixture. However, this research only considers the volumetric properties evaluation of


the mixtures. This paper presents findings and evaluation of the volumetric properties of asphaltic mixtures of similar gradations designed by Superpave and Marshall method. The important properties comprises of percent air voids (AV), voids in mineral aggregate (VMA) and voids filled with asphalt (VFA). At this stage, only moisture induced damage is considered for all mixtures to determine how susceptible these mixtures are with respect to moisture.

2. EXPERIMENTAL PROCEDURE Superpave and Marshall mix design method follows the Asphalt Institute, 2001 (AASHTO TP4) and ASTM D1559 procedures to design asphalt concrete mixes. Gradations of nominal maximum size 12.5mm and 9.5mm were developed for medium to heavy traffic wearing course and the simulation background of the project is based on heavy traffic , categorised as 6 to 30 million equivalent single axle loads (ESALs) with 100 number of gyrations for Superpave mixtures and 75 blows/face for Marshall mixtures to achieve the design density. Unlike in temperate climate countries, zoning of temperature is not considered in this study due to consistent temperature throughout the year in Malaysia. Therefore, as suggested in the Public Works Department (PWD), Malaysia specification, only asphalt binder of penetration grade 80-100 and 60-70 were considered.

3. DESIGN OF ASPHALTIC CONCRETE USING MARSHALL AND SUPERPAVE METHODS

Kajang Rock Quarry granite aggregates were obtained for aggregate of sizes 12.5mm, 9.5mm and quarry dust. Two different types of aggregate gradations were developed and asphalt binder types used were of penetration grade 80-100 (B1) and 60-70 (B2).


Four different types of mixtures were produced from the design matrix consisting of two aggregate gradations of nominal maximum size (NMS) 12.5mm and NMS 9.5mm and two different asphalt binder types. In this study, the design aggregate structure was designed to meet both Superpave and Malaysian gradation limits as shown in Figure 1 and 2 for 9.5mm and 12.5mm aggregate gradations. The aggregate gradation and proportioning of the 12.5mm and 9.5mm mixes are presented in Table 1.

Figure 1. 9.5mm gradation aggregate structure

9.5 mm Nominal Sieve Size - KJG Rock Quarry

0.60

1.18

2.36

4.75

9.50

12.5

0

0.07

50.

15

0.30

0

10

20

30

40

50

60

70

80

90

100

Sieve Size (mm)

Per

cent

Pas

sing

12.5 mm Nominal Sieve Size - KJG Rock Quarry

0.60

0.30

0.15

0.07

5

19.0

0

12.5

0

9.50

4.75

2.361.18

0

10

20

30

40

50

60

70

80

90

100

Sieve Size (mm)

Per

cent

Pas

sing Superpave upper limits

Superpave lower limits

Marshall lower

Marshall upper limits

Gradation


Figure 2. 12.5mm gradation aggregate structure

Table 1. Aggregate gradings

Marshall Mix Design for Asphaltic Concrete Marshall specimens were prepared with blended mineral aggregates at an increment of 0.5% of asphalt binder from 4.5% to 7.0% by weight of mineral aggregate. Three specimens were compacted using 75 blows per face in a cylindrical mould of 100mm diameter for different percentages of asphalt binder. The compacted specimens were then tested for bulk density, stability and flow value to determine the optimum asphalt content (OAC) by taking average of the maximum stability, maximum density and at 4% air voids. The optimum asphalt content is calculated as per Asphalt Institute in MS-2 considering 4% as design air void

Kajang Rock Quarry (KJG) Mixture 12.5 mm

NMS 9.5mm NMS

Metric Sieve (US) Gradation (% Passing)

19mm (3/4in) 100 100 12.5mm (1/2in) 93.0 100.0 9.5mm (3/8in) 80.0 94.0 4.75mm (No.4) 58.0 70.0 2.36mm (No.8) 42.0 52.0 1.18mm (No.16) 23.0 30.0 0.6mm (No.30) 16.0 19.0 0.3mm (0.50) 11.0 11.0 0.15mm (No.100) 6.5 6.0

0.075mm (No.200) 4.0 4.0


content. Table 2 shows the design parameters to determine OAC for Marshall prepared specimens. Higher OAC and VMA values are expected for 9.5mm mixtures due to finer gradation of the mix compared to 12.5mm gradation. The surface area of aggregates in finer mixture needs more asphalt to coat the aggregates The stability and flow values are within the specified limits of the JKR requirement of a durable mix.

Table 2. Design parameters for optimum asphalt content

Superpave Mix Design for Asphaltic Concrete Design aggregate structure is one of the major features in Superpave mix design. The design aggregate structure, when blended at the optimum asphalt binder content, should yield acceptable volumetric properties based on the established criteria. Initially, trial blends were developed for each gradation to select the design asphalt content of the mix. The Superpave mix design compaction method uses the Superpave Gyratory Compactor (SGC) at an angle of 1.250 gyration, having consolidation pressure of 600 kPa and a speed of 30 rpm to compact the specimens and to provide measure of specimen density throughout the compaction procedure. The target compactive effort is based on the 6,000,000 ESALs typical roadway applications which is medium to high traffic and based on three points: the compaction parameters for

Parameter 12.5-B1 MS

12.5-B2 MS

9.5-B1 MS

9.5-B2 MS

Specified Limits (PWD)

OAC (%) 5.6 6.1 6.2 6.4 -

Stability (kN) 10.1 10.2 10.2 10.7 >8kN

Flow (mm) 3.5 3.5 3.2 3.3 2 – 4 mm

VMA (%) 16.4 17.2 17.2 17.6 -

VFA (%) 77 75 75 75 70 – 80


initial compaction (Ninitial = 8 gyrations), design compaction (Ndesign = 100 gyrations) and maximum compaction (Nmaximum = 160 gyrations). The dimensions of each compacted specimens is 150 mm in diameter and approximately between 110 – 120 mm height, depending on the weight of mixtures. At this stage, each specimen is compacted to Ndesign gyrations and the estimated bulk specific gravity (Gmb) and theoretical maximum specific gravity (Gmm) of mixtures were determined. Further evaluation of this data is to determine the estimated binder content, Pb,est to achieve 4% air voids (96% Gmm) at Ndesign. Selection of the design OAC consists of varying the amount of asphalt binder in the design aggregate structure to obtain acceptable volumetric properties when compared to the established mixture criteria based on the SGC specimens with 4% air voids. Values obtained for the OAC of 12.5-B1-SP, 12.5-B2-SP, 9.5-B1-SP and 9,5-B2-SP are 5.1%, 5.3%, 5.4% and 5.7% respectively. The volumetric properties evaluation which consists of VMA, VFA, air voids and dust proportion is one of the major components in determining stability and durability of asphaltic mixtures. The volumetric properties of design mixtures corresponding to OAC of the mixtures along with mix design criteria is as shown in Table 3.


Table 3. Summary of volumetric properties of Superpave mixes

Mix Design Properties

12.5-B1-SP

12.5-B2-SP

9.5-B1-SP

9.5-B2-SP

Criterion

OAC (%) Air Voids (%) VMA (%) VFA (%) %Gmm @ Nini % Gmm @ Nmax Dust Proportion

5.1 4.0 14.7 72.8 87.0 96.1 0.8

5.3 4.0 15.6 74.4 86.7 95.9 0.8

5.4 4.0 15.6 74.4 87.0 95.9 0.8

5.7 4.0 16.375.586.395.80.7

- -

14.0 min 65-75 89% max 98% max

0.6-1.2

Note : B1- asphalt binder penetration grade 80-100; B2- asphalt binder penetration grade 60-70; For design traffic levels >3 million ESALs, (9.5mm) 3/8” nominal maximum size mixtures, the specified VFA range shall be 73-76 percent

4. COMPARISON OF MARSHALL AND SUPERPAVE MIX DESIGN METHOD

The four different types of gradations were designed using Marshall and Superpave method of mix design. Results exhibit different volumetric properties evaluation for similar gradations using different mix design method. This study attempts to evaluate and compare the results and design criteria between Superpave and Marshall mixtures. The volumetric properties are important element in a mix design to ensure that the mix conforms and meets the criteria set by the mix design method. The optimum asphalt content by weight of aggregate showed that less asphalts were needed for Superpave designed mixtures compared to mixtures designed using Marshall method. The OAC for Superpave


mixtures range from 5.1% to 5.7% while Marshall mixtures range from 5.6% to 6.4% asphalt binder. This shows that lower asphalt content is consumed using the Superpave mix design procedure compared to Marshall mix design procedure. The OAC of the mixtures is as shown in Figure 3.

0

1

2

3

4

5

6

7

12.5B1 12.5B2 9.5B1 9.5B2

M ix De s ign

Opt

imum

Asp

halt

Con

tent

(%)

Superpave mix tures Marshall mix tures

Figure 3. Optimum asphalt content of mixtures

Void in mineral aggregate is one of the important design parameters of a mix design. Literature review has indicated that the rationale behind minimum VMA requirement was to incorporate at least minimum permissible asphalt content into mix in order to ensure its durability. Figure 4 shows a comparison of VMA of similar mixtures designed using different method. Results showed that VMA for Superpave mixtures is much lower than Marshall mixtures although these mixtures meet the minimum VMA requirement. The lower VMA values of the Superpave mixtures can generally be contributed to the increased compactive effort by Superpave gyratory compactor. The densification values of the mixtures as shown in Figure 5 is obviously directly related to the compactive effort which also indicates that the aggregate structure packing within the mix is better compared to Marshall compactor.


13.5

14

14.5

15

15.5

16

16.5

17

17.5

18

12.5-B1 12.5-B2 9.5-B1 9.5-B2

Mix Design

VMA

(%)

Superpave Marshall

Figure 4. Voids in mineral aggregate of mixtures

2.27

2.28

2.29

2.3

2.31

2.32

2.33

2.34

2.35

12.5-B1 12.5-B2 9.5-B1 9.5-B2

Mix Design

Den

sity

(%)

Superpave Marshall

Figure 5. Density of mixtures

Moisture Susceptibility Evaluation of Mixtures The moisture susceptibility or the deterioration of hot mix asphalt mixtures due to detrimental influences of moistures is called stripping. Stripping produces loss of strength through weakening of bond between asphalt cement and aggregate which may contribute to rutting and shoving in the wheel paths due to gradual loss of strength over a period of years. To evaluate stripping, specimens were compacted to approximately 7% air voids at optimum asphalt binder content. One subsets of three specimens were tested as control specimens without any conditioning process


while another subsets of three specimens were first saturated between 70-80% before immersing in water for 24 hours at 600C, according to the AASHTO T283 procedures. The ratio of the average tensile strengths of the conditioned (saturated) subset to the average tensile strength of the controlled (unconditioned) subset is the result of the moisture damage that occurred in the mixtures. The specimens were placed between the steel loading strips which were attached to bearing plates of the testing machine. The load was applied to the specimen by constant head rate, at 50 mm/minute and maximum compressive force was recorded until the specimen cracked. The tensile strength values of all the saturated mixtures are lower compared to unconditioned specimens. Results tabulated in Table 4 showed that all mixtures meet the minimum 80% tensile strength ratio (TSR), hence the mix will be able to resist deterioration due to moisture. From the indirect tensile strength values, Superpave mixtures exhibit higher tensile strength values compared to Marshall specimens. The average tensile strength (TS) values for Superpave mixtures of nominal maximum size 12.5mm and 9.5mm is approximately 30% and 20% higher than same mixtures designed using Marshall method. Figure 6 shows the indirect tensile strength of all mixtures tested.


Table 4. Indirect tensile strength and TSR values of mixtures

.

0

100

200

300

400

500

600

700

800

12.5-B1 12.5-B2 9.5-B1 9.5-B2Mix Design

ITS

(%)

Superpave mixtures Marshall mixtures

Figure 6 Indirect tensile strength values of mixes

Mix Design KJG 12.5 B1

KJG 12.5 B2

KJG 9.5 B1

KJG 9.5 B2

Superpave Mixtures

Av (%) 7.2 7.2 7.2 6.5 UnSat

TS(KPa) 626 702 671 661

Av (%) 7.35 6.9 7.0 6.8

Sat (%) 74.0 73.2 74.4 73.6

TS(KPa) 514 616 627 626

Sat

@

600C TSR 82.2 87.7 93.4 94.8

Marshall Mixtures

Av (%) 7.3 6.7 7.3 6.9 Unsat

TS(KPa) 469 464 544 507

Av(%) 7.3 7.2 7.5 7.4

Sat (%) 74.0 73.2 73.4 72.4

TS(KPa) 452 452 464 491

Sat

@

600C TSR 96.4 97.3 85.4 96.8


5. CONCLUSION

This research was conducted to evaluate and compare the Superpave and Marshall mixtures using local aggregates and also how susceptible these mixtures are when induced in moisture. Conclusions drawn based on the findings of this study are :

1. The optimum asphalt content is lower for mixtures designed using Superpave method compared to Marshall mixtures for the same type of gradation.

2. Densification of the mixtures also varies between the two methods. Superpave mixtures exhibit lower VMA values and better aggregate packing using lesser asphalt binder for the mixtures.

3. The Superpave mixtures also exhibit higher tensile strength values however, all mixtures regardless of mix design method are not moisture susceptible, therefore least affected by water.

ACKNOWLEDGEMENT The authors would like to acknowledge Ministry of Science, Technology and Innovations (MOSTI) for funding this research study under the eScience grant.

REFERENCES Asphalt Institute (2001) Superpave Mix Design Series No. 2 (SP-

2), Asphalt Institute Research Center, Lexingon, KY

Asi IM (2004). Role of roads in traffic safety. Traffic safety everybody’s responsibility symposium, Jordan : Hashemite University

Kennedy, T.W., Huber, G.A., Harrigan, E.T. et l., (1994) Superior Performing Asphalt Pavements (Superpave): The Product of the


SHRP Research Program. SHRP-A-410, National Research Council, Wahington D.C.

McGennis, R.B., Nderson, R.M., Kennedy, T.W. and Solaimanian, M. (1995) Background of Superpave Asphalt Mixture Design and Analysis. Report No. FHWA-SA-95-003, Asphalt Institute, Lexington, KY.

Sousa J, Harvey J, Painter L, Deacon J, Monismith C. (1991) Evaluation of laboratory procedures for compacting asphalt-aggregate mixtures. Report SHRP-A/UWP-91-523, University of California – Berkely


5

CRACK PROGRESSION MODELS FOR FLEXIBLE PAVEMENTS

Sugeng Wiyono Universitas Islam Riau, Indonesia

Othman Che Puan, Mohd Rosli Hainin

Universiti Teknologi Malaysia, Skudai, Johor

1. INTRODUCTION The deterioration of paved roads is defined by the damage trend of its surface condition over time. The defects of a pavement surface, which is usually quantified through a pavement condition survey, are classified under three major models of distress, namely; cracking, disintegration, and permanent deformation. The main focus of this paper is on the crack damages because cracking often triggers the application of maintenance treatments and cracking can be the decisive factor in determining the most appropriate rehabilitation option among others. Cracking is perhaps one the most important distresses in bituminous pavements. The development of cracking is considered directly in most mechanistic design procedures and indirectly in most empirical design procedures. A primary bituminous pavement design objective is to minimize cracking. Cracking is a distress that is readily identifiable and universally acknowledged as a sign of pavement deterioration. However, the modelling of cracking is quite complex. There are many factors that can affect the


development of cracks, and once present, the proliferation of cracking may be affected by the same factors, probably the different factor with the other, or a combination of both.

There are many different approaches to analyze and understand bituminous cracking. Cracks may be defined by their shape (e.g., crocodile, block, and linear), their primary causative action (e.g., fatigue, reflection, and thermal), and their location on the pavement (e.g., edge, wheel path). Cracks are often described by the manner they develop (e.g., top-down or bottom-up) and the life of the pavement when the cracks develop. This paper considers empirical models developed by Paterson (1987) and Bennett et al (1995) for three types of crack progressions, i.e. the time of initiation of structural crack, transversal or thermal crack and structural cracks.

The time of initiation of structural crack, measure in years, as suggested by Paterson (1987), is in the form of Eqn 1, i.e.

( )21 4 /

0a YE SNC

icxICX K a e= (1) where ICX is the time of initiation of structural cracking (in years), Kicx is the structural cracking initiation factor, YE4 is the annual number of equivalent standard axle loads (ESAL), SNC is the modified structural number of the pavement and, a0 and a1 are the calibration parameters. For predicting transverse crack (ACT), measure in linear meter per 1000m, Simpson et al (1994) suggested a model as shown in Eqn. 2.

1000CWxACTCRKSPACE

= (2)

Where CW is the width of the paved area (in meter) and CRKSPACE is the spacing between thermal cracks (in meter) estimated from Eqn. 3.


0 10.305 2 10a aCRKSPACE AGE= (3)

where AGE2 is the age of the pavement surface and, a0 and a1 are the calibration parameters. The third model considers in this paper is the prediction of the incremental area of crack as given in Eqn. 4. This model is adapted from Paterson (1987).

1

21 22

0(1 )50 0.5 (1 )50a

a atCRX z z za NEci z z⎡ ⎤= − + + + −⎣ ⎦

(4) where CRXt is the incremental area of cracking at time t (for this study), z is the simoidal model parameter, NEci is the cumulative ESA since cracking initiation and, a0, a1 and a2 are the calibration parameters. Paterson (1987) suggests that z = 1 if TCI < t50; otherwise z = –1, a0

1 = a0 SNCa1(for this study). TCI is the time since cracking initiation and t50 is 50a2 - 0.5a2/a0

1 a2; i.e., time to 50 per cent area cracked (for this study). The models given in Eqns. 1–3 are used in the HDM–IV manual. However, because the basis of the data used to develop the models is different from the Indonesian roads in terms of traffic loading and climate, it is believed that the models are not directly applicable to the analysis of Indonesian highways. Therefore, there is a need to define the parameters that are directly associated with the local environments. In this study, the calibration parameters a0, a1 and a2 in the corresponding models are derived using a simulation model developed and empirical data collected for the environments in Indonesia. The study focused on three types of pavements, i.e. Asphalt Mix on Asphalt Pavement (AMAP), Asphalt Mix on Stabilized Base (AMSB) and Asphalt Mix on Granular Base (AMGB).


2. FIELD DATA COLLECTION PROCEDURE Two sections of main trunk roads in Pekan Baru, Riau Province, Sumatera, i.e., Kandis (5km long) and Sorek (3km long), were used for the collection of information pertaining to the analysis of cumulative standard axles and pavement surface distresses. Each section was subdivided into a subsection of 1km long. Figure 1 shows location of the study area.

SINGAPORE

P.KATEMAN

KEC. RUPAT

P. RUPAT

S E L A T M E L A K A

P.SINGKEP

KAB. KEPULAUAN RIAU

P. SAWANG NABANG

P. PANJANG

P. NATUNA BESARW E S T M A L A Y S I A

SUMATERA BARAT

J A M B I

KAB. INHU

KEC. KAMPAR KIRI

PEKANBARU

KEC. MANDAU

K A B U P A T E N K A M P A R

KEC. KARIMUN

KEC. KUNDUR

P. TEBING TINGGI

KEC. KUBU

KEC. TANAH PUTIH

KABUPATEN BENGKALIS

Bagan siapi-api

KEC. BANGKO

KAB. INHIL

SELAT

PANJANG

KEC. PANGKALANKURAS

DUMAIP. BENGKALIS

KEC. BENGKALIS

KEC. BATAM

Langai

P. BINTAN

P. UJUNG BETINGP. LINGGA

Kuala Enok

KEC. KUALA INDRAGIRI

KEC. PERANAP

KEC. KUANTAN MUDIK

Peranap

KEC. SINGINGISei. Laiak

KEC. PASIR PENYU

KEC. RENGATKEC. GAUNG ANAK SERKA

KEC. MANDAH

Guntung Kateman

KATEMANP. BURUNG

P. MENDOL

P. TUPAN

P. RANGSANG

P. PADANG

KEC. SEI APIT

Rokan kiri

KEC. BUKIT BATU

Palu Panjang

Perawang

Kota TengahTanjung Medan

KEC. RAMBAH

KEC. KEPENUHAN

KEC. KUNTO DARUS

Kota Lama

KEC. ROKAN IV KOTO

TALUK KUANTAN

Langgam

KEC. LANGGAM

Sei BuatanKEC. SIAK SRI INDRA PURA

Pasir Pangarayan Bengkalis

KEC. BANGKINANG

KEC. XIII KOTO KAMPAR

Bangkinang

P. MATAK

P.TEREMPA

KEC. BUNUT

ROKAN N II

Tj. Medang

Sei GaroMinas

Sei Galuh

RIAU PROVINCE MAPE

101° 102° 103° 104°

2°lL

U1°

lLU

0°1°

lLS

NORTH

RENGAT TEMBILAHAN

BENGKALIS

BANGKINANG

BATAM

TANJUNG PINANG

MELACA

JOHOR

Duri

Kandis

Sp. Lago

Sp. Kualo

Sp. Japura

LOCATION MAPE

Figure 1. Location of the study area

The data for each section of the roads was collected three times at an interval of six months. This means that traffic and pavement distresses on all sections were monitored over a period of 18 months. The data gathered during each collection exercise were the traffic characteristics, progression of crack intensities, progression of rutting and potholes, surface temperature, mean monthly precipitation, and pavement deflection using a Benkelman Beam instrument. The pavement structural number is computed using the equation suggested by Paterson (1987).


The pavement where crack existed was cored to evaluate crack mechanisms as well as to establish mix properties. The California Bearing Ratio of the sub–grade was determined using the DCP method. 3. RESULTS In terms of traffic characteristics, both sections of the roads are considered as heavily loaded with traffic since the average ESA was greater than 0.6 million per lane per year or about 1666 ESA per lane per day (Paterson et. al 1987). In general, traffic composition was dominated by commercial vehicles which contributed to about 57% – 63% of the daily traffic. The temperatures of the pavements and atmosphere during data collections were in range of 40°C–55°C and 30°C–33°C, respectively. The average monthly rainfall was 40 mm – 125 mm and the category of the environment tropic region and wet no freezes. The Marshall tests carried out for the cored samples showed that the pavements considered in the study are elastic, weak and low capacity strength. The cored samples indicated that about 80% of the crack mechanisms occurred top–bottom. Table 1 summarises some of the data of ICX, ACT and CRX measured on site and the values resulted from the applications of the equations 1–4 using the default parameters. In general, the differences between the measured and predicted data for each measurement of cracks are significant. This suggests for the need to calibrate all parameters in the existing models to reflect the local environments. Based the data collected, the parameters associated with crack prediction models as given by Eqns. 1–4 that reflect the local environments are summarised in Tables 2–4, respectively. All variables are as defined earlier. The models developed in this study were validated using a new set of data and the respective R2–values are tabulated in Table 5. Based on these R2–values, it may be inferred that calibration parameters derived suit the Indonesian


environments for asphalt pavement crack analysis. Table 1. Tabulation of data observed from the field and calculated values

for comparisons

ICX (year)

ACT (m/1000m)

CRX (%)

Date

Type of

pavement Measured

Paterson (1987)

Measured

Simpson et al

(1994)

Measured

Paterson (1987)

Mac 2002

1.5 3.7182 0.4950 0.4421 0 0

Sept 2002

0.6091 0.6579 1.78095 1.78090

Mac 2003

AMAP (Kandis Section)

1.0104 1.4502 3.2746 23.6307

Mac 2002

2.5 3.7182 0.06964 0.14914 1.00310 1.18797

Sept 2002

0.20714 0.22363 3.24964 19.0242

Mac 2003

AMSB (sorek

Section)

0.27767 0.31191 6.64695 86.5511

Mac 2002

3.5 3.7182 0.16888 0.22306 0.86850 0.68509

Sept 2002

0.36520 0.75674 3.87280 7.72444

Mac 2003

AMGB (Sorek

Section)

0.59539 0.93366 5.48571 41.8153

Table 2. Parameters for structural cracking initiation

Model ( )2

1 4 /0

a yE SNCicxICX K a e=

A0 a1 Kicx Pavement Type This

study Paterson (1987)

This study

Paterson (1987)

This study

Paterson (1987)

AMAP 9.48 8.61 -25.8 -24.4 0.75 to 2 0.5 to 2.5

AMSB 9.17 8.61 -25.1 -24.4 0.43 0.5 to 1.3


AMGB 8.8 8.61 -24.8 -24.4 0.49 -

Table 3. Parameters for transversal or thermal crack

Model 0 10.305 2 10a aCRKSPACE AGE=

Value of a0 Value of a1

Climatic Zone This study

Simpson et al., (1994)

This study

Simpson et al., (1994)

Wet-No Freeze -0.571 to 0.856 -1.12 3.23 to

3.89 -

Table 4. Parameters for predicting structural crack progression

Table 5 Statistic Value of validate progression structural crack and thermal crack

R2 Value for Structural Crack

R2 Value for Thermal Crack

AMAP New 0.8765 AMAP Existing

0.8264 0.823

AMGB 0.5925 0.879 AMSB 0.8158 0.751

4 CONCLUDING REMARKS The findings of this study may be summarized as follows:

Model 1

21 22

0(1 )50 0.5 (1 )50a

a atCRX z z za Neci z z⎡ ⎤= − + + + −⎣ ⎦

Parameters Model Statistics A0 a1 a2 CV (%) R2

\ This

Study Paterson (1987)

This Study

Paterson(1987)

This Study

Paterson (1987)

This Study

Paterson (1987)

This Study

Paterson(1987)

AMAP 1895 3330 -5.22 -4.25 0.27 0.25 4.3 54 0.845 0.314 AMGB 2855 3330 -4.65 -4.25 0.26 0.25 0.23 54 0.593 0.314 AMSB 2650 3330 -4.24 -4.25 0.25 0.25 0.28 48 0.816 0.304


(i) In general, each calibration parameter set in the models

used in HDM–IV is constant for all types of pavements. However, this is not the case for the Indonesian highways. The result shows that each calibration parameter varies with types of pavement.

(ii) Compared with the default values given in the HDM–IV, for crack initiation, the difference in the coefficients is in the range of 5.7% to 20% for AMAP, 2.8% to 16.2% for AMSB and 1.6% to 2.2% for AMGB. In crack progression model, the difference in the coefficients is in the range of 8.0% to 75% for AMAP, 4.0% to 16.6% for AMSB and 0% to 25.6% for AMGB.

(iii)The models developed are able to predict the progression of cracks with reasonably good accuracy, i.e the R2–values for the models are in the range of 0.59 to 0.88.

REFERENCES Bennett, C.R , Hoban, C.J., and Covarrubias, J.P. (1995). Modeling

Road Deterioration and Maintenance Effects. International Study of Highway Development and Management Tools. N.D. Lea International Ltd. Canada.

Nagel and Schreckenberg, M. (1999). Microsopic Simulation of Urban Traffic Based on Cellular Automata. International Journal of Modern Physics, Part C. 8 (5): 1025-1036.

Paterson, W.D.O (1987). Road Deterioration and Maintenance effect: Models for Planning and Management. The Highway Design and Maintenance Standards Series. Baltimore. Maryland. USA.: The John Hopkins University Press.


Simpson, A.L et al (1994). Sensitivity Analyses for Selected Pavement Distresses. SHRP-P-393. Strategic Highway Research Program. National Research Council. Washington D.C.


6

INVESTIGATION ON TYRE/ROAD NOISE USING ULSTER LOAD TYRE

ROAD ASSIMILATOR

Haryati Yaacob, Mohd Rosli Hainin Universiti Teknologi Malaysia

David Woodward, Alan Woodside

University of Ulster, UK

1. INTRODUCTION The first measuring standards for tyre/road noise were established in 1973 by the Japanese Automobiles Standard Organisation (JASO). These were later revised in 1981 (JASO, 1981). This standard specified a coast by method and a laboratory drum method. In the UK, the influence of different road surface types on traffic noise has been taken into account by the Calculation of Road Traffic Noise (CRTN) prediction method (Department of Transport and Welsh Office, 1988).

The influence of traffic noise was based on the average sand-patch texture depth used to specify requirements for skidding resistance. In recent years considerable research has been conducted to develop improved methods of predicting noise levels from surface parameters. Some have considered laser-based surface texture profile measurement systems (Phillips and Kinsey, 2000) and mobile methods for measuring the sound absorption of road surfaces (Morgan, Watts and Phillips, 2001).


These texture-based techniques have increased the accuracy of noise predictions, but are not yet sufficiently developed for use as a means of discriminating between different surface types. Since they are empirical in nature they are not considered reliable for the characteristics of noise for new types of surfaces.

An alternative approach, currently favoured in the UK, involves the direct measurement of road surface noise. Direct techniques utilise measurements taken either at a distance from the vehicles tyres, normally at the roadside, or in close proximity to special test tyres. The on-site measurement of tyre/road noise is affected by many factors. Review of the limited available data concluded that this was not sufficiently rigorous to accurately categorise surfacings with respect to one another due to possible variations with on site measurement conditions.

To overcome these drawbacks, the internal drum test equipment can be used. This type of facility is limited to research laboratories. Typically the drums have diameters of 1.5 to 2.5m.

It is necessary to have the drum fitted with a surface that resembles an actual road. It has been found that indoor drum tests on simulated surfaces which have been carefully manufactured as true replicas of actual road surfaces give good correlation with pass-by tests (Ejsmont, 1982).

The drum test have the advantage that rapid, controlled tests can be made under laboratory conditions (Walker and Williams, 1980). A comparison of road and laboratory measurements was done by Ejsmont (1982). A special trailer for road measurement and a drum facility was built. The drum was covered with a replica road surface. Analyses of correlations between road and laboratory measurements were performed. As the correlation coefficients generally were high, it was concluded that it was possible to replace road measurements by laboratory ones.

The drum method is suitable where high accuracy is needed. Investigation in noise emission from a large number of tyres under various operating conditions can be carried out in a relatively short


time. The method is useful for research and development work and for detecting small differences in noise emissions from different tyres. It is independent of weather conditions and requires little space and only one tyre per sample test. Long measuring times can be used to reduce errors. Speeds can easily be varied over a wide range without safety problems.

This paper will discussed findings from new method developed using a rotating internal drum equipment at University of Ulster, UK.

2 DEVELOPMENT OF THE NOISE MEASUREMENT METHOD USING ULTRA

The Ulster Load Tyre Road Assimilator (ULTRA) apparatus is an internal drum designed to simulate trafficking conditions. It was originally developed at the University of Birmingham by A.R. Williams in 1971 to measure the skid resistance of aggregate in the laboratory.

The diameter of the drum is 1.12m and it can be fitted with a tyre size of 2.25 –8. The drum has variable speed from 0 to 112kph. Fifteen test sections fill the drum, each section being 30mm thick, 240mm long and 127mm wide. This is more than twice the width of the contact patch of the 2.25-8 tyre.

Preparations of On - Site Road Surface Peels

A method was developed to make latex peels of different types of asphalt surfacing. Each site was first brushed to remove loose detritus and then assessed to determine that it was representative of the road section. The latex peels were all taken from the in-side wheel-path.


Latex was heated in an oven until it had a free flowing constituency. A rectangular surround with internal dimensions 250 x 140 x 5mm was placed on the road surface. Sufficient hot latex was poured to fill in the surface texture flush with the top of the rectangular surround.

The latex was allowed to cool for a few minutes after which it could be peeled from the surface. This gave a rectangular latex negative copy of the road surface. Three peels were taken for each road surface. The texture depth of each road surface was determined using the sand patch method.

Preparation of Road Surface Test Specimens The ULTRA machine has an internal rolling road test surface that requires concave test specimens. These were made using specially manufactured steel moulds that could be dismantled to remove the finished test specimen. The latex peel obtained from the road surface was trimmed to fit the internal dimensions of the steel mould i.e. 240 x 125mm. The textured surface of the latex peel was placed up.

Two sheets of aluminium mesh were fixed within the steel mould to strengthen the final test specimen. A rigid backing plate was securely fixed. This had a central hole through which a two-part Nitomotar PE resin was poured. Use of this type of hard resin remove the effect of variables associated with rock type, aggregate wear, aggregate polishing, redistribution of bitumen coatings and other changes in test surface texture during testing.

The resin was poured through the central hole of the backing plate. It was sufficiently fluid to in-fill around the exposed surface of the latex peel. Resin was added until no more could be poured through the hole. The mould was subjected to a small amount of vibration to help remove any air bubbles present.

After hardening the mould was stripped to expose the test specimen. The latex peel was removed from the resin test


specimen to reveal an exact copy of the road surface texture. The same peel could be reused to make the required 15 test specimens.

Mounting Test Specimens on the ULTRA Fifteen curved test specimens, for each surface, were mounted on the ULTRA machine. The test specimen is shaped at the sides to allow it to be securely fixed to the inside of the drum using bolted metal straps. This gave a continuous road surface of test specimens separated by transverse joints. Care was taken to ensure that the joints were of regular spacing and there was no height difference between adjacent test specimens that could induce unwanted vibration.

Each set of 15 test specimens was conditioned for two hours prior testing to ensure that each textured surface had received a similar degree of simulated trafficking. This conditioning period was stopped after 15-20 minutes and the bolts fixing the locating straps re-tightened. A further check was made prior to noise testing.

Tyres details The original tyres selected for the drum machine were no longer available. Rather, a commonly available and cheap tyre was chosen for the investigations carried out in this research. The tyres chosen were Maxis SLC kart racing slick and wet 10 x 4.50-5 tyres. These had the advantage of being a similar size to the GripTester tyre used for measuring road surface skid resistance.

An unfortunate limitation was their maximum inflation pressure of 4kg/cm2 or 56.8psi. The slick tyre was smooth whereas the wet tyre had a pronounced tread depth of 6mm. Nitrogen was used to inflate the tyre to the required test pressure. During prolonged periods of testing, air tends to expand and may decrease slightly the size of the contact area. The ratio of the small test tyre with the drum diameter resulted in minimal curvature effects.


Microphone Positions

A single microphone using a Cel-593 Sound Level Analyser was used to take the noise measurements. The position selected was in front of the rolling tyre at an angle 45º to the rolling direction, 100mm above the contact surface and 200mm from the un-deflected sidewall of the tyre as shown in Figure 1. This is similar to the position used in the CPX method (ISO/CD 11819-2).

Initial testing had shown that the noise recorded in front of the rolling wheel was slightly different than that recorded behind the rolling wheel. However, it was not possible to position the microphone behind the rolling tyre because of the loaded wheel assembly.

Figure 1. Microphone position used in the ULTRA noise measurement method

h=100mm

Tyre seen from above

d2= 200mm

d1=200mm

Microphone

Microphone

45º


Experimental Conditions and Noise Measurements

Eight types of surfacing were selected and latex peels obtained from actual road surfacings. These are detailed in Table 1. They were chosen to have a range of factors including surfacing type, texture depth, texture type, trafficking condition and age. An additional smooth surface was made from the resin to look on the effect of joints. The smooth steel internal running surface of ULTRA was also assessed to simulate a continuous smooth surface with no joints. Table 1. Details of road surfaces used for noise measurement

Type Age Rock type Texture depth

Grip Num

20mm HRA

2 years

PSV>60 gritstone

2.00mm 0.65

14/6mm SD

1 week

PSV>60 gritstone

3.24mm 0.6

10/6mm SD

1 year PSV>60 gritstone

0.97mm 0.74

10mm SD

1 week

PSV>60 gritstone

2.33mm 0.75

6mm SD

2 years

Blend of 2 PSV >60 aggregates –gritstone + igneous

1.12mm -

14mm SMA

2 years

PSV>60 gritstone

1.50mm 0.65

10mm bitmac

1 week

PSV>60 mm gritstone

1.03 mm

0.54

10mm Thinpave

<1 year

- 0.62mm -

Smooth resin

NA NA NA -

Smooth steel

NA NA 0mm -


The noise measurement tests were carried out under one-third octave sound pressure level for one minute duration. The one-third octave band spectra gives a reasonable frequency resolution and does not contain so much data that it is difficult to make evaluations. The measurements were recorded for each tyre / pressure / speed / loading combination across the 12Hz to 20kHz range. This is similar to the CPX method. Both a smooth (slick) and treaded test tyre was used.

A typical testing cycle consisted of:

(i) Set the required tyre pressure.

(ii) Mount the required load.

(iii) Lower the loaded tyre unto the test specimen surface.

(iv) Bring the ULTRA machine to the required speed.

(v) Measure road/tyre noise for 1 minute

The test variables are given in Figure 2. This shows the mass of the weights loaded unto the test wheel assembly, the tyre pressure and speed of the drum as it rotated. It should be noted that this mass is not the total loading applied through the contact patch of the tyre.

Table 2 gives the actual total loading found by McErlean (2006). This is a combination of the individual weights and the weight of the test wheel assembly. Subsequent analysis of the test data uses the values given in Figure 2.


Figure 2. List of variables for noise measurement

PHASE 1

Ulster Tyre Road Assimilatpr

Surfaces (6mm SD, 10mm SD, 14/6mm SD, 10/6mm SD, Smooth Resign,

Smooth Steel, Playsafe, 20mm HRA, Thinpave, 14mm SMA, 10mm Bitmac)

Loading Speed Tyre Type Tyre Pressure

80kg 40kg 20psi 30psi

50km/hr

100km/hr 40psi

95 km/hr *Playsafe only

65km/hr *Playsafe only

80 km/hr *Playsafe only

Smooth Tyre

Treaded Tyre


Table 2.Total loading applied to the contact patch (McErlean, 2006)

Load applied to tyre assembly (kg)

Actual load applied under contact patch (kg)

0 27

40 67

80 107

3 RESULTS

Given the number of surface type and test variable combinations a considerable amount of noise data was recorded. These were downloaded into the Cel 593 -Sound Track – dB2 software sound analyser and then converted into Excel for analysis.

Relationships between Tyre/Road Noise and Texture Depth Figures 3 and 4 summarise the texture depth and tyre/road noise for all of the surfaces using a smooth and treaded tyre. It shows measured Leq at 50 km/hr, loading of 40kg and tyre pressure at 20psi. Texture depth was measured using Sand Patch test. In general terms, there appears to be a relation between noise and texture depth.

The linear trends show weak correlation suggesting that increasing texture means increasing tyre/road noise. However, this can only be considered as a generalised assumption. For example, the 10mm surface dressing was quieter than HRA although the surface dressing texture was greater. This phenomenon has been found by previous studies e.g. Franklin et al (1979) and Sandberg and Ejsmosnt (2002).


Correlation between tyre/road noise and texture depth appears to be best when using tyres which have a shallow or non-aggressive tread pattern. However, when using tyres with a winter or aggressive tread pattern, the correlation is much poorer or else does not exist.

The single parameter of texture depth appears to be not good enough to explain tyre/road noise for all the surfaces assessed. The findings suggest that reliance on the specification of texture depth may not always ensure a quieter or noisier road surface.

However, within the same road surface type increasing the maximum chipping size generally means increased tyre/road noise. This is important as its suggests that this new laboratory technique appears to simulate what other researchers have found and is able to rank the performance of road surfaces.

y = 1.657x + 101.9R2 = 0.4666

y = 1.9075x + 91.483R2 = 0.691990

95

100

105

110

0 1 2 3 4

Texture depth (mm)

Road

/tyre

noi

se (d

B)

50 km/hr 100 km/hr

Figure 3. The relation between tyre/road noise and texture depth at load 40kg and pressure 20psi using a smooth tyre


y = 0.6955x + 104.71R2 = 0.3042

y = 0.6826x + 94.395R2 = 0.4390

95

100

105

110

0 1 2 3 4

Texture depth (mm)

Roa

d/ty

re n

oise

(dB

)

50 km/hr 100 km/hr

Figure 4. The relation between tyre/road noise and texture depth at load 40kg and pressure 20psi using a treaded tyre

The Effect of Surface Texture on Tyre/Road Noise Test specimens made from resin were assessed to determine the effect of no road surface texture on tyre/road noise. Figure 5 shows data for the smooth resin surface compared to the other surfaces. When using a smooth tyre, this shows a gap between the smooth resin surface and the other surfaces at the frequency range 800 to 2500Hz.

Within this frequency range, the textured surfaces were approximately 5 to 10dB(A) noisier than the smooth resin surface. However, at the lower frequency range of 100 to 500Hz and the higher frequency range of 5000 to 10000Hz the smooth resin surface plotted within the other surface data set.

Changing to the treaded tyre, Figure 6 shows that the smooth resin surface is 5 to 10dB(A) higher than the texture surfaces only in the frequency range 2500 to 3150Hz

The data shows that a smooth textured surface is not necessarily the quietest. The effect of different texture surfaces sometimes


result in a surface that is quieter or noisier than a non-textured or smooth surface, depending on the tyre/road noise mechanisms that are at work. The different noise generation mechanisms appear to be related to different textured surfaces with different tyre types. The overall influence is governed by the generation mechanism that dominates for that particular case.

30

40

50

60

70

80

90

100

10 100 1000 10000 100000

Frequency (Hz)

Soun

d le

vel (

dB)

Figure 5. Noise differences between textured surfaces and smooth resin

using a smooth tyre.

Smooth resin Sample with texture


30

40

50

60

70

80

90

100

10 100 1000 10000 100000Frequency (Hz)

Soun

d le

vel (

dB)

Figure 6. Noise differences between textured surfaces and smooth resin

using a treaded tyre.

The Effect of Joints

The effect of joints between test specimens was evaluated. Figure 7 is a plot of the sound level data using the smooth and treaded tyres for the internal smooth steel surface of the ULTRA and the smooth resin test specimens separated by joints.

The resin with joints surface was found to be noisier than the steel surface. The treaded tyre was found to be noisier than the smooth tyre. For these two types of surface, the effect of joints for both tyres was greater at lower frequencies i.e. the range where the impact mechanism is important. The joints around the ULTRA drum appear to induce additional vibration.

This makes it very important to ensure that the test specimens are properly made i.e. to the required height and that they are securely fixed to the test equipment prior to testing. In terms of their effect on the data obtained, their effect is assumed to be similar to all of the surfaces assessed.

Smooth resin Sample with texture


40

50

60

70

80

90

100

10 100 1000 10000 100000Frequency (Hz)

Soun

d le

vel (

dB)

Resin with joints (smooth tyre)Smooth steel (smooth tyre)Resin with joints (treaded tyre)Smooth steel (treaded tyre)

Figure 7. Noise difference measured between a smooth steel surface and a smooth resin surface with joints using smooth and treaded tyres

The Effect of Speed In the earlier investigations, only two speeds were used i.e. 50km/hr and 100km/hr. The results in Table 3 shows that tyre/road noise increases approximately 10dB with the doubling of speed from 50km/hr to 100km/hr.

The effect of speed for both smooth and treaded tyre can be seen stronger in the frequency range of 800 to 2500Hz. The values are similar to Sandberg (1984) who found that tyre/road noise will increase about 9 to 12 dB (A) per doubling of speed.


Table 3. Noise difference after increasing speed from 50km/hr to 100km/hr Surface / tyre type Noise at

50km/hr (dB) Noise at 100km/hr (dB)

Noise Difference (dB)

10Bitmac (smooth) 91.7 101.3 9.6

10Bitmac (tread) 96.3 106.6 10.3

14SMA (smooth) 95.2 105.5 10.3

14SMA (tread) 94.3 104.7 10.4

14/6SD (smooth) 97.6 106.9 9.3

14/6SD (tread) 97.0 107.9 10.9

10SD (smooth) 95.0 104.4 9.4

10SD (tread) 95.6 104.9 9.3

10/6SD (smooth) 93.2 104.7 11.5

10/6SD(tread) 95.1 105.9 10.8

6SD (smooth) 93.0 102.8 9.8

6SD (tread) 94.7 105.2 10.5

HRA (smooth) 96.8 107.6 10.8

HRA (tread) 96.0 106.4 10.4

A further investigation on the effect of speed was carried out using Playsafe (PS) and Thinpave (TP). These are two examples of propriety thin surfacing. The Thinpave test specimens came from an actual road whereas the Playsafe test specimens were prepared using a latex peel taken from a laboratory prepared slab.

Five speeds were evaluated using the two surfaces i.e. 50, 65, 80, 95 and 100km/hr. Testing also considered 20, 30 and 40psi tyre pressure. Both treaded and smooth tyres were used. The results are


shown in Figure 8 . Analysis of the data found that tyre/road noise level increases logarithmically with speed.

However, there was a difference depending on tyre type. The smooth tyre data set plotted approximately 2dB lower than the treaded tyre noise-speed data set. The results using the smooth tyre showed a bigger difference between each test combination especially in the lower speeds of approximately 2-3dB.

92

94

96

98

100

102

104

106

108

40 50 60 70 80 90 100 110

Speed (km/hr)

Noi

se d

ata

(dB

)

Smooth tyre

Treaded tyre

Figure 8. Noise data for Playsafe and Thinpave at different speeds using a smooth and treaded tyre

The Effect of Loading and Tyre Pressure According to Sandberg and Ejsmont (2002) the change of noise for a loaded truck of 13.23 tons in comparison to an empty truck of 5.58 tons was as high as 6.5dB(A) for traction tyres but only 0.5dB(A) for ribbed tyres. The ribbed tyres were known to give less noise compared to tyres with an aggressive tread pattern. However, in the ULTRA the investigation on the effect of similar


loading and tyre pressure can not be done due to the limitation of maximum load and inflation pressure of the tyre.

The loading influence is known to be much higher at lower speeds than at higher speeds (Sandberg and Ejsmont, 2002). This trend was not found using the ULTRA tyre/road noise measurements. The effect of higher loadings of 80kg and 100km/hr speed could not be done due to safety reasons related to tyre loading limitations.

Figure 9 shows that by doubling the weight from 40 to 80kg, the tyre/road noise was found to vary from only 0 to 1dB higher or lower

Figure 10 shows the effect of tyre pressure on tyre/road noise. Similar to loading, tyre pressure does not influence tyre/road noise in a consistent way. The 10Bitmac and 10SD surfaces appear to have reduced tyre/road noise with increased tyre pressure whereas 10/6SD, 14SMA and HRA were noisier. The exact behaviour of noise with varying load and tyre pressure appears to be a complicated function of tyre type, speed and road surface.

90919293949596979899

100

6SD

10/6S

D

10Bitm

ac10

SD

14/6S

D

14SMA

HRA TP PS

Soun

d le

vel (

dB)

40 kg 80 kg

Figure 9. Effect of loading at 50km/hr and 20psi for a smooth tyre


86

88

90

92

94

96

98

100

6SD

10/6SD

10Bitm

ac10S

D

14/6SD

14SMA

HRA TP PS

Soun

d le

vel (

dB)

20 psi 30 psi 40 psi

Figure 10. Effect of tyre pressure at 50km/hr using a smooth tyre

4 CONCLUSIONS In this paper , factors such as different types of surface with differing amount and type of texture depth, loading, tyre type, tyre pressure, speed and joints were considered. It was found out that the biggest factor to tyre/road noise was speed, followed by surface texture, tyre pressure, loading and the presence of joints.

When relating tyre/road noise with surface texture, it was initially assumed that higher texture depth would result in higher tyre/road noise. However, no relation was found between tyre/road noise and texture depth when using a treaded tyre. A weak linear relation was established when using a smooth tyre.

These findings support Swedish studies reported in Sandberg and Ejsmosnt (2002). The first study was made 1977 to 1979 and involved three types of tyre i.e. a smooth tyre, a summer tyre and a


winter tyre. A second study took place in the CPX experiment of 1988 to 1999. The findings in this research confirm both the Swedish studies where a less aggressive tread pattern results in a weak noise / texture relation whilst an aggressive tread pattern results in no correlation with texture depth.

These findings show that new test method developed using the ULTRA apparatus is capable in quantifying tyre/road noise for different type of surface with the findings similar to reported research. The method provides the highway industry and specifying bodies a laboratory means of both ranking noise and for understanding the mechanisms involved in its generation.

REFERENCES Department of Transport and Welsh Office 1988, Calculation of

Road Traffic Noise (CRTN), Her Majesty's Stationery Office, London, UK.

Ejsmont, J.A. 1982, Comparison of road and Laboratory Measurements and Influence of Some Tire Parameters on Generation of Sound, Swedish Road and Transport Research Institute, Linkoping, Sweden.

Franklin, R.E., Harland, D.G. & Nelson, P.M. 1979, Road surface and traffic noise, Transport and Road Research Laboratory, UK.

JASO 1981, Test Procedures for Tire Noise-Japanese Automobile Standard JASO C606-81, Japanese Automobile Standards Organization, Japan.

McErlean, P. 2006, A study of the relations between texture depth, rolling resistance and noise for highway surfacing materials., University of Ulster.

Morgan, P.A., Watts, G.R. & Phillips, S.M. 2001, "Trials of a mobile MLS technique for characterising road surface absorption", Inter-Noise 2001Congress Secretariat, The Netherlands, pp. 2063-6.


Phillips, S. & Kinsey, P. 2000, Advances in identifying road surface characteristics associated with noise and skidding performance., TRL Limited, Crowthorne, UK.

Sandberg, U. & Ejsmont, J.A. 2002, Tyre/Road Noise References Book, INFORMEX Ejsmont and Sandberg Handelsbolag, Sweden.

Sandberg, U. 1984, "Reduction of tyre/road noise by drainage asphalt", International Seminar on Tyre Noise and Road ConstructionNational Swedish Road and Traffic Research Institute, Linkoping, Sweden.

Walker, J.C. & Williams, A.R. 1980, "The improvement of noise and traction due to tyre/road interaction", International Tire Noise ConferenceStyrelsen Foer Teknisk Utveckling, Stockholm, Sweden, pp. 261-71.


7

PERFORMANCE OF HOT MIX ASPHALT USING FINE CRUMB RUBBER

Norhidayah Abdul Hassan, Mohd Rosli Hainin, Haryati

Yaacob Universiti Teknologi Malaysia

1. INTRODUCTION The concept of modifying asphalt mixes is not new in fact, since years ago there have been numerous efforts to modify asphalt mixes in order to get a better performance. The use of crumb-rubber modifier (CRM) in hot mix asphalt can be traced back to the 1840s when natural rubber was introduced into bitumen to increase its engineering performance (Heitzman, 1992). However it was not thoroughly discovered until the late of 1980s when people start to realise about the need to improve the conventional asphalt mixes and recycled tire crumb rubber became one of the alternative materials (Epps, 1994). Initially, only coarse rubber was being used in dry process. However, experience with the mix indicated better durability with an increase of fine rubber content. Hence, after 1981, 20% of the originally used coarse rubber was replaced with fine rubber (passing 850 μm sieve) (Esch, 1984).

Takkalou et al., (1985) reported that the required asphalt content is 1.5 to 3% higher than the conventional mixes with similar size and


type of aggregates. Koh and Talib, (2006) also agreed that rubber modified asphalt concrete (RUMAC) required higher binder content as the percentage of crumb rubber increased. Elliot, (1993) stated that the effect of CRM on the optimum bitumen content (OBC) and volumetric properties is significant for RUMAC mixes with 3% CRM. Studies by Troy, Sebaaly and Epps, (1996) discovered that gap graded CRM mixes. had lower Marshall stabilities than dense-grade CRM mixes. Rutting is a flexible pavement distress caused by the accumulation of permanent deformation in the pavement layers due to the repeated application of traffic. Stroup Gardiner and Krutz, (1992) discovered that the addition of CRM by using dry process does enhance the rutting resistance of the mixes at higher temperatures. Similarly, Rebala et al., (1995) stated the used of CRM in the dry process allows it to serve as discrete particles which may enhance the rutting resistance. While Koh and Talib, (2006) found that rutting of asphalt mixes at 2,000 load cycles was reduced by 22% with the addition of 3% crumb rubber. Another study did by Troy, Sebaaly and Epps, (1996) discovered CRM pavement sections done in Louisiana exhibit similar or lower rut depth than the control sections after five to seven years in service. However, Takkalou et al., (1985) stressed out that performance evaluation is significantly dependent on the crumb rubber gradation, air voids, aggregate gradation, mixing temperature and curing conditions.

Problem Statement

With the increase of traffic loading and number of heavy vehicles, many pavements tend to fail prematurely either structurally or functionally even though they have been designed to last longer. Repeated application of traffic loads can cause structural damage to asphalt pavement which can cause permanent deformation particularly rutting along the wheel tracks. This kind of damage


getting worst especially in hot climatic like Malaysia. Government has spent millions of ringgit to repair and maintain roads in our country. Development of modified asphalt mixes has been explored over the past few decades in order to improve the performance of pavement mixes. Heitzman, (1992) and Epps, (1994) claimed that incorporation of crumb rubber into asphalt mixes will make the mixes more elastic at higher temperature thus enhancing their rutting resistance. Crumb rubber obtained from used tires has been the focus of several research efforts on the purpose of overcoming pavement problem and also helps in recycling mountainous dumping tires at landfill. However, the degree of improvement and the cost effectiveness of using crumb rubber in asphalt mixes have not been firmly established. In most of developed countries such as the United States also facing the same problem because of very limited information available on the effectiveness of using crumb rubber in asphalt mixes and they are not well documented. Although many studies have been performed to investigate the used of crumb rubber in modifying the hot mix asphalt mixes, conflict results have been obtained. These could be due to different devices used, different testing environments, and also the size of the experiment conducted. Thus, there is a need to conduct a study to evaluate the performance of HMA after being modified using crumb rubber according to Malaysian mix design.

Aim and Objective

This study was aimed to investigate the effect of adding crumb rubber on the properties of hot mix asphalt by using dry process according to the Malaysian mix design. The primary objective of this study was to evaluate the rutting resistance of the crumb rubber modified asphalt mixes compared to conventional asphalt mixes. In achieving that, all samples prepared were evaluated in terms of volumetric properties in order to determine the optimum


crumb rubber content that most improves the modified hot mix asphalt mixes. Importance of Study

In Malaysia, the application of crumb rubber as a modifying agent in hot mix asphalt is not significant enough. This is due to the less number of research being conducted in evaluating the potential of crumb rubber as an alternative material to improve the performance of asphalt mixes according to Malaysian condition. Hence, there is a need to conduct a detailed study on the performance of Malaysian hot mix asphalt using crumb rubber as a modifier. It is expected that this practice will not only have environmental significance, but it also have a potential to be cost effective and improve performance of modified flexible pavements as compared to conventional hot mix asphalt. Besides, if the enhanced characteristics of rubber modified asphalt pavement are significant, it could be a potential for crumb rubber to be used as a modifier in HMA mixes.

2. METHODOLOGY For the purpose of this study, Marshall mix design was used together with JKR specifications and ASTM 1992. The laboratory works were divided into several stages beginning with the aggregates preparation and distribution into different particle sizes through sieve analyses. The aggregates were dried sieve and blended meeting the gradation limit fulfilling the JKR specification. Washed-sieve analysis was referred to ASTM C 117 for determining the portion of filler content required in the aggregates gradation. The determinations of specific gravity for coarse and fine aggregates were done according to ASTM C 127 and ASTM C 128. Fine crumb rubber (grinding from truck tires) added was in a form of powder (0.3mm to 0.6mm). The amount of crumb rubber modifier added to the mixes was expressed in the


percentages (1, 2 and 3%) of the total weight of aggregate. 80/100 PEN asphalt cement was used for AC14 both conventional and modified mixes. While SMA14, modified asphalt cement PG76 was used for conventional mix and 80/100 PEN for modified mixes. The second stage was performing the Marshall sample for both SMA14 (50 blows) and AC14 (75 blows). Dry process was adopted in preparing the rubber modified asphalt mixes. Crumb rubber was added as part of the aggregate component before it was blended with the asphalt cement. Some modifications were made for procedures in preparing modified SMA14 mixes, where each sample was cured 1 hour at 160°C before compaction recommended by Arkansas State Highway and Transportation Department and no such curing for AC14. The bulk specific gravity and density of compacted sample were done in accordance to ASTM 2726. The stability and flow test were conducted for Marshall sample according to ASTM D 1559. An average value of theoretical maximum density was obtained as described in ASTM D 2041 using rice method for each different mixes. After obtaining optimum bitumen content, wheel tracking test was carried out with two samples for each mix design selected for measuring rutting potential. 3. RESULTS The specific gravity of different materials used in this investigation are given in Table 1.

Table. 1. Specific Gravity of Materials Used Material Specific Gravity Asphalt Cement 1.03 Coarse Aggregate 2.627 (MRP) 2.667 (Hanson) Fine Aggregate 2.601 (MRP) 2.551 (Hanson)


Crumb Rubber 0.405 Cement (OPC) 3.130 Based on the results obtained, relationship between volumetric properties (stability, stiffness, flow, VTM and VFB) and bitumen and crumb rubber content were evaluated. Then the optimum bitumen and crumb rubber content that most improve the HMA mixes were determined. The results of verified samples were recorded as shown in Table 1 and Table 2 for both mix designs at different content of crumb rubber added.

Table 2. Marshall mix design results of AC14 for conventional and

modified mixes

*Percentage of crumb rubber expressed from the total weight of aggregate blend

Dense graded (AC14) Volumetric Properties Conventi

onal Mix RUMAC 1

RUMAC 2

RUMAC 3

Specification

(JKR/SPJ/rev/2005)

Rubber Content

(%) 0 1* 2* 3* -

OBC (%) 5 (80/100)

5.1 (80/100)

5.3 (80/100)

5.6 (80/100)

-

Stability, S (kg)

1314 1159 1093 587 > 815

Flow, F (mm) 2.27 2.20 2.52 3.68 2.0 - 4.0

Stiffness (kg/mm) 881.9 526.

9 434.6 159.7 > 203

VTM (%) 3.4 3.8 4.3 4.6 3.0 - 5.0 VFB (%) 79.5 75.7 64.6 59.9 70 - 80


Table 3. Marshall mix design results of SMA14 for conventional and

modified mixes Gap graded (SMA14) Volumetric

Properties Conventional Mix

RUMAC 1

RUMAC 2

RUMAC 3

Specification (JKR/SPJ/rev/2005)

Rubber Content

(%) 0 1* 2* 3* -

OBC (%) 7.3 (PG 76) 6.7 (80/100)

6.8 (80/100)

7 (80/100) Min 6 (AASTHO)

Stability, S (kg) 996 812 762 510 > 632

Flow, F (mm) 2.24 2.52 2.60 2.72 2.0 - 4.0

Stiffness (kg/mm) 445.6 322.9 293.0 187.4 -

VTM (%) 3.4 4.2 4.4 4.9 4 ± 1 (NAPA) VMA (%) 19.2 19.9 22.3 24.7 Min 17

*Percentage of crumb rubber expressed from the total weight of aggregate blend

According to the results presented, OBC was determined at 4% air voids as referred to National Asphalt Pavement Association (NAPA). At this OBC, it was found that most of the volumetric properties met the required specifications except result of RUMAC3 for AC14 and SMA14. 4. ANALYSIS AND DISCUSSION Optimum Crumb Rubber Crumb rubber was added as an additive into the mixes and functioning as part of the aggregate. Basically, fine crumb rubber particles were used to generate a reaction between rubber and asphalt cement and at the same time some portion will serve to replace a portion of aggregates in the HMA mixes and act as elastic aggregate.


Modification of AC14 Results indicate that AC14 added with 1 and 2% crumb rubber had no significant effect on OBC compared to conventional mix. While for RUMAC3, the OBC increased from 5.0 to 5.6%. This expected behaviour could be attributed to the absorption of asphalt by crumb rubber which increases the asphalt content. However, stability and stiffness considerably decreased with increasing the crumb rubber percentages almost more than 50% when it reaches 3% crumb rubber. It can be seen that the effect of crumb rubber on other volumetric properties such flow and VTM, the values slightly increased as the percentages of crumb rubber increased but still within the specification range.

Modification of SMA14 Similar trends were observed in case of modifying SMA14. The addition of crumb rubber using dry process seems to reduce the stiffness of the modified mixes as indicated by a reduction in the stability. The decrease in stability with an increase in the percentage of crumb rubber may be an indication that 1 hour of curing does not permit adequate absorption reaction between asphalt and rubber to produce a modified blend. PG76 was used for conventional mix as recommended by JKR specification. While for RUMAC mixes 80/100 PEN was maintained in order to see the effectiveness of using crumb rubber with conventional binder. Result for OBC indicates that asphalt content required by modified mixes increase as the rubber percentages increased. Conventional mix shows higher OBC could be due to viscosity and different mixing temperature for modified binder. Flow result shows insignificant increment for RUMAC mixes compared to conventional mix. Based on both results, it can be summarised that the addition of crumb rubber tend to reduce the stiffness compared to conventional mix. This may due to the elastic behaviour of the crumb rubber added. It was discovered also that gap graded CRM mixes had lower Marshall stabilities than dense graded CRM


mixes. However, it was still within the acceptable ranges except for RUMAC3 for both mixes. RUMAC1 and RUMAC2 for both mixes were found to have better results among all the modified mixes. By having mixes with the present of elastic behaviour it could allow the HMA to recover from deformation under repeated loading. Thus, RUMAC1 and RUMAC2 for dense graded and all RUMAC mixes for gap graded were selected for further consideration to verify on the rutting resistance. Rutting Characteristics of Mixes The aimed of this study was to investigate the effect of adding crumb rubber to asphalt mixes. A major tool for this evaluation was performance related to rutting potential. Testing was performed according to BS 598 using Wessex Wheel Tracker Test to show the effect of increasing the amount of crumb rubber on the rutting performance. This test was conducted after the desired VTM of the sample achieve 7±1%. In order to get the specified percentage of air voids, the calculations were referred in ASTM D 3203-91. The rutting potential of the mixes was determined by measuring the accumulated permanent deformation at interval 25 load cycles until 5,000 load repetitions at temperature of 50°C. Table 4 and 5 summarise the data of rut depth obtained from Wessex Wheel Tracker Test for the conventional and selected modified mixes. The machine was set to stop after 5,000 load cycles or when the rut depth achieves 15mm. Dense graded (AC14) Basically, increasing rutting resistance is possible for modified mixes due to greater elasticity offered by the rubber particles. Excessive rutting was observed for dense graded compared to gap graded mixes. This may be due to the different of aggregates gradation between them. For dense graded, the addition of crumb rubber seems to improve the rutting resistance almost 30% for RUMAC1 and 44% for RUMAC2 respectively after 2,000 load


cycles. Figure 1 describes the similar trend of rutting for RUMAC1 and RUMAC2. Gap graded (SMA14) While result for gap graded, was found that the addition of crumb rubber in modified mixes with asphalt 80/100 PEN cannot compete with the conventional mix using PG76 as binder. PG76 improves rutting resistance more than 70% and 60% compared to modified mixes RUMAC2 and RUMAC3 respectively after 5,000 load cycles. However, RUMAC2 and RUMAC3 offer higher rutting resistance than RUMAC1. RUMAC1 seems to has highest potential of rutting and only can retain after 3,000 load cycles. Figure 2 shows the relationship between rut depth and number of cycles for both conventional and modified mixes and similar trend was observed for RUMAC2 and RUMAC3.

Table 4. Data from Wessex wheel tracker test (AC14) Rut Depth (mm)

Dense graded (AC14) Roller Passes

(cycles) Conventional Mix

RUMAC 1

RUMAC 2

0 0 0 0 500 3.1 3.4 3.1 1000 5.8 5.5 4.7 1500 9.7 7.7 6.4 2000 14.9 10.5 8.3 2500 - 11.4 10.7 3000 - 11.9 11.3 3500 - - -


Rut Depth vs Cycles

0

2

4

6

8

10

12

14

16

0 500 1000 1500 2000 2500 3000 3500 4000Cycles

Rut

Dep

th (m

m)

CONVENTIONAL

RUMAC1

RUMAC2

Figure 1. Results of wheel tracking test for AC14

Table 5. Data from Wessex wheel tracker test (SMA14) Rut Depth (mm)

Gap graded (SMA14) Roller Passes

(cycles) Conventional Mix

RUMAC 1

RUMAC 2

RUMAC 3

0 0 0 0 0 500 1 2.3 2.6 2.5 1000 1.5 3.6 3.8 3.6 1500 1.9 5 4.8 4.6 2000 2.2 7.1 5.9 5.6 2500 2.4 10.2 7.1 6.4 3000 2.6 14.5 8.5 7.3 3500 3.2 - 10.1 8.1 4000 3.3 - 11.6 8.8 4500 3.4 - 13.2 9.1 5000 3.5 - 13.5 9.8


Rut Depth vs Cycles

0

2

4

6

8

10

12

14

16

0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500Cycles

Rut

Dep

th (m

m)

CONVENTIONAL

RUMAC1

RUMAC2

RUMAC3

Figure 2. Results of wheel tracking test for SMA14

5. CONCLUSIONS AND RECOMENDATIONS

Based on this study, it was observed that the performance of HMA mixes was significantly affected with the addition of crumb rubber using dry process. OBC values were found to increase as the percentage of crumb rubber increased. Marshall properties obtained show that stability and stiffness of the modified mixes tend to reduce after the samples were added with crumb rubber. According to the results of the experimental investigations on the rutting performance conducted on conventional and modified mixes for both mixes (AC14 and SMA14) the following conclusions have been drawn.

i. For dense graded, crumb rubber modified mixes displayed lower potential for permanent deformation compared to conventional mix. RUMAC2 shows the least potential of rutting almost 44% improvement after 2,000 load cycles.

ii. Although gap graded mixes gave lower results of stability but they exposed the highest rutting resistance compared to dense graded mixes and retain much better under repeated load over 5,000 load cycles. However, the performance was still lower than the conventional mix using PG76 as binder.


The effect of rubber content was found to be the dominant factor in the permanent deformation resistance, where an increase in rubber content in the mixes increase the rutting resistance due to the highly elastic nature of the rubber particles. Generally crumb rubber appears worthy to be studied in relating with rutting resistance. Based on the overall findings, the addition of 2% crumb rubber was suggested in modifying the HMA mixes. It is recommended that further studies should be conducted on a variety of crumb rubber types, sizes and percentages. Instead of rutting, other pavement deterioration due to fatigue, aging, and stripping should be studied. Besides, other type of gradations also can be put into consideration such as open graded. In addition, the most important effort is to compile all the results obtained from all research that are related and revised by a responsible board in order to make worth of it in improving the HMA performance according to Malaysian condition. REFERENCES Elliott, R. P. (1993). Recycled Tire Rubber in Asphalt Mixes.

Project Proposal Submitted to the Arkansas State Highway and Transportation Department.

Epps, J. A. (1994). Uses of Recycled Rubber Tires in Highways – A Synthesis of Highway Practice. NCHRP Report No. 198.

Esch, D. C. (1984). Asphalt Pavements Modified with Coarse Rubber Particle. Alaska P, Report No.

FHWA-AK-RD-85-07. Federal Highway Administration, Washington, D.C.

Heitzman, M. (1992). State of the Practice – Design and Construction of Asphalt Paving Materials with Crumb Rubber Modifier. Research Report No. FHWA-SA-92-022. Federal Highway Administration, Washington, D.C.

Koh, M. I. and Talib, N. (2006). Use of Scrapped Tires as a Substitute for Fine Aggregates in Asphaltic Pavement. National Seminar on Civil Engineering Research.


Rebala, S. and Estakhri, C. K. (1995). Laboratory Evaluation of CRM Mixtures Designed Using TxDOT Mixture Design Method. Transportation Research Record No. 1515.

Stroup Gardiner, M. and Krutz, N. (1992). Pavement Deformation Characteristics of Recycled Tire Rubber Modified Asphalt Concrete Mixes. Transportation Research Record No. 1339.

Takkalou, H. B., Hicks, R. G. and ESCH, D. C. (1985). Effect of Mix Ingredient on the Behavior of Rubber-Modified Mixes. Report No. FHWA-AK-RD-86-05A. Federal Highway Administration, Washington, D.C.

Troy, K., Sebaaly, P. and Epps, J. (1996). Evaluation Systems for Crumb Rubber Modified Binders and Mixtures,” Transportation Research Record, No. 1530, TRB, National Research Council, Washington, D.C.


8

THE EFFECT OF RAINFALL ON ASPHALT SURFACING MATERIALS

Nursetiawan, David Woodward, Alan Strong

Universiti of Ulster, UK

1. INTRODUCTION This paper reports findings of an investigation to determine the effect of rainfall on the properties of asphalt surfacing materials. The effect of rainfall is important for a number of reasons. For example, it can significantly reduce skid resistance. Excessive spray generation can make driving dangerous due to reduced visibility. Water has also been shown to cause reduction of aggregate / bitumen adhesion due to moisture sensitivity related issues.

This paper considers the development of surface runoff during simulated rainfall. In particular the relationship between rainfall intensity, crossfall and texture depth with Time of Transition Flow i.e. the duration into a rainfall event at which the surface texture has filled with water and surface runoff starts to happen.

2. MATERIALS USED IN THE INVESTIGATION Five types of test surface were assessed in the investigation. The asphalt materials were selected as being representative of the wide


range of road surface materials currently used in the UK. These consisted of the following:

• Proprietary 6mm Open textured asphalt surface (6mm OT).

• 10mm Dense Bituminous Macadam (DBM) wearing course to BS 4987.

• Proprietary 10mm Marshall Asphalt (10mm MA)

• Proprietary 14mm Marshall Asphalt (14mm MA).

• Plywood to give a very low value of texture depth (assumed to be close to zero as measured using the sand-patch test method).

3. RESEARCH METHODOLOGY The research developed a laboratory based methodology to investigate the interaction of rainfall with a range of road surfaces. This resulted in development of the Ulster Rainfall Simulator (URS).

Development and subsequent use of the URS involved the following stages i.e. preparation large test specimen slabs, mounting the large test specimen slabs at three crossfalls, simulation of three rainfall intensities under controlled laboratory conditions allowing the inter-relationships between rainfall intensity, rainfall duration, texture depth and crossfall to be determined.

The test specimens were 1400 x 600 x 50mm in size. The different types of hot mix asphalt were sampled at a number of mixing plants in the UK. Sufficient 25kg bags of asphalt were reheated in the laboratory and placed in a specially designed compaction mould.

The compaction mould used a concrete shuttering panel to provide the strong base with the sides made with 50mm thick timber. A


vibratory pedestrian roller was used to compact the hot asphalt. Reheating and compaction temperatures were recorded.

The finished surface texture of each asphalt test specimen was determined at 8 locations using the Sand Patch method in accordance with BS 598:Part 105 (2001).

The Ulster Rainfall Simulator (URS) is shown in Figure 1. It consists of the following main components i.e. mains water supply and storage tank, water pump to pump water to the spray nozzles, a flume on which the asphalt slab is located, a flow meter and water pressure gauge, a spray nozzle and water capture tank.

The spray nozzle was located 1.43m above the test surface. Three different capacity nozzles were used. Initial tests were carried out to standard the appropriate water pressure and nozzle type to give three rainfall intensities.

The edge of the asphalt slab in contact with the edge of flume was sealed using mastic sealant to avoid water leakage. The slope of the flume arrangement was adjusted using a hydraulic jack.

Figure 1. The Ulster rainfall simulator

tank

nozzle

Storage tank

pump

Water Supply

Pressure gauge

Bituminous material

Flow Meter


The uniformity of rainfall intensity distribution across the large asphalt test slab was determined by placing 13 catch cans evenly across its surface. The simulated rainfall for each of the three spray nozzles was collected over a 10 minute duration for each catch can.

This data was analysed to determine the uniformity of each spray nozzle distribution using Christiansen Uniformity Coefficient (CU) as detailed by Zoldske and Solomon (1988):

A summary of the nozzle performance data is given in Table 1. This shows the pressure and flow rate values used for each of the spray nozzles to give the increasing rainfall intensities used during the experiments.

Table 1. Summary of nozzle performance data

Nozzle Number

Pressure(bar)

Flow rate (l/s)

Average Rainfall Intensity(mm/h)

CU (%)

1 0.75 2.00 31.4 86.6 2 1.20 4.20 54.2 91.2 3 1.60 5.75 78.3 85.7

4. TESTING Table 2 shows the four groupings of variable assessed i.e. those belonging to the test surface, the test condition, variables measured during testing and those calculated from the test data.


Table 2. Summary of test variables

Variable Group

Variable

Test surface

Surface type Texture depth Air voids Hydraulic conductivity

Test condition

Crossfall Rainfall intensity Duration of testing

Measured during testing

Runoff Percolation

Calculated from test data

Time for transition flow (Tft) Time for steady state flow (Tsf) Total flow rate

A typical test consisted of setting up the slab at the required crossfall angle. The nozzle type and pump conditions selected to ensure the required rainfall intensity.

The test was then started and the amount of water captured at the down slope end of the flume every 10 seconds. Testing typically lasted for 10 minutes during which period runoff reached equilibrium.

Some of the asphalt materials were designed to be porous and so surface runoff and percolation down through the material was also recorded.


0

5

10

15

20

0 120 240 360 480 600 720 840

Time (s)

Flow

Rat

e (c

ubic

cm

/s)

2% crossfall 4% crossfall

6% crossfall

Figure 2. Example of flow rate plotted against time (14mm Marshall Asphalt, Rainfall Intensity 78.3mm/h and crossfalls of 2, 4 and 6%)

Figure 2 shows an example of flow rate against time for 14mm Marshall Asphalt at a rainfall intensity of 78.3mm/h and crossfalls of 2, 4 and 6%. This example shows that at the greatest rainfall intensity crossfall had a minimal effect in the time it took for the surface texture to reach a condition of steady flow.

This time period has been termed the Time of Transition Flow (Ttf) and relates to the time it takes for the surface voids and interconnected voids within a mix to become infilled by water. After this period surface runoff will start.


0

5

10

15

20

0 120 240 360 480 600 720 840

Time (s)

Flow

Rat

e (c

ubic

cm

/s)

2% crossfall 4% crossfall

6% crossfall

Figure 3. Example of flow rate plotted against time (14mm Marshall Asphalt, Rainfall Intensity 31.4mm/h and crossfalls of 2, 4 and 6%)

Figure 3 shows an example of flow rate against time for the 14mm Marshall Asphalt at a rainfall intensity of 31.4mm/h and crossfalls of 2, 4 and 6%. This example shows at the reduced rainfall intensity crossfall has a significant effect on Time of Transition Flow(Ttf).

0

60

120

180

240

0 1 2 3 4 5 6 7 8

Crossfall (%)

Ttf (

s)

Wood 10mm DBM10mm Superflex 14mm Superflex

Fig. 4. Effect of crossfall on Time of Transition (Rainfall Intensity 31.4mm/h)


Figure 4 plots the effect of crossfall on Tft for the 5 surfaces assessed at rainfall intensity 1 (31.4mm/h). The plots have strong linear relationship and show that Tft reduces as crossfall increases.

0

60

120

180

240

0.0 0.5 1.0 1.5

Texture depth (mm)

Tft (

s)

Crosfall 2% Crossfall 4% Crossfall 6%

Figure 5. Effect of texture depth on time of transition flow (Rainfall Intensity 31.4mm/h)

Figure 5 plots the effect of texture depth measured using the sand patch method on Tft. Again this shows strong linear relationships i.e. as the surface texture of the road increases it takes longer for Tft to be reached.

As texture depth increases the road surface is able to act as a reservoir until such time as it reachs capacity and there is excess water to cause runoff.


0

60

120

180

240

0 0.5 1 1.5

Texture depth (mm)

Tft (

s)

RI 31mm/h RI 54mm/h RI 78mm/h

Figure6. Effect of texture depth on time concentration for slope 2%

Figure 6 shows the effect of rainfall intensity on the relationship between Tft and texture depth. The general relationship is similar to that shown in Figure 5.

5. ANALYSIS OF DATA Multivariate regression analysis was carried out to model the data. The dependant variable was Tft with the independent variables being rainfall intensity, crossfall and texture depth. This resulted in the general equation (1):

Ttt = 189.39 – 1.53 x RI -10.63 x S +58.22 x TD (1)

Where:

RI = rainfall intensity (mm/h)

S = crossfall (%)

TD = Texture depth (mm)


Figure 7 plots calculated and predicted Tft values using this model and shows strong linear correlation.

y = 0.8445x + 17.245R2 = 0.8445

0

60

120

180

240

0 60 120 180 240

Predicted Ttf (s)

Cal

cula

ted

Ttf (

s)

Figure 7. Plot of predicted and calculated Tft data

6. CONCLUSIONS The investigation found predictable correlations between the main variables. The ranking of variables in order of importance was found to be rainfall intensity, slope and texture depth.

Multivariate regression analysis was used to model the variables and derived a linear equation model relating rainfall intensity, crossfall and texture depth with Tft.

The research shows that it possible to derive fundamental understanding of surface water runoff using simulatory laboratory investigation without the need for expensive full-scale road trials.


ACKNOWLEDGEMENTS The authors would like to acknowledge the support provided by Aggregate Industries UK.

REFERENCES

CIRIA Publication C523 (2001). Sustainable urban drainage systems – best practice manual, UK.

CIRIA Publication C609 (2004). Sustainable Drainage Systems – Hydraulic, Structural and Water quality advice, UK.

Domenichini, L. and Lopencipe, G. (2004). Validation of DESTTRA Water Film Depth Prediction Model. The SIIV 2nd International Congress, Italy.

Gallaway et al. (1971). The Effects of Rainfall Intensity, Pavement Cross Slope, Surface Texture, and Drainage Length on Pavement Water Depths. Research Report Number 138-5, Texas Transportation Institute, TAMU College Station, Texas, USA.

Highways Agency (1999). Design Manual for Road and Bridges, Volume 7: Pavement Design and Maintenance, HSMO, London.

NCHRP (1998). Improved Surface Drainage of Pavements. Final Report, The Pennsylvania Transportation Institute, Transportation Research Board, National Research Council, USA.

Ross, N.F. and Russam, K. (1968). The Depth of Rain Water on Road Surfaces. Road Research Laboratory, Report LR 236, UK 1968.

Simone, A., Vignali, V., Bragali, V. and Maglionico, M. (2004) Surface Runoff: A Rainfall Simulator for Wash-Off Modelling and Road Safety Auditing Under Different Rainfall Intensities. The SIIV 2nd International Congress, Italy.


Zoldske, D.F. and Solomon, K.H. (1988). Coefficient of Uniformity – What It Tells Us? Irrigation Notes, California State University, USA.


9

RESEARCH INTO SUSTAINABLE ASPHALT SURFACING MIXES IN THE

UK

David Woodward, Alan Woodside, Alan Strong Universiti of Ulster, UK

Paul Phillips and Bob Allen

Aggregate Industries

1. INTRODUCTION In the United Kingdom (UK) sustainability has become the most important issue in highway engineering. In the last 10 years there has been tremendous change in terms of government policy, specifications, design to product development and the laying of asphalt materials.

The many definitions of sustainability from Brundtland (WCED, 1987) to the English Highways Agency (2003) address issues such as global warming, greenhouse gases, carbon emissions and the carbon footprint.

Sustainability is about recycling, reuse and minimizing use and has led to many types of waste and by-product being evaluated. However, only a small number have been found application in surfacing materials.

Sustainability is also about how improved stiffness and fatigue characteristics of the load bearing layers has allowed reduced layer thicknesses, less aggregate and bitumen use. Alternatively, these


layers can be designed thicker to increase pavement life and so reduce thickness of the surfacing layer.

Sustainable highway construction is about appreciation of the bigger picture e.g. the inevitable future shortage of bitumen. Sustainability is about understanding the interconnections between practises, expectations, policies, technologies, whole life costing to better understanding the limitations of the materials used.

In summary, sustainability in practical terms it is about making better use of our resources to minimize the risk in providing longer-term performance of what may be conflicting in-service property expectations.

This paper draws examples from research at University of Ulster in areas from aggregate to mix properties such as skid resistance, noise and rolling resistance to highlight how better understanding of material properties is necessary to meet the simple ideal of providing more sustainable highway surfacings.

2. UK HIGHWAY SURFACING EXPECTATIONS In the UK, the requirement for a skid-resistant surfacing in wet conditions underpins all other expectations. People complaining now place road noise almost as important as safety. However, more sustainable roads is not simply about providing higher levels of grip or making them quieter.

It requires understanding of the combination of road surface characteristics, contact patch phenomena, tyre technology, braking, suspension systems and vehicle dynamics.

A sustainable approach would be to construct the road surface to last as long as possible and rely on another means such as greater tyre tread depth to remove water or use of ABS braking and traction control.


3. ENVIRONMENTAL FACTORS Highway design tends to concentrate on number of the commercial vehicles. If the consequences of global warming are considered, the UK climate is expected to experience greater variation in temperature and rainfall variation.

Little consideration is given to simple things such as loss of surfacing aggregate strength due to rainfall. For example, some types of surfacing aggregate can loose up to 70% of their dry strength each time is rains (Woodward, 1995). In the UK the main types of surfacing materials are high stone content mixes with variable interconnected void contents i.e. SMA, propriety thin surfacing and semi porous asphalts.

These materials have negative and porous textures and tend to trap surface water either beneath the tyre or force it into the mix via its network of interconnected voids. As water is not compressible the resulting very high pressures may be affecting the longer-term cohesion of different mixes in different ways.

For asphalt mixes designed to be porous the energy associated with this water can probably be dissipated. However, for lower void content mixes the water may be forced deep within the mix and may start to affect properties such as adhesion and cohesion.

Current research by Nursetiawan (2008) is asking simple questions such as how long does it takes an asphalt mix to dry out between simulated rainfall events. The research aims to show that by taking longer to dry out then mix longevity may be detrimentally affected.

4. UNIQUE TEST FACILITIES Despite everyone on this earth relying on some version of a road infrastructure there are only a relatively small number of highway research facilities and an even smaller number of academic centers.


Research into highway materials is a very small community compared to areas such as medicine or other aspects of engineering.

Research at the University of Ulster has concentrated on surfacing materials and has a number of unique pieces of test equipment that have facilitated PhD studies into areas such skid resistance, noise generation and rolling resistance..

For example the ULTRA is an internal drum machine that can be can be used to assess how these three properties relate to speed. The Road Test Machine (RTM) simulates the wear caused by two full-scale tyres on surfacing properties such as wear, skid resistance, texture depth and mix integrity.

It has been used in projects ranging from comparison of laboratory and in-service early life properties of asphalt mixes, assessing the effect of diesel spillages on mix cohesion to the colour retention of coloured surfacings. The equipment is also used to test high friction surfacings for wear testing to Appendix H of TRL Report 176 (Nicholls, 1997).

5. AGGREGATE In the UK BS EN 13043 (2002) specifies aggregate properties with guidance given in PD 6682-2 (2003). The main aggregate properties are grading, fines content of both the coarse and fine aggregate, flakiness index, resistance to fragmentation, PSV, AAV and durability measured with water absorption and magnesium sulphate soundness if water absorption is >2%.

The skid and abrasion resistance properties of aggregate for surface courses are specified in HD36/06 of the Design Manual for Roads and Bridges (DMRB, 2006). While the bitumen industry continues to develop new types of super-binder the same types of aggregate are expected to perform to ever increasing levels of stress made possible by the improved binders.


Aggregate properties have a natural limit beyond which they will fail. Improved sustainability implies understanding when and how this limit is reached and the consequences of failure. Woodward (1995) compared the ability of aggregate test methods to predict performance and highlighted many issues that contrasted the limitations of laboratory testing with in-service performance.

The research found that PSV is gained at the expense of almost every other property such as strength and durability. In sustainability terms surfacing aggregate is exposed to rain and oxygen the same as bitumen and leads to an important junction between what a national guidance document may recommend and the extra that is required to minimize risk and enhance longer-term performance i.e. modified and non-standard testing.

It requires that the contribution of the aggregate to mix performance be better understood e.g. use of the German Wehner Schulze test for testing the skid resistance of the asphalt mix.

6. THE SURFACING / TYRE INTERFACE The dynamics of a vehicle are transmitted through the pavement structure via the tyre / road surface interface. It is subjected to direct contact with the tyre and all of the imposed stressing associated with the moving vehicle e.g. acceleration, braking, cornering, steady speed.

The texture and grip characteristics of the road surface reflect those of the tyre. Both road surface and tyre require minimum levels of both to remove water films in wet weather and minimize aquaplaning.

Liu (1993) considered the road / tyre interface and its application to road design criteria. He found that the contact envelope between tyre and surface is elliptical in shape and directly proportional to wheel load and inversely proportional to tyre pressure.


Whereas wheel load is important to pavement design, tyre inflation pressure is more important for the conditions experienced at the road / tyre interface. The tread pattern of a tyre is typically only 14 to 18% of the contact envelope. The actual contact area may be much lower than the computed contact area leading to underestimation of the actual contact stress at the trafficked aggregate surface.

Siegfried (1999) and Douglas (2000, 2007) found that the stresses involved due to a tyre rolling over a road surface are highly concentrated and will exploit any weakness present such as micro-texture on the aggregate surface, the chipping edges or inferior quality constituents.

This helps to explain how factors such as grip, noise and rolling resistance are closely inter-related to not only one another but also to the amount and type of texture.

7. MOISTURE SENSITIVITY Premature failure of the aggregate / bitumen bond will lead to stripping related problems and / or ravelling and break-up of the surfacing layer. One the main reasons for this type of failure is moisture.

McKibben (1987) developed a modified version of immersion wheel track testing where the tyre was locked in one direction causing it to be dragged across the tracked surface of the test specimen. The test samples were soaked in water prior to testing and the dragged tracking action exploited any weakness in either the aggregate / bitumen bond or mix integrity.

The Strategic Highway Research Program (SHRP) developed the Net Adsorption Test as a relatively fast and simple test method to quantify the adsorptive nature and water sensitivity of bitumen-aggregate combinations. The American research found that aggregate played a more important role than previously acknowledged.


Woodward (1995) and Russell (1997)_ considered the relationships between factors such as aggregate skid-resistance and soundness to the information obtained by the NET test. Aggregates were assessed to represent a wide range of both bulk chemistry composition and engineering properties such as PSV and MSSV.

The amount of initial bitumen adsorption ranged from 48.0% for quartz vein to 97.3% for unsound lateritic basalt. This would be expected as the quartz vein had been chosen for its poor adhesion characteristics whereas the lateritic basalt was a major problem with regard to making dry mixes. The introduction of water caused the adsorption values to drop to 44.6% to 80.8% for the same aggregates respectively.

Plotting these values resulted in overlapping fields based on rock type in the general order limestone, granite, gritstone and basalt. In terms of adhesion to bitumen higher PSV aggregates for any group either require more bitumen for optimal adhesion and / or may be potentially more susceptible to the detrimental effects of moisture (Woodside et. al. 1994).

A simple method to rapidly assess moisture sensitivity has recently been evaluated. The method is called the Mini Moisture Effect test and consists of placing single laboratory prepared plugs in a simple kitchen pressure cooker along with a predetermined amount of water.

When the pressure release value went off about 9 minutes later the plug was removed and allowed to cool. The effect of moisture was assessed by before and after ITSM. This simple test allows the effect of greater number of conditioning cycles to be quickly assessed.

8. TIME SCALES Time forms the basis of defining sustainability. Time scales of varying length are fundamental to material performance. Geological time ranges from initial formation to exposure at the quarry face or


extraction as crude oil and is expressed in 10’s or 100’s of millions of years.

The conditions experienced during geological time will govern whether the aggregate is mechanically strong and abrasive, its level of skid-resistance or whether a specific quarry or part of a quarry may be prone to producing a mechanically weak, unsound or flaky chipping. In pavement design, the structure is designed to last 10’s of years.

Engineering time is similar to design time and is in the order of years. Many engineers still consider that construction materials such as aggregate and bitumen are inert and do not appreciate that their properties are constantly changing over time.

Woodward (1995) proposed the concept of performance time. An example of short-term performance time would be the use of a very high skid resistant aggregate that is susceptible to the wearing action of traffic. Whilst, this may provide initial high levels of skid-resistance, trafficking quickly decreases texture depth resulting in premature loss of skid-resistance.

Longer-term performance time would be the use of a harder but lower PSV aggregate were the level of skid-resistance is at a lower level but maintained for a much longer period of time.

9. NOISE, ROLLING RESISTANCE AND TEXTURE In the UK noise is the second most important property after skid resistance. Anderson (2000) considered the generation of road / tyre noise in relation to porous asphalt mixes and concluded that aggregate size and shape was important together with mix tortusoity i.e. how the voids were interconnected.

Yacoob (2006) considered the use of smaller sized types of surface dressing as a sustainable option to re-surfacing high stone mixes or concrete surfaces. The research concentrated on assessing the role of texture in the generation of noise.


A methodology was developed to take road surface latex castes and produce specimens for testing using a rotating drum. Test conditions of speed, load and tyre type were varied and noise measurements were used to develop a noise index (Woodward et. al. 2005).

Whereas Anderson (2000) had found that a smooth road surface and negative texture with interconnected voids was ideal for a hot mix, Yaccob (2006) found that reducing the texture depth of positive textured surface dressing resulted in significant reductions in road/ tyre noise.

Using a simple coast-down technique McErlean (2006) found that surface texture was the main surfacing property that had to be considered when developing more fuel-efficient surfacing materials.

Comparison of noise and rolling resistance characteristics showed a clear relationship between the two factors concluding that significant sustainability improvements could be gained by moving toward smaller nominal size materials and textures.

10. SKID RESISTANCE / GRIP This area has been investigated for many years at the University of Ulster. Woodside (1981) considered the limitations and stressing during the PSV test.

Many subsequent studies have evaluated the relationship between aggregate petrology and PSV (Woodward, 1995) and concluded that almost all other aggregate properties are gained at the expense of PSV.

Perry (1997) considered the PSV of Northern Ireland greywacke whilst Jellie (2003) evaluated the diverse range of factors involved in the provision of skid resistance.

The SKIDPREDICT project with TRL revaluated the PSV test to determine whether an ultimate state of polish for an aggregate


existed (Roe and Woodward 2004). New variations of the PSV test were further developed and used in SKIDGRIP (Woodward, 2003). This found that by offsetting the test wheel to induce greater stress further reductions of 20+ points less than the standard PSV value were possible.

This research identifies that the PSV test should only be regarded as a ranking test and not as a method to predict in-service performance.

The SKIDGRIP project used a GripTester to investigate the development of early life skid resistance. Collaboration with industrial partners facilitated development of new asphalt materials and monitoring of their in-service development with time (Jellie et. al. 2004; Woodward et. al. 2007).

11. CONCLUSIONS This paper has considered a range of different research areas investigated at University of Ulster that underlie the prediction of what may be termed sustainable surfacing performance.

They help to illustrate the complexity of the factors involved. Reliance on standard test methods will not provide the necessary in-depth understanding.

The demand for higher performance in an increasingly diverse range of applications and environments has severely stretched what natural materials are capable of achieving.

The finished high quality product may in reality be more susceptible to a greater range of mechanisms that lead to poor sustainable performance.

Surfacing aggregate is exposed to many types of stressing which are not adequately accounted for in current specifications or test methodologies.

The stressing experienced in-service will vary considerably depending upon the use it is put. It may be possible to enhance the


use of a ‘lower quality material’ (in terms of PSV) by using it in ways different to what is done traditionally i.e. reducing the nominal aggregate size to increase the contact area.

Initial adhesion and subsequent moisture sensitivity of asphalt surfacing mixes are areas that still need much work as these are now recognized as being high risk issues.

This is particularly so with the introduction of performance specifications and long term warranty / maintenance periods that are / have replaced existing recipe methods allowing the designer / contractor / material supplier much greater flexibility.

Greater knowledge of the materials being used and the interconnection of properties has become an essential pre-requisite to sustainable highway construction practise.

ACKNOWLEDGEMENTS The authors would like to acknowledge the assistance of Aggregate Industries in many of the research projects summarized in this paper.

REFERENCES

Anderson, G. (2000) An Investigation of the Factors which affect the

Acoustical Characteristics of Bituminous Porous Road Surfacing PhD thesis, University of Ulster.

BS EN 13043. 2002. Aggregates for bituminous mixtures and surface treatments for roads, airfields and other trafficked areas. BSI, London.

Design Manual for Roads and Bridges. (2004). Volume 7 Pavement Design and Maintenance, Section 3 Pavement Maintenance Assessment, Part 1 Skidding Resistance HD 28/04.


Douglas, R. A. (2007) Tyre/road contact stress distributions measured and modelled in three coordinate directions Current project funded by Land Transport New Zealand.

Douglas, R. A., Woodward, W.D.H. and Woodside, A.R. (2000) Road contact stresses and forces under tires with low inflation pressure. Canadian Journal of Civil Engineering. Vol 27 pp 1248-1258 ISSN 1208-6029.

Liu, G. X. (1993) The area and stresses of contact between tyres and road surface and their effects on the road surface. PhD thesis, University of Ulster.

Highways Agency. (2003) Building Better Roads: Towards Sustainable Construction. London.

Jellie, J.H., Woodward, W.D.H. and A.R. Woodside (2004) The early life safety of high stone content surfacings. European Eurobitume Conference, Vienna, May, 2004.

Jellie. J. (2003) A study of factors affecting skid resistance characteristics. PhD thesis, University of Ulster.

McErlean P. (2006) A study of the relationship between texture depth, rolling resistance and noise for highway surfacing materials. PhD thesis, University of Ulster.

McKibben M. (1987) A study of the factors affecting the performance of dense bitumen macadam wearing courses in Northern Ireland. PhD thesis, University Polytechnic.

Nicholls J.C. (1997) Laboratory tests on high-friction surfaces for highways TRL Report 176.

Nicholls, J.C. (2005). Design guide for road surface dressing Road Note 39 (Fifth edition), TRL Limited.

Nursetiawan (2008) Hydraulic properties of road surface materials. Current PhD study, University of Ulster.


Perry. M. (1997) A study of the factors that influence the polishing characteristics of gritstone aggregate. PhD thesis, University of Ulster.

PD 6682-2 2003. Aggregates for bituminous mixtures and surface treatments for roads, airfields and other trafficked areas – Guidance on the use of BS EN 13043.

Roe PG and WDH Woodward (2004) Predicting skid resistance from the polishing properties of the aggregate SKIDPREDICT Final Report PR CSN/31/03.

RUSSELL, T.E.I (1997) The effect of aggregate properties on the aggregate/bitumen adhesive bond. PhD thesis, University of Ulster.

Siegfried. (1999) The study of contact characteristics between tyre and road surface. PhD thesis, University of Ulster.

Woodside A.R. (1981) A study of the characteristics of road stones particular reference to polishing and skidding resistance. MPhil thesis, Ulster Polytechnic.

WOODSIDE, A.R., WOODWARD, W.D.H, RUSSELL, T.E,I. and PEDEN, R.A. (1994). The relationship between aggregate mineralogy and the adhesion of bitumen to aggregate. First Symposium on Performance and durability of Bituminous Materials, Department of Civil Engineering, University of Leeds, 30th to 31st March.

Woodward, W.D.H. (1995) Predicting the performance of surfacing aggregate. DPhil Thesis, Faculty of Engineering, University of Ulster.

Woodward, W.D.H. (2003) Predicting Early Life Skid Resistance of Highway Surfacings (SKIDGRIP), EPSRC funded project, IGR Final Report GR/R09022/01.

Woodward, W.D.H., Woodside, A.R., Phillips, P., Shahmohammadi, R. and I. Walsh. (2007). The development of very early life skid resistance. 4th International Conference


Bituminous Mixtures and Pavements, Thessaloniki, Greece, 19-20 April

Woodward, W.D.H., Woodside, A.R., Yaacob, H. & McErlean, P. (2005). Development of the USI laboratory test to predict tyre / road noise. International Journal of Pavements, Vol. 4, No 1-2, January-May. pp 72-81. ISSN 1676-2797.

World Commission on Environment and Development. (1987). Our common future. Oxford: Oxford University.

Yaacob H. (2006). A study of the effect of texture on surface dressing characteristics. PhD thesis, University of Ulster.


10

EXTENT OF PAVEMENT DISINTEGRATION ASSOCIATED WITH

LOW COST ROAD FAILURE

J. Ben-Edigbe Universiti Teknologi Malaysia, Skudai, Johor

1. INTRODUCTION Road pavements do not fail suddenly. It is generally considered to begin to deteriorate after entering service and then gradually get worse as time progresses until a failure condition is reached. Bituminous surfacing may crack for a variety of reasons that include lack of good bond between the surface layer and the course underneath, excessive pavement deflection, expansion, shrinkage and contraction of the sub grade. In early stages the crack patterns can indicate the cause. However, when cracks have developed over a large area and become sufficiently wide and numerous to allow the entry of surface water or disturbances into the pavement, road begins deteriorate. Pavement distress is prevalent in Nigeria especially on state roads. The road system in Nigeria is classified into four main categories: i. The Federal Express Roads, maintained by the Federal Government; ii The Federal highways, formerly under state ownership, but were taken over by the Federal Government, with a view to upgrading them to motorways; iii. The State low cost roads, maintained by the state governments; and iv. Local government highways; each tier of government has the


responsibility for planning, construction and maintenance of the network of roads under its jurisdiction. The current total road network of roads, estimated at 193,000 kilometres as shown below in Table 1. Nigeria’s population has grown from 55 million in 1960 to an estimated 150 million in 2007 making it one of the most densely populated countries and dependence on road transportation is near total. Vehicle population is up from 55,000 in 1970 to 2.5 million in 2007, passenger car traffic grew at an annual rate of 3.5 per cent between 1998 and 2002, and is estimated to grow at 4.5 per cent between 2002 and 2007 according to Federal Ministry of Works, Sheladia Inc and Yolas Network report of 1998. Commercial vehicle traffic growth rate is estimated at 3.5 per cent per annum for the period 1998-2003. These large increases in traffic flows have road management, road maintenance, and also, environmental management consequences. Further, the transition to higher level of vehicle ownership and consequently road use has become an increasing burden to governments in Nigeria. However, state roads are usually designed and constructed on ‘low-cost’ concept. According to Robert Petts (2002), ‘Low Cost Surfacing is concerned with supporting sustainable improvements in low cost road surfacing and basic access to support poverty reduction initiatives in rural communities. This implies the effective use of local resources, particularly human resources, locally available and alternative materials, and readily available and low cost intermediate equipment wherever possible. In the situation of scarce financial resources, it also requires the application of affordable and appropriate standards and adoption of techniques suitable for use by the indigenous private sector (particularly small domestic construction enterprises) and local communities. The application of good management practices coupled with adequate technical inputs are also encouraged.’ In other words, those embracing low cost road concepts are at liberty to bend design standards if need be it. Cook J.R and Gourley C.S (2002) suggested that ‘The Key to the success of these innovative solutions is recognition that conventional assumptions regarding


road design criteria need to be challenged and that the concept of an appropriate, or environmentally optimised design, approach provides a way forward. Low volume road standards and designs need to support the function that the road is providing as well as recognising the important influences of the deterioration mechanisms The use of locally available, but frequently non-standard, pavement construction materials plays a significant role within this concept’.

Table 1. Nigeria road network 1996

Road Length - Km

Road Class Paved Unpaved Total %

Federal

State

Local

26,500

10,400

2,600

5,600

20,100

128,000

32,100

30.500

130,600

16

16

68

Total 36,500 153,700 193,200 100

% 20 80 100

But why are most state low cost roads life cycle short lived it may be queried. The answer lies with a road system that’s poorly thought out, where the drive to construct roads is carelessly pursued with little considerations for the long time financial implications. Otherwise how can the results of nationwide road condition surveys carried out by the governments in 2002 that clearly show that 70 percent of the roads in Nigeria are in poor conditions be explained. The state of some road sections at the time of survey suggests that road pavement distress is characterised by substantial potholes, edge damage, multiple cracks and wheel ruts. In fact, many factors can be called to account for the occurrence of such vase pavement distress and they may include among others, poor design, poor construction, and


poor maintenance. The objective of this paper is to examine the extent of pavement distress associated with low cost roads in Nigeria. For that purpose, the remainder of the paper has been divided into; Section 2, literature; Section 3, data collection; Section 4, findings; and Section 5, Conclusions. The deterioration of paved roads is defined by the damage trend of its surface condition over time. The defects of a pavement surface, which is usually quantified through a pavement condition survey, are classified under three major models of distress, namely; cracking, disintegration, and permanent deformation. The main focus of this paper is on the crack damages because cracking often triggers the application of maintenance treatments and cracking can be the decisive factor in determining the most appropriate rehabilitation option among others. 2. LITERATURE Since Nigeria Road designs are based on British Standards, the notion that low cost road concept is most appropriate for developing countries were not challenged by engineers and researchers in Nigeria. More so, because the research programs were usually founded by the United Kingdom since Nigeria has no credible research base from which to launch any constructive counter concept. In any case the concept was embraced holistically by all state governments as cost saving and politically correct. It made politicians look good and they adore it. Consequently, 13,000km of such so-called roads were and are still being constructed. The state of low cost roads has remained poor for number of reasons. The number one problem is poor quality, resulting from faulty designs, lack of gutters and very thin coatings that are easily washed away by floods and can hardly withstand heavy traffic loadings. Second, funding of road maintenance has been grossly inadequate and poorly managed. Thirdly, excessive and overburden use of the road network, given the near dependence on road transport of goods and people. Fourth, there is


no articulated road maintenance management program at any level of government; most works are usually carried out intuitively. Road maintenance decisions are influenced by politics and not necessarily on the actual maintenance needs, consequently most roads have been neglected and laid to waste. State and Local governments maintained paved roads (13,000km) are usually low cost design. In highway surveys conducted by the governments from the 11th to 13th December 2002 in Nigeria results indicate that most roads are in poor conditions and dire need of rehabilitation. On many roads, the shoulder, a major component of the road had eroded off, putting the roads in near impassable condition. The survey also revealed that from February 1997 to December 2001 (no data for 2002), a total of 96 road contracts, mainly rehabilitation, reconstruction and expansion, were awarded by the Federal Ministry of Works, at a total contract sum of N186.999 billion ($1.5billion). According to Nigeria Federal Ministry of Works (1998) over 70 per cent of the national road surfaces that are in poor condition are located in the southern region of the country, so the study on road pavement distress surveys was conducted in southern. Asphaltic highway pavement (flexible road), which is of interest to us, consists of a series of structural layers (sub base, road base and surfacing) on the naturally occurring soil generally referred to as the sub-grade or if you like the formation level. The sub-grade is the structure that must eventually support all of the loads that come onto the pavement. Clearly, the strength of the sub-grade has to be established and cannot be assumed. The sub-base is a layer of relatively weak material and its thickness depends upon the projected intensity of traffic loadings and the strength of the sub-grade. The road-base is usually asphalt, and designed to absorb and redistribute loads such that deformation in the road pavement remains within acceptable limits. In a true flexible road the whole of the construction is in compression, the loads from traffic being so distributed by the construction that the load at formation level is not greater than the sub-grade can accommodate without


permanent deformation according to Lister (1977). Bituminous surfacing is a generic term for wearing and base courses that are essential for good ride quality to be combined with the appropriate resistance to skidding in asphaltic roads. And for this, texture and durability under traffic are required. In simplified highway pavement structure, the wearing course (uppermost layer) usually 40mm completes the flexible road pavement construction and is designed to withstand the direct effects of road traffic together with the action of weather and temperature conditions, O’Flaherty (2002). It was further suggested that a requirement for base course is usually 60mm; therefore, high stability can be obtained by using mixes with a high stone and low void content. Also, the aggregate should preferably be crushed stone having a high impact resistance and the use of harder bitumen grades is beneficial for increased resistance to deformation and fatigue cracking. As a resulting, the bituminous surfacings thickness ranges from 60mm to 100mm usually placed on concrete road base. Under traffic loading the various courses in a bituminous-bound pavement are subject to repeated stressing and the possibility of damage by fatigue cracking is usually considered to continually exist. Pavement distresses are usually manifested in form of cracking, rutting, ravelling, potholes, roughness, edge break, surface texture and polished surface. Shoving, cracking, rutting, ravelling and flushing may lead to break up of pavement. Specifically, pothole was defined in HDM111 as open cavity in road surface with at least 150mm diameter and at 25mm depth. Pothole may be defined as any localised loss of material or depression in the surface of a pavement that compromises the ride quality of the pavement. In general terms potholes may result from, deficiencies in the pavement, such as cracks, settlement in the utility cut, repair failure, overlay failure, poor construction and water is an important contributor to pothole formation mainly through loss of support caused by a saturated base.


3. DATA COLLECTION Since the majority of the failed roads are located in the southern region of Nigeria; the research boundary is confined to this area. Circumscribing the sample roads to those in the southern region was also thought to make the survey manageable in terms of the time and resources available to the researcher. Within the research boundary roads were also selected based on the following criteria; road Geometry ≥ class ‘B’ road FMWH design specifications, clear visibility and level terrain; road must exhibit visible multiple bituminous surfacing distresses that are capable of impairing significantly traffic movements. The surveyed sites are single-carriageway lanes type ‘B’ low cost state roads. All road links were coded for convenience referencing in alphabetical orders. The data for each day were screened for bad weather, incidents, equipment malfunctioning or usual traffic operation and general recording errors. Study crew would not normally check their own-recorded data at the close operation. The team was made up of a team leader, and four men. Team members were trained in equipment handling, public relations, and data recording. Tally sheets; work sheets; a packet of pen (red, blue, and black); tape measure and markers; pieces of stopwatch; pairs of walking-talkie; numbers of video camera; road cones; and a survey bus. Each road is divided into subsections of 100 or 200 metres with the road register marker posts used for reference. Then for each distress mode, the extent and severity of the defect are recorded supplemented by an assessment of their possible causes. This regime has been incorporated into this study by way of recording the numbers of potholes, area of distress, and the maximum depth of pothole; noting that the distribution of potholes is random and it may not necessary follow a particular pattern. Simple measurements are required for pavement distress as contained in most literature and they include: type of distress, length, width, depth, affected area, number (nos.), and the relative percentage of distress (see Table 2). Pavement distresses are classified into three classes; slight, moderate and severe. UK Department of Transport


DTp road note advice 20/84 (1997) suggested that for validity carriageway lane must not be less that 2.5m, therefore potholes, ravelling, and edge damages with transverse widths greater than 500mm on a 6.1m carriageway (allowing 100mm for road markings) would have violated lane width tolerance level. Low cost roads are usually single lane carriageway with 6.1m road width.

Table 2. Proportional distributions of potholes relative to pavement distress area

Potholes (PH) / Pavement Distress Area (PDA)

Road Code

Road Name Length

m

Width

m

Depth

mm

PDA

M2

% per km/Ln

PH

Nos

Severity

AN001 Enugu/Onitsha Road Onitsha 86 2.9 330 249.4 8.6 15 severe

DL004 Refinery Road Warri 44 2.5 220 110.0 4.4 13 severe

DL005 Warri / Sapele Road Warri, 48 2.6 200 124.8 4.9 07 severe

ED006 Ogida Road Benin 72 2.8 300 201.6 7.2 09 severe

ED007 Upper Sakponba Road Benin 53 2.6 150 137.8 6.3 11 severe

ED008 Upper Siliko Road Benin 40 2.5 200 100.0 4.1 17 severe

EK009 Ajilosun Street Ado-Ekiti 64 2.7 200 172.8 6.4 15 severe

OG010 Lantoro Road Abeokuta 46 2.5 200 115.0 4.6 12 severe

OG011 Aiyetoro Road Abeokuta, 28 2.4 300 67.2 3.0 14 severe

OG012 Oba Simolade Str. Shagamu 35 2.4 300 84.0 3.0 10 severe

OY018 Awolowo Avenue Ibadan 27 2.4 150 64.8 2.7 16 severe

OY019 Oyo Road Ibadan 54 2.6 200 140.4 6.4 13 severe

Average 50 2.6 230 130.7 5.1 13


Figure 1. Typical pavement distress 4. FINDINGS The study was based on the hypothesis that bituminous failure resulting from low cost road design is significant, costly, inefficient and unsustainable. It was conducted at twelve selected locations in the southern part of Nigeria. Bituminous surfacing of roads is classified as failing when evidences emerge of functional and / or structural deteriorations. The empirical results obtained from surveys conducted at twelve locations in the southern part of Nigeria suggest that potholes were the most significant form of bituminous surfacing distress. The highest recorded number of potholes at one location is 17 per km and the lowest is 7 per km. The highest recorded depth of pothole at one location is 330mm and the lowest is 150mm. The highest recorded area of bituminous distress at one location is 249m2 per km and the lowest is 65m2 per km. From the study it was found that site OY018 has the lowest relative pavement distress area of 65m2 while site AN001 has the highest relative pavement distress area of 249m2. Site DL005 has


the least nos. of pothole (7) while site ED008 has the highest nos. potholes (17). Interestingly, even though site ED006 had the second lowest road capacity loss of 25% with the second lowest level of pavement distress, the sites had the largest proportion of commercial vehicle suggesting that commercial vehicles suffer less from of pavement distress when compared to passenger cars and should be an area for further research. Considering that surfacing thickness of these so called low cost roads is between 30mm to 35mm, pothole depth greater than 35mm would allow ingress of surface water and expose the road base to weathering action. This in turn would accelerate pavement deterioration as evidenced above in Figure 1. Bituminous roads in tropical climates often deteriorate in different ways from those in the more temperate regions partly because of harsh climatic conditions. Harsh climatic conditions aside, it’s often the case that the traffic mix may include fairly significant proportions of pedestrians, bicycles and animals competing for road space with over-capacitated buses, Lorries, and trailers. Under such circumstances, one would expect the roads to be designed to minimum design standard with 100mm surfacing placed on concrete road base. On the contrary state and local governments opt for low cost roads with 30mm bituminous surfacing placed on un-stabilised laterite bases; an infertile soil and heavily leached tropical subsoil. Laterite is a red-coloured clay-like kind of infertile soil found in the tropics. It consists usually of aluminium oxyhydroxides with smaller amounts of iron oxyhydroxides and a little bit of a clay mineral called halloysite; when exposed and dried it sometimes is rock-like. It is not a fertile soil and does not satisfy the requirement for base course materials however, it can be stabilized using Portland cement or bitumen when used for base course construction. Cement is expensive and sparingly used in this part of the world. This should provoke discussion on an alternative stabilising agent for example bitumen that is in abundance supply and reduce dependence on Portland cement. In any case, it is evident from the discussion so far that low cost road


pavement designs are defective, so, it’s not surprising that the road surfacing fails few weeks after opening. 5. CONCLUSIONS The major materials used in road construction in Nigeria are lateritic soils, gravel and bitumen. Nigeria has abundant deposit of laterite, which does not satisfy the requirement for base course materials and also large a deposit of bitumen. Typical constituency of low cost road pavement has between 30mm to 35mm bituminous surfacing laid on 150mm and 250mm road base and sub-base respectively. The bases are usually made up of laterite, which is a red-coloured clay-like kind of soil found in the tropics. It can be concluded that:

• The extent of pavement distress associated with low cost failure is significant

• There are significant numbers of potholes at surveyed sites • Edge damage and attendant erosion are common low cost roads • Low cost road pavement designs are not ideal for state and

local maintained roads in Nigeria • Although the design and construction costs are low,

maintenance costs would make such designs expensive over time

• The notion that low cost road design is an alternative design approach is costly and defective

• It’s is traffic volume not costs that is the over-riding road design factor

Whether it is poor road maintenance or the lack of it that is responsible for the poor road conditions or it is quite possible that poor design, badly executed construction or a combination of any or all these factors can be called to account for road defects. The paper concludes that unless road policies are put in place to redress the imbalance between low cost roads and the standard road design, persistent socio-economy instability resulting from poor


road transportation will entrench poverty and adversely affect road sustainability.

REFERENCES Ben-Edigbe J (2002) Influence of Pavement Distress on Highway

Capacity Loss And Their PCE Values’ PhD Thesis University of Strathclyde Library, Glasgow Scotland

Cook J.R & Gourley C (2002), ‘A Framework for the Appropriate Use of Marginal Materials’ World Road Association (PIARC)-Technical Committee C12, Seminar in Mongolia,

Federal Ministry of Works and Housing (2000) ‘National Road Network Statistics for Maintenance Road Maintenance Draft Document,’ FMWH, Abuja FCT-Nigeria

Lister N.W. ‘Transport and Road Research Laboratory Report 375’ Transport Research Laboratory Crowthorne, England

O’Flaherty C.A. (2002) ‘Highway Traffic Design and Analysis’ 4th Edition Published by Butterworth-Heinemann, Jordan Hill Oxford, England

Petts R., ( 2002)‘The Low Cost Road Surfacing Initiative’ LCS March Working Paper, Transport Research Laboratory Crowthorne England

bookchapter_fka07

Documents

Transcript of bookchapter_fka07