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 1

Public Work Department Road
Specification
Azmi bin Hassan, Zulakmal Hj. Suffian
Chapter 2 Providing a Safe Economic and Durable 18

Highway Surface
Alan Woodside
Chapter 3 Determination of Periodic Pavement 37

Rehabilitation Treatment using
Pavement Condition Assessment (PCA)
Ahmad Kamel bin Abdul Malik
Chapter 4 Evaluation of Malaysian Asphaltic 49

Concrete Mixture using Superpave and
Marshall Mix Design Method
Juraidah Ahmad, Mohd Yusof Abdul
Rahman, Mohd Rosli Hainin, Mustaque
Hosasain
Chapter 5 Crack Progression Models for Flexible 63

Pavements
Sugeng Wiyono,Othman Che Puan,
Mohd Rosli Hainin
Chapter 6 Investigation on Tyre/Road Noise using 72
Ulster Load Tyre Road
Haryati Yaacob, Mohd Rosli Hainin, Alan
Woodside, David Woodward
Chapter 7 Performance of Hot Mix Asphalt using 93

Fine Crumber
Norhidayah Abdul Hassan, Mohd Rosli
Hainin, Haryati Yaacob
Chapter 8 The Effect of Rainfall on Asphalt 107

Surfacing Materials
Nursetiawan, David Woodward, Alan Strong
Chapter 9 Research into Sustainable Asphalt 119

Surfacing Mixes in the UK
David Woodward, Alan Woodside, Alan
Strong, Paul Phillips, Bob Allen
Chapter 10 Extent of Pavement Disintegration 133

associated with Low Cost Road Failure
J. Ben-Edigbe
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
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 FT 1, Jalan Beranang -
Gilsonite Seremban
2. Porous asphalt with FT 1, Tampin and
natural rubber Route 2, Kuantan
3. Porous asphalt with FT 2, Federal Highway
Europrene Sol-T
4. Porous asphalt with FT 3, Jalan Tebrau and
Cariphalte 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 Surface HRA Porous Safepave ULM SMA

type dressing asphalt
Structural nil **** ** * ** ***
Rutting nil *** **** **** ***** ****
*
Cracking * ***** *** ** *** ***
Ride nil *** **** ***** **** ****
Suitable ** *** ***** ***** **** ****
texture
Skidding
- Early ***** **** *** *** *** ***
- Normal ***** **** **** **** **** ****
Noise 82 80 74 76 76 76
dB(A)
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.
12.5 mm Nominal Sieve Size - KJG Rock Quarry
100
Marshall upper limits

90
80
70 Gradation
Percent Passing
60 Superpave upper limits Marshall lower
50
40
30
Superpave lower limits
20
10
12.50
19.00
0.075
1.18
4.75
0.15
0.30
0.60
2.36
9.50
Sieve Size (mm)
Figure 1. 9.5mm gradation aggregate structure
9.5 mm Nominal Sieve Size - KJG Rock Quarry
100
90
80
70
Percent Passing
60
50
40
30
20
10
0
12.50
0.075
0.60
2.36
9.50
1.18
0.15
0.30
4.75
Sieve Size (mm)

Figure 2. 12.5mm gradation aggregate structure
Table 1. Aggregate gradings
Kajang Rock Quarry

(KJG)
Mixture
12.5 mm 9.5mm
NMS NMS
Metric Sieve
Gradation (% Passing)
(US)
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
6.5 6.0
(No.100)
0.075mm
4.0 4.0
(No.200)
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
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

12.5- 12.5- 9.5- Specified
9.5-B2
Parameter B1 B2 B1 Limits
MS
MS MS MS (PWD)
OAC
5.6 6.1 6.2 6.4 -
(%)
Stability
10.1 10.2 10.2 10.7 >8kN
(kN)
Flow
3.5 3.5 3.2 3.3 2 – 4 mm
(mm)
VMA
16.4 17.2 17.2 17.6 -
(%)
VFA
77 75 75 75 70 – 80
(%)
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
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 12.5- 12.5- 9.5- 9.5-
Properties B1- B2- B1- B2- Criterion
SP SP SP SP
OAC (%) 5.1 5.3 5.4 5.7 -
Air Voids 4.0 4.0 4.0 4.0 -
(%) 14.7 15.6 15.6 16.3 14.0 min
VMA (%) 72.8 74.4 74.4 75.5 65-75
VFA (%) 87.0 86.7 87.0 86.3 89%
%Gmm @ 96.1 95.9 95.9 95.8 max
Nini 0.8 0.8 0.8 0.7 98%
% Gmm @ max
Nmax 0.6-1.2
Dust
Proportion
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.
7 Superpav e mix tures Mars hall mix tures
Optimum Asphalt Content (%)

6
0
12.5B1 12.5B2 9.5B1 9.5B2
M ix De s ig n
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.
18 Superpave Marshall
17.5
17
16.5
VMA (%)
16
15.5
15
14.5
14
13.5
12.5-B1 12.5-B2 9.5-B1 9.5-B2
Mix Design
Figure 4. Voids in mineral aggregate of mixtures
2.35
Superpave Marshall
2.34
2.33
Density (%)
2.32
2.31
2.3
2.29
2.28
2.27
12.5-B1 12.5-B2 9.5-B1 9.5-B2
M ix Des ign
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

KJG KJG KJG KJG
Mix Design 12.5 12.5 9.5 9.5
B1 B2 B1 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
Sat (%) 74.0 73.2 74.4 73.6
@
. TS(KPa) 514 616 627 626
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
Sat (%) 74.0 73.2 73.4 72.4
@
TS(KPa) 452 452 464 491
600C
TSR 96.4 97.3 85.4 96.8
Superpave mixtures Marshall mixtures

800
700
600
500
ITS (%)
400
300
200
100
0
12.5-B1 12.5-B2 9.5-B1 9.5-B2
Mix Design
Figure 6 Indirect tensile strength values of mixes

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.
ICX = K icx a0e
( a YE 4 / SNC )
1
2
(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.
CWx1000
ACT = (2)
CRKSPACE
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.
CRKSPACE = 0.305 AGE 2a0 10a1 (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
a2
CRX t = (1 − z )50 + z ⎡ za02 NEci + z 0.5a1 + (1 − z )50a2 ⎤

⎣ ⎦
(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, a01 = a0 SNCa1(for this study). TCI is the time
since cracking initiation and t50 is 50a2 - 0.5a2/a01 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.
LOCATION MAPE
101° 102° 103° 104°
MELACA P. PANJANG
2°lLU
KEC. KUBU
Bagan siapi-api Tj. Medang SE
LA
KEC. BANGKO KEC. RUPAT T P. NATUNA BESAR
P. RUPAT M WEST MALAYSIA
EL
Palu Panjang AK
ROKAN N II A
KEC. TANAH PUTIH DUMAI KEC. BENGKALIS JOHOR

KEC. BUKIT BATU P. BENGKALIS
BENGKALIS P. SAWANG NABANG
Rokan kiri KABUPATEN BENGKALIS
SINGAPORE
Kota Tengah KEC. BATAM

P. PADANG
1°lLU
Tanjung Medan KEC. KEPENUHAN Duri

BATAM Langai
KEC. MANDAU P. BINTAN

TANJUNG PINANG
SE
KEC. KARIMUN
LA
Kandis P. RANGSANG
T
KEC. RAMBAH
KEC. KUNTO P. TEBING TINGGI
DARUS Perawang KEC. SEI APIT PA
NJA Bengkalis
Pasir Pangarayan NG
Minas
Kota Lama P. TUPAN KEC. KUNDUR
Sei Garo
Sp. Lago
K A B U P A T E N K A M PA R KEC. SIAK SRI INDRA PURA P. MENDOL
Sei Galuh Sei Buatan
KEC. ROKAN IV KOTO
KEC. BANGKINANG
Bangkinang
PEKANBARU KEC. BUNUT
KAB. KEPULAUAN RIAU

Sp. Kualo P. BURUNG KATEMAN
BANGKINANG KEC. LANGGAM

P.KATEMAN
KEC. XIII KOTO
KAMPAR Guntung Kateman
Langgam
SU
0°
KEC. MANDAH
MA
KEC. PANGKALAN
KEC. KAMPARKIRI
TE
KURAS
RA
P. LINGGA
P. UJUNG BETING
BA
RA
KEC. RENGAT
TALUK KUANTAN KEC. GAUNG ANAK SERKA
T
Sp. Japura
KEC. KUALA INDRAGIRI
P. MATAK KEC. PASIR PENYU
RENGAT
TEMBILAHAN
KEC. SINGINGI
Sei. Laiak
P.SINGKEP
P.TEREMPA
Kuala Enok
KAB. INHU
NORTH
Peranap KAB. INHIL
KEC. PERANAP
1°lLS
KEC. KUANTAN MUDIK
J A M BI RIAU PROVINCE 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
Type of ICX (year) ACT (m/1000m) CRX (%)

Date pavement Paterson Simpson Paterson
Measured (1987) Measured et al Measured (1987)
(1994)
Mac AMAP 1.5 3.7182 0.4950 0.4421 0 0
2002 (Kandis
Sept Section) 0.6091 0.6579 1.78095 1.78090
2002
Mac 1.0104 1.4502 3.2746 23.6307
2003
Mac AMSB 2.5 3.7182 0.06964 0.14914 1.00310 1.18797

2002 (sorek
Sept Section) 0.20714 0.22363 3.24964 19.0242
2002
Mac 0.27767 0.31191 6.64695 86.5511
2003
Mac AMGB 3.5 3.7182 0.16888 0.22306 0.86850 0.68509

2002 (Sorek
Sept Section) 0.36520 0.75674 3.87280 7.72444
2002
Mac 0.59539 0.93366 5.48571 41.8153
2003
Table 2. Parameters for structural cracking initiation
Model ICX = Kicx a0e

( a yE 4 / SNC )
1
2
A0 a1 Kicx
Pavement
Type This Paterson This Paterson This Paterson
study (1987) study (1987) study (1987)
0.5 to
AMAP 9.48 8.61 -25.8 -24.4 0.75 to 2
2.5
0.5 to
AMSB 9.17 8.61 -25.1 -24.4 0.43
1.3
AMGB 8.8 8.61 -24.8 -24.4 0.49 -
Table 3. Parameters for transversal or thermal crack
Model CRKSPACE = 0.305 AGE 2a010a1

Value of a0 Value of a1
Climatic Zone Simpson Simpson
This This
et al., et al.,
study study
(1994) (1994)
-0.571 to 3.23 to
Wet-No Freeze -1.12 -
0.856 3.89
Table 4. Parameters for predicting structural crack progression

1
a2
CRX t = (1 − z )50 + z ⎡ za02 Neci + z 0.5a1 + (1 − z )50a2 ⎤

Model
⎣ ⎦
Parameters Model Statistics
A0 a1 a2 CV (%) R2
\
This Paterson This Paterson This Paterson This Paterson This Paterson
Study (1987) Study (1987) Study (1987) Study (1987) Study (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
Table 5 Statistic Value of validate progression structural crack and

thermal crack
R2 Value for R2 Value for
Structural Crack Thermal Crack
AMAP New 0.8765

AMAP 0.8264 0.823
Existing
AMGB 0.5925 0.879
AMSB 0.8158 0.751
4 CONCLUDING REMARKS
The findings of this study may be summarized as follows:

(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

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.
Microphone
h=100mm
Tyre seen from above
45º
d1=200mm
Microphone
d2= 200mm
Figure 1. Microphone position used in the ULTRA noise measurement

method
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 Grip
depth Num
20mm 2 PSV>60 2.00mm 0.65

HRA years gritstone
14/6mm 1 PSV>60 3.24mm 0.6
SD week gritstone
10/6mm 1 year PSV>60 0.97mm 0.74
SD gritstone
10mm 1 PSV>60 2.33mm 0.75
SD week gritstone
6mm 2 Blend of 2 PSV 1.12mm -
SD years >60 aggregates –
gritstone +
igneous
14mm 2 PSV>60 1.50mm 0.65
SMA years gritstone
10mm 1 PSV>60 mm 1.03 0.54
bitmac week gritstone mm
10mm <1 - 0.62mm -
Thinpave year
Smooth NA NA NA -
resin
Smooth NA NA 0mm -
steel
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.
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 Tyre Pressure Tyre Type Speed
80kg 40kg 20psi 30psi 40psi 100km/hr
Treaded 50km/hr
Tyre
65km/hr
Smooth *Playsafe only
Tyre
80 km/hr
*Playsafe only
95 km/hr
*Playsafe only
Figure 2. List of variables for noise measurement

Table 2.Total loading applied to the contact patch (McErlean, 2006)
Load applied to tyre Actual load applied under

assembly (kg) 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.
50 km/hr 100 km/hr
110
Road/tyre noise (dB)
105
y = 1.657x + 101.9
R2 = 0.4666
100
95
y = 1.9075x + 91.483
90 R2 = 0.6919
0 1 2 3 4
T exture depth (mm)
Figure 3. The relation between tyre/road noise and texture depth at load
40kg and pressure 20psi using a smooth tyre
50 km/hr 100 km/hr
110
Road/tyre noise (dB)

105
y = 0.6955x + 104.71
100 R2 = 0.3042
95
y = 0.6826x + 94.395
90 R2 = 0.43
0 1 2 3 4
Texture depth (mm)
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.
Smooth resin Sample with texture
100
90
Sound level (dB)
80
70
60
50
40
30
10 100 1000 10000 100000
Frequency (Hz)
Figure 5. Noise differences between textured surfaces and smooth resin

using a smooth tyre.
Smooth resin Sample with texture
100
90
Sound level (dB)

80
70
60
50
40
30
10 100 1000 10000 100000
Frequency (Hz)
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.
Resin with joints (smooth tyre)

Smooth steel (smooth tyre)
Resin with joints (treaded tyre)
Smooth steel (treaded tyre)
100
90
Sound level (dB)
80
70
60
50
40
10 100 1000 10000 100000
Frequency (Hz)
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 Noise at Noise
50km/hr (dB) 100km/hr (dB) 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.
108
106
Treaded
104
tyre
Noise data (dB)
102
100 Smooth
98 tyre
96
94
92
40 50 60 70 80 90 100 110
Speed (km/hr)
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.
40 kg 80 kg
100
99
98
Sound level (dB)
97
96
95
94
93
92
91
90
RA
ac
TP
A
PS
D
SD
SD
SD
SM
6S
tm
/6
10
/6
H
Bi
10
14
14
10
Figure 9. Effect of loading at 50km/hr and 20psi for a smooth tyre

20 psi 30 psi 40 psi
100
98
Sound level (dB)

96
94
92
90
88
86
D
A
SD
PS
RA
SD
SD
ac
TP
6S
SM
tm
10
/6
/6
H
10
14
14
Bi
10
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
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
Volumetric Dense graded (AC14)
Properties Conventi RU RUM RUM Specificatio
onal Mix MA AC AC n
C 2 3 (JKR/SPJ/re
1 v/2005)
Rubber
Content 0 1* 2* 3* -
(%)
OBC (%) 5.1 5.3 5.6
5
(80/1 (80/1 (80/1 -
(80/100)
00) 00) 00)
Stability, S 1314 1159 1093 587
> 815
(kg)
Flow, F
2.27 2.20 2.52 3.68 2.0 - 4.0
(mm)
Stiffness 526.
881.9 434.6 159.7 > 203
(kg/mm) 9
VTM (%) 3.4 3.8 4.3 4.6 3.0 - 5.0
VFB (%) 79.5 75.7 64.6 59.9 70 - 80
*Percentage of crumb rubber expressed from the total weight of
aggregate blend
Table 3. Marshall mix design results of SMA14 for conventional and

modified mixes
Volumetric Gap graded (SMA14)
Properties Conventional RUMAC RUMAC RUMAC Specification
Mix 1 2 3 (JKR/SPJ/rev/2005)
Rubber
Content 0 1* 2* 3* -
(%)
OBC (%) 6.7 6.8 7
7.3 (PG 76) Min 6 (AASTHO)
(80/100) (80/100) (80/100)
Stability, S
996 812 762 510 > 632
(kg)
Flow, F
2.24 2.52 2.60 2.72 2.0 - 4.0
(mm)
Stiffness
445.6 322.9 293.0 187.4 -
(kg/mm)
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)

Roller Rut Depth (mm)
Passes Dense graded (AC14)
(cycles) Conventio RUMAC RUMAC
nal Mix 1 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

16
14
12
R ut D ep th (m m )
10
8
CONVENTIONAL
6 RUMAC1
RUMAC2
4
0
0 500 1000 1500 2000 2500 3000 3500 4000
Cycles
Figure 1. Results of wheel tracking test for AC14
Table 5. Data from Wessex wheel tracker test (SMA14)

Roller Rut Depth (mm)
Passes Gap graded (SMA14)
(cycles) Convent RUM RUM RUM
ional AC AC AC
Mix 1 2 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

16
14
12
R u t D e pth (m m )
10
8
CONVENTIONAL
6 RUMAC1
RUMAC2
4 RUMAC3
0
0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500
Cycles
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.
nozzle
Flow
Meter Pressure
gauge
Water Supply
Bituminous material
Storage
tank
tank
pump
Figure 1. The Ulster rainfall simulator

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 Pressure Flow Average CU
Number (bar) rate Rainfall (%)
(l/s) Intensity
(mm/h)
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 Variable
Group
Test surface Surface type
Texture depth
Air voids
Hydraulic conductivity
Test condition Crossfall
Rainfall intensity
Duration of testing
Measured during Runoff
testing Percolation
Calculated Time for transition flow (Tft)
from Time for steady state flow (Tsf)
test data 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.
2% crossfall 4% crossfall
6% crossfall
20
Flow Rate (cubic cm/s)

15
10
0
0 120 240 360 480 600 720 840
Time (s)
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.
2% crossfall 4% crossfall
6% crossfall
20
Flow Rate (cubic cm/s)

15
10
0
0 120 240 360 480 600 720 840
T ime (s)
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).
Wood 10mm DBM

10mm Superflex 14mm Superflex
240
180
Ttf (s)
120
60
0
0 1 2 3 4 5 6 7 8
Crossfall (%)
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.
Crosfall 2% Crossfall 4% Crossfall 6%
240
180
Tft (s)
120
60
0
0.0 0.5 1.0 1.5
Texture depth (mm)
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.
RI 31mm/h RI 54mm/h RI 78mm/h
240
180
Tft (s)
120
60
0
0 0.5 1 1.5
Texture depth (mm)
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.
240
y = 0.8445x + 17.245
180 2
Calculated Ttf (s)

R = 0.8445
120
60
0
0 60 120 180 240
Predicted 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 26,500 5,600 32,100 16
State 10,400 20,100 30.500 16
Local 2,600 128,000 130,600 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 Length Width Depth PDA % per PH
Code km/Ln
Road Name m m mm M2 Nos Severit
y
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

Caricato da

Informazioni sul documento

Descrizione originale:

Titolo originale

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Bookchapter Fka07

Caricato da

Copyright:

Formati disponibili

First Edition 2008

All rights reserved. No part of this publication may be reproduced or transmitted

Perpustakaan Negara Malaysia Cataloguing-in-Publication Data

Pereka Kulit: MOHD. NAZIR MD. BASRI

Diatur huruf oleh / Typeset by

Diterbitkan di Malaysia oleh / Published in Malaysia by

Dicetak di Malaysia oleh / Printed in Malaysia by

Chapter 1 Speciality Mixes in the New Malaysian 1

Chapter 2 Providing a Safe Economic and Durable 18

Chapter 3 Determination of Periodic Pavement 37

Chapter 4 Evaluation of Malaysian Asphaltic 49

Chapter 5 Crack Progression Models for Flexible 63

Chapter 7 Performance of Hot Mix Asphalt using 93

Chapter 8 The Effect of Rainfall on Asphalt 107

Chapter 9 Research into Sustainable Asphalt 119

Chapter 10 Extent of Pavement Disintegration 133

Highway materials and construction is a vast field covering a

Hasanan Mohd Nor

The flexible pavement structure in Malaysia consists typically of

another common surfacing used mainly for FELDA and FELCRA

2. REQUIREMENTS FROM BITUMINOUS MIXES

The basic requirements from bituminous mixes are as follows:

a. Stability Resist permanent deformation

What gives us the requirements?

a. Stability Stone contact:

3. MODES OF PAVEMENT DISTRESS IN MALAYSIA

Top- down cracking

caused by the diurnal temperature change, or by traffic induced

normally provides adequate grip, but in wet conditions (about 25

4. NEW BITUMINOUS MIXTURE IN MALAYSIA

As an option to prolong the life and enhance the performance of

Polymer Modified Asphalt

subsequently in major rehabilitation projects. Various polymers

The inclusion of modifiers will incur an increase in cost of the

W In the new PWD road specifications, a PG-76 polymer modified

i. Improvement of skid resistance at high speeds, especially

ii. Reduction of splash and spray.

iii. Suppression of hydroplaning effects.

iv. Reduction of headlight reflection and glare on wet

v. Improvement of visibility at night.

Porous asphalt was introduced in Malaysia only in 1991 when the

Performance of Porous Asphalt

With the exception of Gilsonite modified Porous Asphalt, all other

Stone Mastic Asphalt

Advantages of SMA include;

i. Expected longer life than dense graded mixtures

In 1999, the PWD constructed a test section on SMA on Route 1,

Of the many types of mixes in the new PWD specification, the

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

11. Fernando M J and Guirguis H R. Natural rubber for

The economic growth of a nation can be directly related to the

field" sites in the UK or in many parts of Western Europe due to

3. WHOLE LIFE TESTING

The new European Standards state that "any aggregate may be

Nation and a Community or a Union one must develop levels of

Work carried out by Woodside and Woodward on unwanted,

minimum pull-off force necessary to remove chippings from

6. RECYCLING OF HOT ROLLED ASPHALT

The recycling of an existing road structure would appear to be an

is not without its problems, as experienced by the author, viz the

However, recent research at UUJ has shown that Penetration

7. THE REQUIREMENTS FOR ROAD PAVEMENT