You are on page 1of 44

Asphalt surfacing to bridge decks

Prepared for SSR Directorate (Highways Infrastructure)


Pavement Engineering Team, Highways Agency

J C Nicholls, R W Jordan and K E Hassan

TRL Report TRL655


First Published 2006
ISSN 0968-4107
ISBN 1-84608-654-X
Copyright TRL Limited 2006.

This report has been produced by TRL Limited, under/as part


of a contract placed by the Highways Agency. Any views
expressed in it are not necessarily those of the Agency.

TRL is committed to optimising energy efficiency, reducing


waste and promoting recycling and re-use. In support of these
environmental goals, this report has been printed on recycled
paper, comprising 100% post-consumer waste, manufactured
using a TCF (totally chlorine free) process.

ii
CONTENTS
Page

Executive Summary 1

1 Introduction 3

2 Questionnaire 3
2.1 Format 3
2.2 Responses 3

3 Laboratory test programme 3


3.1 Asphalt mixture properties 3
3.1.1 Objectives and test programme 3
3.1.2 Air voids content 4
3.1.3 Permeability 5
3.1.4 Indirect Tensile Stiffness Modulus 6
3.1.5 Wheel-tracking 8
3.1.6 Review of asphalt results 8
3.2 Asphalt flexibility 10
3.2.1 Objectives and test programme 10
3.2.2 Indirect tensile fatigue test 10
3.2.3 4-point bending fatigue test 10
3.2.4 Semi-circular bending test 10
3.2.5 Asphalt bending test 12
3.2.6 Discussion 12
3.3 Waterproofing system properties 15
3.3.1 Objectives and test programme 15
3.3.2 Torque bond tests 15
3.3.3 Interface permeability 16

4 Considerations for specification of surfacing for


bridge decks 17
4.1 General approach 17
4.2 Drainage 17
4.2.1 Drainage requirements 17
4.2.2 Removal of water 17
4.2.3 Sub-surface drainage 18
4.3 Waterproofing system 18
4.3.1 Bond to asphalt 18
4.3.2 Membrane stiffness 18
4.3.3 Effect of laying and compaction temperatures on
the waterproofing system 19

iii
Page

4.4 Asphalt properties 20


4.4.1 Deformation resistance 20
4.4.2 Texture depth 21
4.4.3 Skid resistance 21
4.4.4 Flexibility and fatigue 21
4.4.5 Permeability and air voids content 21
4.4.6 Protection for waterproofing systems 22
4.5 Joints and interface between layers 22
4.5.1 General 22
4.5.2 Joints 22
4.5.3 Bond between layers 22
4.5.4 Bond to waterproofing system 23
4.5.5 Permeability of interface between asphalt and
waterproofing system 23
5 Conclusions 24

6 Acknowledgements 24

7 References 25

Appendix A: HA draft notes for bridge-deck overlays 27

Appendix B: Questionnaire 28

Appendix C: Permeability tests 32

Appendix D: Semi-circular bending test 33

Abstract 40

Related publications 40

iv
Executive Summary

The current requirements in the Specification for Highway equal ranking. Material optimisation is usually avoiding
Works require waterproofing systems on concrete bridge any excessively adverse property rather than getting the
decks to be overlaid with a 20 mm thick sand asphalt best performance in all.
protection layer and binder and surface courses so that the The flexibility test programme involved applying four
total thickness of the three layers is 120 mm. These different tests to three control mixtures and six stone
requirements include that the material that directly mastic asphalt mixtures. The tests undertaken were the
overlays the waterproofing system should be sand asphalt indirect tensile fatigue test, the 4-point bending fatigue test
as well as considerations about the bond of the surfacing to at low temperature and low frequency, the semi-circular
the waterproofing system and the sub-surface drainage. In bending test from the Netherlands and an asphalt bending
service, it has been found that the performance of test from China. The results obtained from the test
surfacing on concrete bridges is generally satisfactory if programme are reviewed together with theoretical
the total thickness of the asphalt layers is at least 120 mm. considerations and the semi-circular bending test is
However, the total thickness on some bridges has to be identified as the most promising. This test was found to be
reduced for practical and/or economic reasons and, in such practical, equivalent to a controlled stress fatigue test and
cases, a number of premature failures have occurred when ranks materials similarly to binder content, a known
the asphalt has broken up and potholes have developed. component of flexibility if other things are constant. It is
Therefore, HA commissioned TRL to develop a suggested that initially values of 3.0 N/mm2 and 19 N/mm3/2
specification for the asphalt surfacing on bridge decks that could be set as the minimum values required for tensile
is suitable even when the surfacing has a total thickness of strength and fracture toughness, respectively. These limits
less than 120 mm. The objective of this work was to are considered practical because they were exceeded by
investigate parameters that would allow the development the majority of the mixtures tested.
of a specification for surfacings on concrete bridges that The results from the torque bond test on specimens
enhances the probability of achieving reasonable durability including waterproofing systems without secondary
when they are less than the standard thickness. compaction varied by an order of magnitude, with the two
The research has included a literature search, a waterproofing systems used ranking the surfacing
questionnaire and laboratory test programmes. The materials differently. In all cases, the failures were at the
literature search provided little useful information. The interface between the waterproofing membrane and the
results from the questionnaire, sent to individuals and tack/bond coat, so the failure stresses were more dependent
companies involved in the manufacture and laying of the on the properties of the waterproofing systems than the
current materials used on bridge decks, were asphalt mixtures. For one waterproofing system, the
disappointing, with only eight responses received. highest failure stresses were measured with a tack coat and
The laboratory test programme of asphalt mixtures that sand carpet whilst, for the other waterproofing system, the
have been, or could be, used for surfacing bridge decks has failure stresses differed for two different tack coats. The
identified some differences in their properties that have been high values for mastic asphalt were probably due to the
used as the basis for the specification. A supplementary test high temperature at which the material was laid and
programme was also undertaken to look at tests for
compacted and at which the tack/bond coat was activated.
measuring the flexibility of asphalt materials, an important
Based on these findings and other considerations,
property for both bridge deck surfacings (particular those
various additions and changes to the Design Manual for
with a thin deck) and other situations with relative thin
Roads and Bridges, Specification for Highway Works and
pavements over soft substrates. Also, tests were undertaken
Notes for Guidance on the Specification for Highway
on composite samples of concrete, waterproofing and
Works have been proposed. The main changes include:
asphalt to assess the bond achieved.
From the tests on a series of twelve asphalt mixtures, ! Sub-surface drainage is emphasised.
mastic asphalt was found to have the most suitable air ! Bond requirements are strengthened.
voids content, permeability and stiffness modulus ! Deformation requirements are specified for all mixtures
properties and was fifth at wheel-tracking, making it the within 100 mm of the surface.
best material overall for these tests. However, mastic
! Maximum air voids content limits on all asphalt
asphalt is a relatively expensive mixture, a factor that
mixtures.
cannot be excluded. A dense 0/10 SMA was the next best
despite having the 8th highest air voids content whilst an Aspects that were not fully covered are permeability
open 0/10 SMA was the worst, showing that the precise testing of the asphalt and at the interfaces. Potential tests
mixture design can be critical. The remaining mixtures have been identified that could be developed for
showed relatively similar overall ratings, but with the sand standardisation if these aspects are considered critical. The
carpet only ranked 8th overall. Nevertheless, when asphalt permeability can be covered by the surrogate of air
considering the appropriate materials, the choice is often a voids content, but this approach is more difficult at
trade off between properties and they will not usually have interfaces where more than one material is involved.

1
2
1 Introduction 2.2 Responses
The reaction to the questionnaire was very limited with
The current clause in the Specification for Highway Works
only eight responses received, of which only four
(MCHW 1) requires waterproofing systems on concrete
contained useful comments. The contents of the
bridge decks to be overlaid with a 20 mm thick sand asphalt
constructive replies are given in Section B.2 of Appendix B.
protection layer and binder and surface courses so that the
However, any conclusions from the survey can only be
total thickness of the three layers is 120 mm. A previous
regarded as indicative.
research project by TRL for the Highways Agency (HA) The replies were not consistent other than that the aim
found that the performance of surfacing on concrete bridges should be to produce surfacings with a service life of about
is generally satisfactory if the total thickness of the asphalt 15 years. This aim is consistent with the top end of the
layers is at least 120 mm. However, the total thickness on durability found for thin surfacing systems in conventional
some bridges has to be reduced for practical and/or situations (Nicholls and Carswell, 2004) given that the
economic reasons and, in such cases, a number of premature surfacing will need to be maintained at, if anything, a
failures have occurred when the asphalt has broken up and higher standard than typically on a highway. Nevertheless,
potholes have developed. The failures have been attributed it is an indication of what engineers, let alone the driving
to a number of factors, including: public, expect.
! The accumulation of sub-surface water in the asphalt. One reply indicated reservations about currently used
! Poor bond of the asphalt layers to the waterproofing materials but the respondent did not think anything else
system. currently available would produce a practical alternative
! Low compressive modulus of the waterproofing system. that was economic. A second respondent offered epoxy
asphalt, but that tends to be an expensive product that
! Low stiffness modulus of the asphalt layers. presents considerable risks if there is any distance between
Therefore, HA commissioned TRL to develop a the batching plant and the site. It is, therefore, only
specification for the asphalt surfacing on bridge decks that appropriate for important and/or conveniently located
is suitable even when the surfacing has a total thickness of bridges, although it was for such site that the responder put
less than 120 mm. The specification is intended to develop it forward.
further the existing HA advice for bridge deck surfacing, Another respondent did identify red sand carpet as a
which is reproduced as Appendix A. particular material that should be replaced. This
The research has included a literature search, replacement was already being considered because the
questionnaire and two laboratory test programmes. The limited deformation resistance it provides will become
results from the literature search and the questionnaire to more of a liability as the thickness of material over it
those that manufactured and laid the current materials used reduces. Nevertheless, some material with relatively small
on bridge decks were disappointing, with only eight aggregate particles will be needed to protect some types of
responses received from the questionnaire. However, the the waterproofing systems from being punctured.
initial laboratory test programme of asphalt mixtures that
have been, or could be, used for surfacing bridge decks has
identified some differences in their properties that have
3 Laboratory test programme
been used as the basis for the specification. The second
3.1 Asphalt mixture properties
laboratory test programme identified a test method for
assessing the flexibility of asphalt materials. 3.1.1 Objectives and test programme
The object of the initial laboratory test programme on
asphalt mixtures was to identify the values achieved by
2 Questionnaire various materials in tests used to measure the properties
assumed to be relevant to bridge deck surfacings. These
2.1 Format properties are air voids content, permeability, stiffness and
A questionnaire was issued to individuals and companies deformation resistance. Air voids content and permeability
involved in the manufacture and laying of the current were investigated because drainage of water is particularly
materials used on bridge decks in order to ascertain if they important on bridges whereas the stiffness and deformation
were aware of any deficiencies in the present systems and, resistance are required as standard properties needed for all
probably more importantly, whether they considered that surfacings. Flexibility was not included within this
some new materials, or combination of materials, could programme because there are no accepted tests to define
produce better results. It was anticipated that the replies the property, but is covered separately (Section 3.2).
would focus on practical aspects. There was a deliberate The results were required in order to be able to specify
policy to keep the questionnaire relatively short but to achievable levels of performance for future surfacing. The
allow recipients to extend their thoughts if they wished. known limitations on the site performance of some of the
The questionnaire, which is reproduced as Section B.1 of materials for specific parameters can be used to ensure that
Appendix B, was sent to a total of 32 people from 23 the requirements do not allow it to be used where it would
different organisations. not perform satisfactorily.

3
The materials tested had to represent those that have Table 3.1 Composition of mixtures
been, or could be, used on bridge decks. The mixtures that
were considered covered the range of potential mixture 30% 35% 55% 55%
Mastic Sand 0/10 0/14 0/14 0/10
types that might be used on bridge decks whilst not being Type asphalt carpet HRA HRA HRA HRA
too extensive. The mixtures included a variety of hot rolled
asphalts (HRA), dense bitumen macadams (DBM), stone 28 mm – – – – – –
mastic asphalts (SMA) and mastic asphalts. From these, 20 mm – – – 100 100 –
14 mm – – 100 93 95 100
the following twelve mixtures were selected for studying 10 mm 100 – 92 71 82 95
in the initial test programme: 6.3 mm 95 100 75 – – –
! Mastic asphalt BS 1447 (BSI, 1988). 3.35 mm 80 – – – – –
2.36 mm 60 97 66 61 45 45
! 0 % 0/2 HRA (Sand carpet) Column 6/1 in BS 594 0.600 mm – 90 58 53 35 35
(BSI, 2002). 0.300 mm – – – – – –
0.212 mm – 47 32 28 17 17
! 30 % 0/10 HRA – surface Column 6/3 in BS 594 0.075 mm 30 15 10 9 5.5 5.5
course (BSI, 2002).
! 35 % 0/14 HRA – surface Column 3/3 in BS 594 Binder content 8.0 10.3 7.8 6.4 6.5 6.5
Binder grade 15 pen 50 pen 50 pen 50 pen 100 pen 100 pen
course (BSI, 2002).
! 55 % 0/14 HRA – base Column 2/2 in BS 594 0/10 0/10 0/14 0/14
(roadbase), binder course (BSI, 2002). 0/20 0/6 SMA SMA SMA SMA
Type DBM SMA (open)* (dense) (open) (dense)
and regulating course
! 55 % 0/10 HRA – base Column 2/1 in BS 594 28 mm 100 – – – – –
(roadbase), binder course (BSI, 2002). 20 mm 97 – – – 100 100
14 mm 75 – 100 100 90 100
and regulating course
10 mm 62 100 90 100 60 75
! 0/20 DBM – binder course Tables 15 and 16 in 6.3 mm 47 95 35 50 40 50
BS 4987 (BSI, 2001). 3.35 mm 39 – 24 34 24 34
2.36 mm – 37 21 26 19 29
! 0/6 SMA Table 2, column 3 of 0.600 mm – – – – – –
prEN 13108-5 (CEN, 2000a). 0.300 mm 14 – – – – –
0.212 mm – – – – – –
! 0/10 SMA (open) Table 2, column 5 of
0.075 mm 5.5 10.5 9 11 9 11
prEN 13108-5 (CEN, 2000a).
! 0/10 SMA (dense) Table 2, column 5 of Binder content 4.7 7.7 6.7 6.9 7.4 7.6
Binder grade 100 pen 50 pen 50 pen 50 pen 50 pen 50 pen
prEN 13108-5 (CEN, 2000a).
! 0/14 SMA (open) Table 2, column 6 of * Also 0/10 PMB SMA (open) mixture except for binder grade for the
prEN 13108-5 (CEN, 2000a). flexibility programme (Section 3.2).

! 0/14 SMA(dense) Table 2, column 6 of


A separate core from the cored slab was tested for bulk
prEN 13108-5 (CEN, 2000a).
density and the air voids content calculated in accordance
For convenience, the sieve sizes current at the start of the with BS EN 12697-8 (CEN, 2003b) using the maximum
work were used to define the gradings for the first seven density for the previous core from the same slab. The
mixtures rather than those implemented in 2004. Similarly, permeability of the cores was then measured using the TRL
the SMA mixtures were derived from gradings given in the permeability cell, as described in Section C.1 of Appendix C.
draft for development of the harmonised European Standard The remaining two slabs of each mixture were tested for
(CEN, 2000a) after conversion to the ‘old’ sieve sizes. The wheel-tracking in accordance with BS EN 12697-22
target gradings, based on the mid point of the envelopes for (CEN, 2003c) using Procedure A with the small size
most of the mixtures but on the envelope boundaries for the device, one slab at 45 ºC and the other at 60 ºC.
target gradings of 0/10 and 0/20 SMA mixtures to give the
permitted extremes, are given in Table 3.1. 3.1.2 Air voids content
Four 300 mm by 300 mm by 50 mm thick slabs were The results of the density and air voids content
mixed to BS EN 12697-35 (CEN, 2004a) and compacted measurements are given in Table 3.2, where ‘ITSM’ and
by roller-compactor to BS EN 12697-33 (CEN, 2003a) for ‘Permeability’ refer to the other test carried out on the
each of the mixtures. Five 100 mm diameter cores were samples on which the measurements were made.
then cut from two of the slabs of each material in The bulk densities measured varied within a range of
accordance with BS EN 12697-27 (CEN, 2000b). 2.18 to 2.46 Mg/m³ and the maximum densities within a
One core from each cored slab was tested, consecutively, range of 2.27 to 2.51 Mg/m³. These values are of limited
for bulk density to BS EN 12697-7 (CEN, 2002a), indirect value in themselves because they will be more dependent
tensile stiffness modulus at 20 °C and then at 40 °C to on the density of the aggregate than the properties of the
BS DD 213 (BSI, 1993) and maximum density to BS EN asphalt. Such variation is hidden for this investigation
12697-5 (CEN, 2002b). The air voids contents were then because aggregate from the same source was used for all
calculated in accordance with BS EN 12697-8 (CEN, 2003b) mixtures. However, the combination of bulk and
using the bulk density and maximum densities. maximum density to produce air voids content is

4
Table 3.2 Air voids content results

Sample Bulk density Max density Air voids content

Material ID ITSM Permeability ITSM ITSM Permeability

Mastic asphalt A 2.38 2.37 2.38 0.2 % 0.4 %


B 2.38 2.38 2.37 -0.4 % -0.4 %
Mean 2.38 2.38 0.0 %

0/2 HRA A 2.19 2.22 2.27 3.3 % 2.1 %


(Sand carpet) B 2.20 2.18 2.27 3.2 % 3.9 %
Mean 2.20 2.27 3.1 %

30 % 0/10 HRA A 2.34 2.36 2.39 2.2 % 1.4 %


B 2.29 2.32 2.38 3.9 % 2.5 %
Mean 2.33 2.39 2.5 %

35 % 0/14 HRA A 2.28 2.31 2.42 5.9 % 4.7 %


B 2.26 2.31 2.37 4.9 % 2.7 %
Mean 2.29 2.40 4.6 %

55 % 0/14 HRA A 2.37 2.38 2.45 3.0 % 2.7 %


B 2.41 2.36 2.45 2.0 % 3.7 %
Mean 2.38 2.45 2.8 %

55 % 0/10 HRA A 2.39 2.38 2.44 2.0 % 2.4 %


B 2.38 2.40 2.45 2.6 % 1.9 %
Mean 2.39 2.45 2.2 %

0/20 DBM A 2.46 2.44 2.51 2.0 % 3.0 %


B 2.44 2.43 2.50 2.6 % 2.8 %
Mean 2.44 2.51 2.6 %

0/6 SMA A 2.38 – 2.42 3.8 % –


B 2.33 – 2.42 2.6 % –
Mean 2.35 2.42 3.2 %

0/10 SMA A 2.22 2.21 2.41 8.2 % 8.3 %


(open) B 2.21 2.21 2.42 8.6 % 8.7 %
Mean 2.21 2.42 8.5 %

0/10 SMA A 2.35 2.37 2.44 3.5 % 2.8 %


(dense) B 2.35 2.35 2.43 3.5 % 3.6 %
Mean 2.35 2.43 3.3 %

0/14 SMA A 2.27 2.24 2.42 6.0 % 7.2 %


(open) B 2.25 2.25 2.42 7.1 % 7.1 %
Mean 2.25 2.42 6.8 %

0/14 SMA A 2.36 2.37 2.45 3.5 % 3.0 %


(dense) B 2.30 2.37 2.43 5.3 % 2.6 %
Mean 2.35 2.44 3.6 %

significant. For the mixtures tested, the air voids content where: ρbulk = bulk density (Mg/m³)
ranged from zero for the mastic asphalt to 8.5 % for the ρmax = maximum density (Mg/m³)
0/10 SMA (open) mixtures. The width of the range and the νair = air voids content (%)
ranking of mixtures within it are consistent with what was bc = binder content (%)
expected, with the denser, more binder rich mixtures
having less air voids content than the more open mixtures. There is little correlation between binder content and air
The relationships between the binder content of the various voids content because the binder contents are selected
mixtures and their densities are shown in Figure 3.1 and based on the mixture type and aggregate skeleton.
that between the binder content and air voids content in However, the correlation between binder content and
Figure 3.2. Included on the figures are also linear trend lines, maximum density is surprisingly robust.
the equations for which are given as Equations (3.1) to (3.3).
ρbulk = 2.56 – 0.033 bc (R2= 0.34) ....(3.1) 3.1.3 Permeability
The permeability testing was carried out on asphalt cores
ρmax = 2.69 – 0.039 bc (R2= 0.84) ....(3.2)
(100 mm diameter and height between 70 mm and 100 mm).
νair = 5.03 – 0.198 bc (R2= 0.014) ....(3.3) Air was applied at the desired pressure (P) and the flow

5
2.55
2.50
Max density
2.45

Density (Mg/m³)
Bulk density
2.40
2.35
2.30
2.25
2.20
2.15
4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0

Binder content (%)

Figure 3.1 Relationship between binder content and density

10.0

8.0
Air voids content (%)

6.0

4.0

2.0

0.0
4.0 5.0 6.0 7.0 8.0 9.0 10.0 11.0
Binder content (%)

Figure 3.2 Relationship between binder content and air voids content

rate was calculated using a bubble flowmeter with diameter, SMA and open SMA. The most obvious outlier is 35 % 0/14
Df, length, Lf, and the average flow time, taverage. The flow HRA, which has a higher air voids content and, to a greater
rates were used for the calculation of the air permeabilities, extent, permeability than the other HRA mixtures.
as described in Section C.1 of Appendix C, and the results Nevertheless, this ranking is quite similar to that reported
are given in Table 3.3. earlier (Daines, 1995). It is important to highlight that the
The permeability ranking of the different asphalt mixtures permeability results are obtained from laboratory specimens
is shown in Table 3.4. The lowest permeability values were with high quality control, which may not represent the
obtained from the mastic asphalt mixture, which exhibited actual values of the same mixtures constructed on site.
the lowest air voids content of around 0 %. The permeability However, the ranking is expected to be the same.
of the HRA mixtures varied over a range of two orders of Figure 3.3, with the data points plotted on a previously
magnitude for air voids content between 2.3 % and 5.4 %. found relationship (Zoorob, 1999), shows the influence of
The permeability and air voids content of the sand carpet air voids content on the permeability of asphalt. Whilst there
and the 50 % stone HRA mixtures were lower than those of
are not adequate data to establish a reliable relationship,
the 30 % and 35 % stone HRA mixtures. In fact, for the
particularly in terms of mixtures with high air voids
same aggregate size (0/14), increasing the aggregate content
contents, a similar trend can be fitted to that found for
from 35 % to 55 % reduced the air voids content from 5.4 %
concrete. Regardless of the mixture type, low permeability
to 2.4 % and the permeability by approximately two orders
values are obtained when the air voids content is 4 % or
of magnitude. The DBM and dense SMA mixtures gave
below and the permeability increases rapidly for air voids
permeability values within the lower band of HRA. In
contrast, the high air voids content of the 0/6 and open SMA contents above 5 %. In the absence of a specification limit
mixtures with >5 % air voids contents resulted in high for permeability, it appears reasonable to use the air voids
permeability values. These values were out of the limit of content as an indicator for permeability.
the flowmeter used with the TRL permeability cell.
The results in Table 3.4 shows that the permeability 3.1.4 Indirect Tensile Stiffness Modulus
ranking, from low to high, of the different asphalt mixtures The results of the Indirect Tensile Stiffness Modulus
is generally as follows: mastic asphalt, HRA, DBM, dense (ITSM) test are given in Table 3.5.

6
Table 3.3 Permeability results

Sample Flow meter Permeability (10-17 m2)

Df Lf P taverage
Material ID (mm) (mm) (bar) (s) Sample Mean

Mastic asphalt A 1.7 50 0.8 25.0 0.55 0.73


B 1.7 50 0.8 15.0 0.91
0 % 0/2 HRA A 1.7 100 0.8 13.3 1.96 2.27
B 1.7 100 0.8 10.1 2.58
30 % 0/10 HRA A 1.7 100 0.6 3.3 12.52 8.09
B 1.7 100 0.6 10.1 3.66
35 % 0/14 HRA A 10 100 0.4 7.6 276 219
B 10 100 0.4 14.0 162
55 % 0/14 HRA A 1.7 50 0.4 20.0 1.56 1.85
B 1.7 50 0.4 14.9 2.14
55 % 0/10 HRA A 1.7 100 0.6 21.2 1.84 2.50
B 1.7 100 0.6 12.4 3.16
0/20 DBM A 10 – 0.8 – Permeable 3.41
B 1.7 100 0.8 8.0 3.41
0/6 SMA A 10 – 0.8 – Permeable Permeable
B 10 – 0.8 – Permeable
0/10 SMA (open) A 10 – 0.8 – Permeable Permeable
B 10 – 0.8 – Permeable
0/10 SMA (dense) A 1.7 100 0.8 20.5 1.34 2.93
B 1.7 100 0.8 6.0 4.52
0/14 SMA (open) A 10 – 0.8 – Permeable Permeable
B 10 – 0.8 – Permeable
0/14 SMA (dense) A 1.7 100 0.8 4.8 5.64 4.47
B 1.7 100 0.8 9.0 3.30

Table 3.4 Permeability ranking of the asphalt mixtures Table 3.5 ITSM results

Air voids Air permeability ITSM @ 20 °C (GPa) ITSM @ 40 °C (GPa)


Mixture content (%) (10-17 m2)
Material A B Mean A B Mean
Mastic asphalt <1 0.73
HRA 2.3 – 5.4 1.85 – 219 Mastic asphalt 5.13 6.38 5.75 0.55 0.54 0.55
DBM 2.3 3.4 0/2 HRA 1.43 1.54 1.48 0.11 0.15 0.13
Dense SMA 3.5 – 4.4 2.9 – 4.5 30 % 0/10 HRA 2.12 2.29 2.20 0.17 0.24 0.21
Open SMA 5.2 – 8.4 Permeable 35 % 0/14 HRA 3.00 2.22 2.61 0.32 0.22 0.27
55 % 0/14 HRA 0.99 1.07 1.03 0.15 0.15 0.15
55 % 0/10 HRA 1.08 1.25 1.16 0.15 0.18 0.16
0/20 DBM 1.84 1.47 1.65 0.15 0.19 0.17
300
0/6 SMA 1.89 2.08 1.99 0.16 0.16 0.16
250 0/10 SMA (open) 1.38 1.43 1.40 0.10 0.09 0.09
0/10 SMA (dense) 2.10 2.05 2.08 0.17 0.16 0.16
Air permeability (10-17 m2)

200
0/14 SMA (open) 1.57 1.41 1.49 0.12 0.10 0.11
0/14 SMA (dense) 1.51 1.64 1.58 0.10 0.19 0.15

150
The ITSM of the mixtures ranged from 1.0 GPa to 5.8 GPa
100
at 20 °C and 0.09 GPa to 0.55 GPa at 40 °C. However, the
results for mastic asphalt are effectively outliers, with the
50
range reducing to 1.0 GPa to 2.2 GPa at 20 °C and 0.09 GPa
0 to 0.27 GPa at 40 °C without them. The high value for
0 1 2 3 4 5 6 7 8 9 mastic asphalt is probably due to the high binder content
Air voids (%)
giving relative high tensile strength. Ignoring mastic asphalt,
the lack of any correlation between ITSM and air voids
Figure 3.3 Influence of air voids content on the content is shown in Figure 3.4. The plotted trend line at
permeability of asphalt 20 °C, including mastic asphalt, has a correlation coefficient
of R² = 0.23, but the value reduces to 0.0006 when the
mastic asphalt is omitted.

7
6.0
o
5.0 ITSM @ 20 C

Indirect tensile stiffness


o
ITSM @ 40 C

Modulus (kPa)
4.0

3.0

2.0

1.0

0.0
0.0 2.0 4.0 6.0 8.0 10.0
Air voids content (%)

Figure 3.4 Stiffness modulus variation with air voids content and temperature

The effect of temperature on ITSM is also shown in Table 3.7 Wheel tracking test results
Figure 3.4. The ratio of ITSM at 20 °C to ITSM at 40 °C
ranged from 6.9 to 15.4 for the mixtures tested with the Testing @ 45 ºC Testing @ 60 ºC

ranking order being marginally different for the two Wheel Wheel
temperatures, as shown in Table 3.6. tracking Rut tracking Rut
rate depth rate depth Overall
Table 3.6 Ranking of ITSM results Material (µm/cycle) (mm) (µm/cycle) (mm) ranking

ITSM (GPa) ITSM Ranking Ratio 0/10 SMA (dense) 0.7 3.2 2.7 5.6 1
0/14 SMA (open) 1.0 3.1 3.6 6.6 2
@ @ @ @ 20 °C 0/6 SMA 1.2 3.4 3.3 8.0 3
Mixture 20 °C 40 °C 20 °C 40 °C Mean / 40 °C 0/10 SMA (open) 1.2 3.0 4.1 7.1 4
Mastic asphalt 0.8 2.3 6.7 11 5
Mastic asphalt 5.8 0.55 1 1 1 10.5 0/20 DBM 0.9 2.7 6.7 7.7 6
35 % 0/14 HRA 2.6 0.27 2 2 2 9.6 0/14 SMA (dense) 2.1 11 4.6 9.4 7
30 % 0/10 HRA 2.2 0.21 3 3 3 10.7 35 % 0/14 HRA 7.6 9.6 14 17 8
0/10 SMA (dense) 2.1 0.16 4 5= 4 12.7
55 % 0/10 HRA 6.0 10 55 84 9
0/20 DBM 1.7 0.17 6 4 5 9.8 55 % 0/14 HRA 4.1 7.1 44 71 10
0/6 SMA 2.0 0.16 5 5= 6 12.3 30 % 0/10 HRA 7.9 12 70 98 11
0/14 SMA (dense) 1.6 0.15 7 8= 7 10.8 0 % 0/2 HRA 18 25 76 103 12
55 % 0/10 HRA 1.2 0.16 11 5= 8 7.2
0/2 HRA 1.5 0.13 8= 10 9 11.5
0/14 SMA (open) 1.5 0.11 8= 11 10 13.8
worst at deformation resistance with the other HRA
55 % 0/14 HRA 1.0 0.15 12 8= 11 6.9 mixtures in a band between the sand carpet and the other,
0/10 SMA (open) 1.4 0.09 10 12 12 15.4 almost non-deforming mixtures. Therefore, sand carpets
are not suitable for use in the top 100 mm of surfacing
The ranking orders are similar for the two temperatures, whilst other HRA mixtures should be designed against
but not identical. The top three materials are the same in deformation to SHW clause 943 (which was not the case
for the trialled mixtures) for use in that region.
both with the material that has the greatest change in
An inconsistency among the mixtures is the 0/14 SMA
ranking being 55 % 0/10 HRA, which is 11th at 20 °C but
(dense) at 45 ºC, in which the rate of deformation started
5th= at 40 °C. The differences demonstrate that the choice
high before reducing considerably so that the rut depth is
of test temperature will be important if the asphalt stiffness
in the range for the HRA mixtures tested but the wheel
needs to be a parameter when specifying the stiffness of tracking rate is only marginally higher than the other SMA
the waterproofing system. mixtures. Therefore, this result is assumed to be an outlier
caused by excessive bedding in at the beginning of the test.
3.1.5 Wheel-tracking
The results of the wheel tracking test are given in Table 3.7 3.1.6 Review of asphalt results
with the plots of the tests undertaken at 45 ºC and 60 ºC The rankings from the various tests can be combined to
been shown in Figure 3.5 and Figure 3.6, respectively. produce a rough overall ranking as in Table 3.8. However,
The plots for the tests at 60 ºC show dramatically that when considering the appropriate materials, the choice is
the HRA mixtures, with the possible exception of the 35 % often a trade off between these properties and they will not
0/14 HRA, were less deformation resistant than the DBM usually have equal ranking. Material optimisation is usually
and SMA mixtures. At the lower temperature of 45 ºC, the avoiding any excessively adverse property rather than
plots show that 0/2 HRA (sand carpet) is significantly the getting the best performance in all. Nevertheless, mastic

8
20

18

16

14
Rut depth @ 45 C (mm)

12
o

10

0
0 100 200 300 400 500 600 700 800 900 1000
Number of load cycles

0% 0/2 HRA 30% 0/10 HRA 35% 0/14 HRA 55% 0/14 HRA 55% 0/10 HRA 0/20 DBM
0/6 SMA 0/10 SMA open 0/10 SMA dense 0/14 SMA open 0/14 SMA dense Mastic Asphalt

o
Figure 3.5 Plot of wheel-tracking results @ 45 C

20

18

16

14
Rut depth @ 60 C (mm)

12
o

10

0
0 100 200 300 400 500 600 700 800 900 1000
Number of load cycles

0% 0/2 HRA 30% 0/10 HRA 35% 0/14 HRA 55% 0/14 HRA 55% 0/10 HRA 0/20 DBM
0/6 SMA 0/10 SMA open 0/10 SMA dense 0/14 SMA open 0/14 SMA dense Mastic Asphalt

o
Figure 3.6 Plot of wheel-tracking results @ 60 C

9
Table 3.8 Overall rankings for asphalt properties The test methods undertaken on the nine mixtures were:
a Indirect tensile fatigue test.
Air voids Perme Stiffness Wheel Sum of
Test content -ability modulus -tracking rankings b 4-point bending fatigue test.
c Semi-circular bending test.
Mastic asphalt 1 1 1 5 8
0/10 SMA (dense) 8 5 4 1 18 d Asphalt bending test.
0/20 DBM 4 6 6 6 22
55 % 0/10 HRA 2 4 8 10 24 The indirect tensile fatigue tests were carried out in
30 % 0/10 HRA 3 8 3 11 25
accordance with Annex E of BS EN 12697-24 (CEN,
0/6 SMA 7 11= 5 3 26 2004b) at 20 °C except that the deformations were measured
55 % 0/14 HRA 5 2 11 9 27 vertically rather than horizontally because Cooper Research
35 % 0/14 HRA 10 9 2 8 29 Technology Limited, manufacturer of the fatigue equipment
0/2 HRA 6 3 9 12 30 used, do not offer a horizontal deflection measurement
0/14 SMA (open) 11 11= 7 2 31 system. The specimens had to go through the test two or
0/14 SMA (dense) 9 7 10 7 33
0/10 SMA (open) 12 11= 12 4 39
three times before they fractured because the test stops at a
deflection of 8 mm. The results are the cumulative number
of cycles and deflections combined.
asphalt is demonstrated to be the best material for all the The 4-point bending fatigue tests were carried out in
properties other than deformation resistance, when it was accordance with Annex D of BS EN 12697-24 (CEN,
still in the top half. However, mastic asphalt is a relative 2004b) with a loading device capable of imparting a fixed
expensive mixture, a factor that cannot be excluded.
adjustable amplitude up to ± 5 mm with a frequency of
The maximum density (and, to a lesser extent, bulk
6 cycles/h at a temperature of 5 ºC and a sample size of
density) correlated well with the binder content, but not
300 mm × 40 mm × 40 mm.
with the air voids content because the binder content was
The semi-circular bending tests were carried out in
selected based on the mixture type and aggregate skeleton.
accordance with Appendix D, repeated on separate
Therefore, assuming air voids content is considered to be
samples with and without a notch, at a temperature of
indicative of permeability, the binder content cannot be
15 °C and a deformation rate of 5.1 mm/min. The draft
used as a surrogate for permeability without reference to
method is based on a method developed in the Netherlands
the mixture type.
and has recently been put forward for standardisation as a
European test method to assess the crack propagation
3.2 Asphalt flexibility property of asphalt.
3.2.1 Objectives and test programme The asphalt bending tests were carried out in accordance
Flexible asphalt has obvious advantages for use in with Appendix E at temperatures of 0 °C and 15 °C and a
situations where there is potential movement. Such deformation rate of 50 mm/min. The draft method is based on
situations include roads on ‘soft’ foundations as well as PRC Requirement T0715-93 (Peoples Republic of China,
bridge decks. However, there are currently no test 1993) except that, for practicality, the deformation was
requirements for flexibility in either the British Standards measured at the top of the specimen rather than the bottom.
for asphalt or in the British Board of Agrément (BBA) During the tests, it appeared that compression of the
Highway Authorities Product Approval Scheme (HAPAS) specimens was minimal (also evidenced by the low test loads
for thin surfacing systems. The inclusion of such a test as achieved before failure) which would imply very similar
an option in the BBA/HAPAS scheme has been requested deformations at the bottom and the top of the specimens.
by representatives from the County Surveyors Society.
Because of the synergy with the bridge deck issue, the 3.2.2 Indirect tensile fatigue test
Highways Agency extended the scope of this work to The mean fatigue lives of the various asphalt mixtures
include a programme of tests to identify a suitable test and tested for different applied tensile stresses are summarised
typical values for the property. in Table 3.9. An approximate ranking of the mixtures has
The test programme devised consisted of undertaking been included.
four different tests that were proposed by members of These values are plotted on Figure 3.7, from where the
BBA-HAPAS Specialist Group 3 on a series of asphalt rough ranking order can be seen.
mixtures. The known ranking for the properties required of
some of the mixtures tested was hoped to provide an
3.2.3 4-point bending fatigue test
insight into which of the tests had potential to assess the
flexibility of asphalt mixtures. The mean fatigue lives in terms of time of the various
From the mixtures used in the asphalt material trials asphalt mixtures tested are summarised in Table 3.10. An
(Section 3.1.1), the mastic asphalt, 30 % 0/10 HRA and 55 % approximate ranking of the mixtures has been included.
0/14 HRA mixtures were selected as the control
mixtures for the flexibility programme with the SMA 3.2.4 Semi-circular bending test
mixtures, including a variant of the 0/10 PMB SMA The values calculated for the tensile strength (from the un-
(open) mixture that used styrene-butadiene-styrene notched samples) and the fracture toughness (from the
block co-polymer (SBS) modified binder instead of notched samples) of the asphalt mixtures tested are given
fibres, as the trial materials. in Table 3.11. An approximate ranking of the mixtures has

10
Table 3.9 Summary of mean fatigue lives in ITFT

Applied tensile stress (kPa)

Material 100 150 250 350 500 570 750 Ranking

Mastic asphalt – – – – 155,000 106,000 5,280 1


0/14 SMA(dense) – 176,000 11,900 6,810 – – – 2=
0/10 SMA (dense) – 162,000 15,600 7,250 – – – 2=
30 % 0/10 HRA – 117,000 16,900 6,470 – – – 2=
0/6 SMA – 94,300 12,500 5,130 – – – 2=
0/14 SMA (open) – 64,200 9,440 2,360 – – – 6
0/10 SMA (open) 43,200 20,200 4,670 – – – – 7=
55 % 0/14 HRA 95,000 19,400 3,100 – – – – 7=
0/10 PMB SMA (open) 27,000 5,620 1,670 – – – – 9

5.5

5.0
Log (Fracture life)

4.5

4.0

3.5

3.0
0 100 200 300 400 500 600 700

Applied tensile stress (kPa)

30 % 0/10 HRA 55 % 0/14 HRA Mastic asphalt


0/6 SMA 0/10 SMA (open) 0/10 PMB SMA (open)
0/10 SMA (dense) 0/14 SMA (open) 0/14 SMA (dense)

Figure 3.7 Plot of mean fatigue lives

Table 3.10 Summary of fatigue lives in 4-point bending Table 3.11 Summary of semi-circular bending test results

Fatigue life (h) for sample: Tensile strength Fracture toughness

Material 1 2 3 Mean Ranking Combined


Material (N/mm²) Ranking (N/mm3/2) Ranking ranking
0/10 PMB SMA (open) 4.87 8.39 6.65 6.63 1
55 % 0/14 HRA 4.85 6.41 3.25 4.83 2= Mastic asphalt 8.29 1 50.24 1 1
0/6 SMA 10.0 1.81 2.00 4.61 2= 30 % 0/10 HRA 3.86 2 24.68 2= 2
0/10 SMA (open) 5.87 2.82 1.94 3.55 4 0/14 SMA (dense) 3.24 3= 23.03 2= 3
30 % 0/10 HRA 3.53 2.92 1.68 2.71 5 0/14 SMA (open) 3.35 3= 19.95 4= 4
0/6 SMA 3.03 5= 19.63 4= 5=
0/10 SMA (dense) 2.92 1.48 1.81 2.07 6
0/14 SMA (dense) 1.99 1.03 1.55 1.52 7= 0/10 SMA (dense) 3.07 5= 19.88 4= 5=
0/14 SMA (open) 1.05 1.86 0.76 1.22 7= 0/10 SMA (open) 2.43 7 17.35 7 7
Mastic asphalt 0.77 1.38 1.05 1.07 9 55 % 0/14 HRA 1.88 8= 14.45 8 8=
0/10 PMB SMA (open) 1.89 8= 11.87 9 8=

11
been included for both properties, which were similar, Table 3.13 Summary of asphalt bending test results
together with an overall ranking by rounding the average @ 15 ºC
of the two rankings.
Bending Maxm Bending
strength bending strain stiff. modulus
3.2.5 Asphalt bending test
The values calculated for the bending strength, maximum Value Ran- Ran- Value Ran-
Material (MPa) king Value king (GPa) king
bending strain and bending stiffness modulus of the
asphalt mixtures tested are given in Table 3.12 for a test 0/14 SMA (dense) 14.86 1 0.0101 9 1503 1
temperature of 0 ºC and in Table 3.13 for a test 0/10 SMA (dense) 6.19 3= 0.0190 7= 348 2=
temperature of 15 ºC. An approximate ranking of the 0/14 SMA (open) 6.71 2 0.0205 4= 331 2=
0/10 PMB SMA (open) 6.26 3= 0.0203 4= 314 4
mixtures has been included for each property plus an
Mastic asphalt 4.81 6= 0.0174 7= 276 5
overall ranking combining them for each temperature.
30 % 0/10 HRA 6.27 3= 0.0289 3 218 6=
0/10 SMA (open) 4.80 6= 0.0217 4= 228 6=
Table 3.12 Summary of asphalt bending test results 55 % 0/14 HRA 3.83 8 0.0324 1= 121 8
@ 0 ºC 0/6 SMA 2.41 9 0.0354 1= 72 9

Bending Maxm Bending


strength bending strain stiff. modulus The bending strength and bending stiffness modulus
rankings are in approximately the reverse order of the
Value Ran- Ran- Value Ran- rankings for maximum bending strain. An overall ranking
Material (MPa) king Value king (GPa) king can be derived by averaging the rankings for the bending
strength and bending stiffness modulus and the inverse
0/14 SMA (dense) 11.0 1= 0.00238 6= 4.69 1
0/10 PMB SMA (open) 8.3 5= 0.00229 6= 3.75 2= (i.e. 10 minus the actual ranking) of the ranking for the
30 % 0/10 HRA 10.9 1= 0.00295 4 3.70 2= maximum bending strain, as shown in Table 3.14.
0/10 SMA (dense) 8.4 5= 0.00223 6= 3.85 2=
0/10 SMA (open) 6.6 9 0.00231 6= 2.85 5= 3.2.6 Discussion
55 % 0/14 HRA 10.7 1= 0.00380 2 2.91 5=
0/14 SMA (open) 7.8 7= 0.00255 5 3.05 5= 3.2.6.1 Practicality
Mastic asphalt 8.0 7= 0.00329 3 2.43 8
The complexity of the potential flexibility tests differ. The
ITFT took considerably longer than the other tests because the
0/6 SMA 9.3 4 0.00501 1 1.88 9
fatigue life has to be between 103 and 106 cycles with each
cycle lasting 0.5 s (0.1 s loading time and 0.4 s rest period),
A plot of the ranking of the six different measures (three giving a total duration of between 8 min and 6 days
properties at two temperatures) is shown in Figure 3.8. continuous working for a single determination with three
The order of the materials has been revised in accordance determinations required per test. Of the other three tests, the
with their average ranking from the bending strength and four-point bending fatigue test was the most complex to set
bending stiffness modulus at both temperatures to assist up whilst the semi-circular bending and asphalt bending tests
the comparison. were both the simplest to set up and quickest to perform.

9.0

8.0
7.0
6.0
Ranking

5.0
4.0
o
3.0 Max bending strain @ 15 C
o
2.0 Max bending strain @ 0 C
o
Bending strength @ 15 C
1.0 o
Bending strength @ 0 C
0.0 o
Bending stiffness modulus @ 15 C
0/6 SMA

0/10 SMA (open)

o
Bending stiffness modulus @ 0 C
Mastic asphalt

55 % 0/14 HRA

0/14 SMA (open)

0/10 PMB SMA (o)

30 % 0/10 HRA

Test
0/10 SMA (dense)

0/14 SMA (dense)

Material

Figure 3.8 Relative rankings of the properties measured in the asphalt bending test

12
Table 3.14 Development of an overall ranking for the The mastic asphalt and 0/10 PMB SMA tend to be the
asphalt bending test extremes on the rankings, certainly for the first three tests
although, compared to the ITFT and semi-circular bending
Ranking tests, the order is reversed for the 4-point bending and, to a
Max Bending
lesser extent, the asphalt bending test. The reason is that
bending Bending stiff. the tests undertaken measure different properties at
strain strength mod. different temperatures.
The properties and temperatures are:
@ @ @ @ @ @
Material 0ºC 15ºC 0ºC 15ºC 0ºC 15ºC Overall ! ITFT: Fatigue life under controlled stress (force)
conditions at 20 ºC.
0/14 SMA (dense) 6= 9 1= 1 1 1 1
! 4-point bending: Fatigue life under controlled strain
0/10 SMA (dense) 6= 7= 5= 3= 2= 2= 2
0/10 PMB SMA (open) 6= 4= 5= 3= 2= 4 3 (displacement) conditions at 5 ºC.
0/14 SMA (open) 5 4= 7= 2 5= 2= 4= ! Semi-circular bending: Tensile strength and/or fracture
30 % 0/10 HRA 4 3 1= 3= 2= 6= 4=
toughness at 15 ºC.
Mastic asphalt 3 7= 7= 6= 8 5 6=
! Asphalt bending: The bending strength, maximum
0/10 SMA (open) 6= 4= 9 6= 5= 6= 6=
55 % 0/14 HRA 2 1= 1= 8 5= 8 8 bending strain and/or bending stiffness modulus at both
0/6 SMA 1 1= 4 9 9 9 9 0 ºC and 15 ºC.
Controlled stress and controlled strain tests are known to
result in reversed rankings. Therefore, it can be assumed
3.2.6.2 Comparison of rankings that the tensile strength and the fracture toughness in the
The overall rankings derived from each of the four tests semi-circular bending test are aligned to constant stress
are compared in Table 3.15 with the ranking plotted in rather than constant strain conditions.
Figure 3.9. Figure 3.10 shows the ranking orders when the materials
are listed in order of their semi-circular bending test results
Table 3.15 Ranking order by different tests whilst Figure 3.11 shows them when ordered for the 4-point
bending test. Both these figures show that there a consistency
4-point Semi- Asphalt in the order for three of the tests (even if the 4-point bending
Material ITFT bending circular bending
test order has to be reversed), but that the overall ranking
0/14 SMA(dense) 2= 7= 3 1 derived from the asphalt bending test shows no consistency
0/10 SMA (dense) 2= 6 5= 2 with the others.
0/10 PMB SMA (open) 9 1 8= 3 Plotting the individual rankings from the asphalt
30 % 0/10 HRA 2= 5 2 4= bending test on graphs with the same axes as Figure 3.10
0/14 SMA (open) 6 7= 4 4= and Figure 3.11 did not produce any obvious equivalence.
Mastic asphalt 1 9 1 6= Therefore, the properties derived from the asphalt bending
0/10 SMA (open) 7= 4 7 6= test are not considered to be well correlated with those
55 % 0/14 HRA 7= 2= 8= 8 from the other three tests.
0/6 SMA 2= 2= 5= 9
Mastic asphalt could be expected to be the most flexible
because of its high binder content, although that binder is

9
8
7
6
Ranking

5
4
3
2
Asphalt bending
1
Semi-circular
0 4-point bending
30% 10 HRA

55% 14 HRA

Mastic asphalt

ITFT
0/6 SMA

0/10 SMA (o)

Test
0/10 PMB SMA

0/10 SMA (d)

0/14 SMA (o)

0/14 SMA (d)

Material

Figure 3.9 Plot of observed rankings for each test

13
10
9
8
7
6

Ranking
5
4
3
2
1
0
Mastic 30 % 0/14 SMA 0/14 SMA 0/6 SMA 0/10 SMA 0/10 SMA 55 % 0/10 PMB
asphalt 0/10 HRA (dense) (open) (dense) (open) 0/14 HRA SMA
(open)

ITFT 4-point Semi-circular Asphalt bending

Figure 3.10 Relative rankings in order of semi-circular bending ranking

10
9
8
7
6
Ranking

5
4
3
2
1
0
0/10 PMB 55 % 0/6 SMA 0/10 SMA 30 % 0/10 SMA 0/14 SMA 0/14 SMA Mastic
SMA 0/14 HRA (open) 0/10 HRA (dense) (open) (dense) asphalt

ITFT 4-point Semi-circular Asphalt bending

Figure 3.11 Relative rankings in order of 4-point bending ranking

harder than that for other mixtures which could lead to less mixture compared to the fibres, as measured by the semi-
flexibility, particularly at colder temperatures. If the binder circular bending test. A possible for this lack of enhancement
content is the overwhelming criteria for flexibility, then the is that the SBS binder was used as a replacement rather than
ranking would be as follows: designing a new mixture with SBS binder.
1 Mastic asphalt (8.0 % binder) The binder content ranking was not consistent with that
obtained from the 4-point bending test.
2 30 % 0/10 HRA (7.8 % binder)
3 0/6 SMA (7.7 % binder)
3.2.6.3 Theoretical considerations
4 0/14 SMA (dense) (7.6 % binder)
There is a need for flexibility in bridge-deck surfacing and
5 0/14 SMA (open) (7.4 % binder) surfacings over weak substrates if they are to overcome
6 0/10 SMA (dense) (6.9 % binder) slow but repeated movements, particularly at low
7= 0/10 SMA (open) (6.7 % binder) temperatures. Therefore, the concept of the 4-point bending
7= 0/10 PMB SMA (open) (6.7 % binder) test at 5 ºC appears be appropriate. However, the need for
controlled stress or control strain conditions in that test
9 55 % 0/14 HRA (6.5 % binder)
depend on the overall structure – constant strain conditions
This ranking is very close to that obtained with the semi- are indicative of thick, stiff pavements (where other layers
circular bending test. The differences are that 0/6 SMA will take more of the load when the surfacing weakens with
was moved up two places and 55 % 0/14 HRA swapped fatigue) while constant strain conditions are indicative of
places with 0/10 PMB SMA (open). These differences thinner pavements (where the deflections will increase as
could be due to an improvement in flexibility with smaller the material weakens and flexes to resist the applied force
aggregate size. without much help from other layers). Therefore, any test
What is unexpected is that the styrene-butadiene-styrene for flexibility should be based on controlled stress rather
(SBS) modified-binder did not enhance the flexibility of the than controlled strain, implying the ITFT rather than the

14
4-point bending fatigue test or, by default, the semi-circular type. Both waterproofing manufacturers offer at least two
test. However, the ITFT takes considerable time and, different tack/bond coats for different types of asphalt, so
therefore, is not particularly practical. they selected tack/bond coats that were deemed most
Therefore, the recommendation from theoretical suitable for these mixtures. Two cores were taken from
considerations aligned to the results of the test programme is each composite specimen and tested for torque bond in
to use the semi-circular bending test. The median value of accordance with the HAPAS thin surfacing guidelines
results obtained were a tensile strength of 3.07 N/mm2 and a (BBA, 2000). The holes were then refilled and the
fracture toughness of 19.88 N/mm3/2, so suggested initial composite specimens re-compacted for 250 passes with the
limits that are known to be attainable are a tensile strength roller compacter set at 1 bar in order to try to simulate
of 3.0 N/mm2 and a fracture toughness of 19 N/mm3/2. trafficking before two further cores were taken and tested
for torque bond.
3.3 Waterproofing system properties Each composite specimen was sliced vertically down the
centre to provide 300 mm by 100 mm cut faces with the
3.3.1 Objectives and test programme
waterproofing system along the centre. These faces were
The objective of the laboratory study of waterproofing
then tested for permeability at the interface between asphalt
systems was to assess the variation of, and typical values
and the waterproofing system using apparatus developed for
for, the bond strength between them and possible
measuring the initial surface absorption of concrete (BSI,
overlaying asphalt materials. It was hoped that the
1996), as shown in Section C.2 of Appendix C.
knowledge would be useful in setting realistic specification
limitations which become more important as the thickness
of the layers is reduced. In addition, the opportunity was 3.3.2 Torque bond tests
taken to investigate the interlayer permeability because The results of the initial torque-bond test are given in
water will always find the weakest link, and that is usually Table 3.16 together with the results after simulated
the joints. A pavement structure will not be impermeable, trafficking in Table 3.17.
even when constructed of impermeable materials, if the In order to estimate the ultimate shear stress of a failure
horizontal and/or vertical joints between those materials surface from a torque bond test, assumptions must be made
are permeable. concerning the stress-strain curve of the material at the
Eight 300 mm by 300 mm by 55 mm thick blocks were failure surface. If the material yields so the shear stress at
manufactured using C40 concrete and waterproofing failure is uniform across the failure surface, the maximum
systems applied. Two different membranes were used, shear stress can be calculated from the theoretical
each with two different tack coats. The waterproofing relationship given in Equation (3.4).
systems were then overlaid with 40 mm depth of asphalt,
compacted using a roller compactor to give an overall 12 000 × T
depth of around 100 mm (block plus waterproofing plus τ= …. (3.4)
π × d3
surfacing). The mixtures used were mastic asphalt, 0/2 HRA
(sand carpet), 30 % 0/10 HRA and 0/10 SMA (dense) from where: = shear stress in N/mm²
Section 3.1.1 such that there was one slab with each of the T = maximum torque in N.m
combinations of waterproofing manufacturer and asphalt d = diameter in mm

Table 3.16 Torque-bond results before simulated trafficking

Waterproofing system
Diameter Height Max. torque Shear stress Failure
Membrane Tack coat Surfacing type ID (mm) (mm) (N.m) (N/mm²) location

B S Mastic asphalt 1 98.49 46.13 72 0.29 – 0.38 1


2 98.20 46.15 50 0.20 – 0.27 1
B R 0 % 0/2 HRA 1 98.27 35.15 118 0.47 – 0.63 1
2 98.33 35.83 145 0.58 – 0.77 1
B S 30 % 0/10 HRA 1 97.87 40.21 70 0.29 – 0.38 1
2 98.47 40.13 45 0.18 – 0.24 1
B S 0/10 SMA (dense) 1 98.54 41.31 42 0.17 – 0.22 1
2 98.26 40.53 20 0.08 – 0.11 1

A P Mastic asphalt 1 97.91 43.17 310 1.26 – 1.68 1


2 97.87 44.76 320 1.30 – 1.74 1
A P 0 % 0/2 HRA 1 98.60 45.93 180 0.72 – 0.96 1
2 98.62 46.62 205 0.82 – 1.09 1
A Q 30 % 0/10 HRA 1 97.95 45.86 20 0.08 – 0.11 1
2 98.28 46.51 48 0.19 – 0.26 1
A Q 0/10 SMA (dense) 1 98.02 43.83 130 0.53 – 0.70 1
2 98.02 44.00 125 0.51 – 0.68 1
Location of failure 1: Debonded at interface between tack/bond coat and waterproofing membrane.
15
Table 3.17 Torque-bond results after simulated trafficking

Waterproofing system
Diameter Max. torque Shear stress Failure
Membrane Tack coat Surfacing type ID (mm) (N.m) (N/mm²) location

B S Mastic asphalt 1 – – – 2
2 – – – 2
B R 0 % 0/2 HRA 1 97.83 40 0.16 – 0.22 1
2 – – – 2
B S 30 % 0/10 HRA 1 – – – 2
2 – – – 2
B S 0/10 SMA (dense) 1 – – – 2
2 – – – 2

A P Mastic asphalt 1 97.65 340 1.39 – 1.86 1


2 97.73 340 1.39 – 1.86 1
A P 0 % 0/2 HRA 1 97.92 260 1.06 – 1.41 1
2 97.73 200 0.82 – 1.09 1
A Q 30 % 0/10 HRA 1 – – – 2
2 – – – 2
A Q 0/10 SMA (dense) 1 97.99 137 0.56 – 0.74 1
2 97.73 117 0.48 – 0.64 1

Location of failure 1: Debonded at interface between tack/bond coat and waterproofing membrane.
Location of failure 2: Material debonded during trafficking.

If the stress is proportional to the strain at the failure 7= Membrane B and tack coat S with (c.30 N.m)
surface, the shear stress at failure will increase from zero at 0/10 SMA, and
the centre to a maximum at the outer diameter. The Membrane A and tack coat Q with
maximum stress will be given by Equation (3.5). 30 % 0/10 HRA.
According to BD 47/99 (DMRB 2.3.4), the minimum
16 000 × T shear bond strength of an asphalt layer to a bridge deck
τ= …. (3.5)
π × d3 waterproofing system should be 0.2 MPa at 23 ºC. Table 3.16
shows that, for each membrane, the failure stresses before
In practice, the stress-strain curve of the material at the simulated trafficking exceeded the minimum for three
failure surface will fall somewhere between the fully mixtures but were below it for the fourth mixture. In all
elastic and fully plastic behaviour – strain hardening is cases, the failures were at the interface between the
unlikely. Therefore, the true shear stress will be within the waterproofing membrane and the tack/bond coat, so the
failure stresses were more dependent on the properties of
range calculated by Equations (3.4) and (3.5) that are
the waterproofing systems than the asphalt mixtures. For
shown in Table 3.16 and Table 3.17. the Membrane B systems, the highest failure stresses were
Of the initial results, the torque bond varied considerably measured with the tack coat R tack coat and sand carpet.
by an order of magnitude with the two waterproofing For the Membrane A systems, the failure stresses were
systems giving different rankings for the different materials. higher for tack coat P than for tack coat Q. The high values
For all materials other than 30 % 0/10 HRA, the results were for mastic asphalt were probably due to the high
higher for Membrane A systems than for Membrane B temperature at which the material was laid and compacted
systems. Overall, the results can be ranked as follows: and at which the tack/bond coat was activated.
The additional passes of the roller compactor at
1 Membrane A and tack coat P with (c.300 N.m)
elevated temperature to simulate trafficking did not prove
mastic asphalt.
successful, with debonding occurring, particularly for the
2 Membrane A and tack coat P with (c.200 N.m) Membrane B systems, rather than the gain in adhesion.
sand carpet. Therefore, the results after simulated trafficking have
3= Membrane B and tack coat R with (c.130 N.m) been ignored from any analysis. However, it should be
noted that most of the samples that debonded had a low
sand carpet, and
shear stress before trafficking.
Membrane A and tack coat Q with
0/10 SMA.
3.3.3 Interface permeability
5= Membrane B and tack coat S with (c.60 N.m)
The permeability at the interface between the asphalt
30 % 0/10 HRA, and
surfacing and the waterproofing was carried out on the
Membrane B and tack coat S with following three composite systems with the results being
mastic asphalt. given in Table 3.18:

16
! Mastic Asphalt – Membrane A and tack coat P likely to have the greatest effect on the durability of the
! 0 % 0/2 HRA – Membrane A and tack coat P surfacing on bridges concern the following properties:
! 0/10 SMA (dense) – Membrane A and tack coat Q ! The permeability of the asphalt and the control and
removal of water.
Table 3.18 Flow rate at interface between asphalt and ! The bond of the asphalt to the waterproofing system.
waterproofing ! The resistance of the asphalt to fatigue cracking.
Average flow rate (ml/m2/s) ! The compatibility with the waterproofing system.

Time Mastic 0% 0/10 SMA


(min) asphalt 0/2 HRA (dense) 4.2 Drainage

1 0.043 7.05 136


4.2.1 Drainage requirements
2 0.044 7.78 125 Whereas sub-surface water can permeate downwards
3 0.043 7.34 135 through several bound and unbound layers as it flows
5 0.042 6.99 131 towards the sub-surface drainage systems on pavements,
10 0.046 7.14 118
15 0.052 6.90 114 sub-surface water on bridges can only flow downwards as
30 0.047 6.83 110 far as the waterproofing system. It must then flow
60 0.045 7.88 111 horizontally towards any sub-surface drainage systems that
are located at the low points and may be well over ten
Mean 0.045 7.18 123
metres away from where the water entered the surfacing.
Therefore, drainage and the movement of water through
The test specimens were taken from the surfaced asphalt have a greater impact on the durability of asphalt
concrete blocks used for the torque bond tests. The on bridge decks than on pavements.
diameter of the test areas was 85 mm and they were Clause 4.1 of BD 47/99 (DMRB 2.3.4) requires surface
approximately 25 mm deep. The measured flow rate was water to be removed from bridge decks by the provision of
dependent on the permeability of the concrete, the falls and suitable surface drainage systems. Clause 4.2 also
permeability of the interface between the asphalt and the requires sub-surface drainage where natural drainage is not
waterproofing system, and the permeability of the asphalt possible. Some expansion joints provide sub-surface
within the test section. Because the permeability of the drainage by incorporating 20 mm square slotted drainage
concrete was low, most of the flow was through the channels that run across the carriageway. However, these
interface and the asphalt itself. drains can become clogged over a period of time so they
A very low flow rate was measured on the mastic asphalt no longer drain away the water quickly enough. Through-
specimen because both the interface and the asphalt were of deck drains at the low points are considered to be a more
low permeability. The flow rate was, on average, about 160 effective means of sub-surface drainage. Whilst these may
times higher on the sand asphalt specimen than on the be installed on new bridges, they are not always installed
mastic asphalt specimen. Because of the nature of sand on old bridges during maintenance works unless specific
asphalt, it is unlikely that the permeability of the interface problems have been encountered.
was significantly different to that of the asphalt itself.
Therefore, Specifications for the asphalt surfacing on
However, it is possible that the permeability of the base of
bridge decks should be ensure:
the layer of sand asphalt was less than that at mid-layer or at
the top of the layer, the latter having an influence on the ! Minimum accumulation of water by identifying suitable
results shown in Table 3.3. requirements for the asphalt used in each layer.
The flow rate was, on average, about 2700 times higher ! Efficient surface and sub-surface drainage systems to
on the 0/10 (dense) SMA specimen than on the mastic facilitate drainage, aided by longitudinal gradients and
asphalt specimen. Because the waterproofing membrane crossfalls of the bridge deck and the overlaying asphalt,
was overlaid with a thick layer of tack coat Q, the interface as are currently specified.
permeability should not have been significantly higher
than that of the SMA. Therefore, most of the flow should 4.2.2 Removal of water
have been through the SMA itself.
The accumulation of water within an asphalt layer,
whether on a bridge or a pavement, has a detrimental
effect in areas trafficked by heavy good vehicles. This is
4 Considerations for specification of
because wheel loading induces high hydrostatic pressures
surfacing for bridge decks in saturated asphalt that are sufficient to weaken the layer
and break it up so that potholes are formed. Also, the
4.1 General approach pressures have an adverse effect on the bond between
The Specification for the surfacing on bridge decks must layers and, in particular, the bond of asphalt to bridge deck
include the requirements normally applied to asphalt waterproofing systems and, hence, on the structural
surfacings, e.g. skidding resistance, resistance to integrity of the asphalt layers on bridge decks. There have
deformation and surface regularity, as well as further been a number of failures of the surfacing on bridge decks
criteria specific to asphalt on bridges. The specific criteria where water has accumulated in the asphalt, particularly on

17
the high side of certain types of expansion joint that form a standard thickness with a sand asphalt layer performs well
barrier to the flow of water. Many of these failures have even if, apparently, it may be poorly bonded to the
occurred after periods of heavy rain. waterproofing system. However, surfacing of less than
The main criterion influencing the penetration and standard thickness has not been durable when it has not
movement of water is the permeability both of the asphalt remained firmly bonded. In these circumstances, water has
itself (see also Section 4.4.5) and of the interface between accumulated on the waterproofing system and weakened/
layers (see also Section 4.5.5). Permeability is a measure failed the bond of the surfacing to the waterproofing system
of the connectivity of pores/voids within the binder matrix, and/or the surfacing itself. The durability of the surfacing
aggregate and the interface between them. The void appears to be more reliant on the bond when the thickness,
content at the interface between the asphalt layers and the and hence dead weight, of the surfacing is low.
waterproofing system is dependent on the characteristics of Where the surfacing is less than 120 mm thick, the bond
the asphalt layer and the waterproofing system, so they strength requirements are much higher in other countries
have been considered together. compared to those in the UK. For example, the current
Japan Highway Public Corporation specification requires
4.2.3 Sub-surface drainage waterproofing systems to be overlaid with asphalt of total
thickness 75 mm. The 35 mm thick layer directly overlaying
If the surface course material is permeable (see Section 4.4.5),
the waterproofing system is SMA with a 0/5 mm aggregate
significant amounts of water could reach the lower asphalt
size. The 40 mm thick surface course is a drainage (porous
layer(s) so that the potential for water to enter these layer(s)
asphalt) layer with a 0/13 mm aggregate size. Performance
through air voids, defects and improperly sealed joints and
tests on waterproofing systems include tests to measure the
reach the waterproofing system will be greater. Therefore, to
shear and tensile bond strength of the weakest interface of
minimise the risk of premature failure, an edge of
specimens comprising a concrete block, the waterproofing
carriageway drainage system should be provided to drain the
system and an asphalt layer. The shear bond strength must
full depth of a permeable surface course. Furthermore, the
not be less than 0.8 MPa at -10 °C and 0.15 MPa at 20 °C.
risk should be reduced further by specifying a surface course
The tensile bond strength must not be less than 1.2 MPa at
system which incorporates a thick bond coat that when
-10 °C and 0.6 MPa at 20 °C.
applied to the lower asphalt layer helps to seal it.
The current German specification requires waterproofing
systems on major roads to be overlaid with surfacing of total
4.3 Waterproofing system thickness from 70 mm to 80 mm. A 35 mm thick protective
4.3.1 Bond to asphalt layer of Gussasphalt with a 0/8 mm aggregate size directly
overlays the waterproofing system. The surface course is
According to Clause B4.2 (l) of BD 47/99 (DMRB 2.3.4),
35 mm to 45 mm thick. The tensile bond strength of the
the minimum tensile bond of an asphalt layer to a bridge
weakest interface of specimens comprising a concrete block
deck waterproofing system should be 0.1 MPa at 23 ºC.
and the waterproofing system must not be less than 0.7 MPa
According to Clause B4.2 (k), the minimum shear bond
at 8 °C and 0.4 MPa at 23 °C. The shear bond strength of
should be 0.2 MPa at -10 ºC and 23 ºC, and 0.1 MPa at
the weakest interface of specimens comprising a concrete
40 ºC. However, because BD 47/99 requires the total
block, the waterproofing system and an asphalt layer must
thickness of the asphalt layers to be 120 mm, it is implied
not be less than 0.15 MPa at 23 °C when the shear force is
that these bond strength requirements are applicable for
applied at an angle of 15º to the plane of the specimen.
surfacing of this standard thickness.
Therefore, it is proposed that the minimum tensile and
The bond of asphalt to a waterproofing system is
shear bond requirements are increased from the values
dependent on the characteristics of both the waterproofing
given in BD 47/99 (DMRB 2.3.4) when the surfacing is
system and the asphalt, e.g. the adhesive and cohesive
less than 120 mm thick. Furthermore, if the surfacing is
properties of the membrane, tack coat and asphalt, not just
120 mm or more thick, higher bond strength requirements
the ‘thickness’ of the tack coat. It is also dependent on the
should be specified if the waterproofing system is not
temperature at which the asphalt is laid and compacted and
overlaid with an additional protective layer of sand asphalt,
whether any tack coat is activated.
because of the increased risk of water accumulating on the
The bond strength requirements in BD 47/99
membrane and weakening the bond. The higher bond
(DMRB 2.3.4) refer to the initial bond, but Clause 2005.5 of
strength should not be achievable unless the contact area
the Specification for Highway Works (MCHW 1) states that:
between the asphalt and waterproofing system is high, i.e.
‘The additional protection layer 1 or surfacing laid on there are few voids at the interface. To ensure that the
the waterproofing system shall be fully bonded to the bond is not susceptible to the presence of water or varies
system for the life of the system.’ with time, the bond strength should be measured both
The bond is prone to failure in service when the surfacing before and after exposure to water.
became saturated and high hydrostatic pressures are
generated by wheels of heavy goods vehicles. Surfacing of 4.3.2 Membrane stiffness
Although well-designed pavements should not suffer
fatigue cracking, there has been concern that the surfacing
1
The additional protection layer is specified in the DMRB and SHW overlaying waterproofing systems on bridge decks can be
as a 20 mm layer of sand asphalt. susceptible to fatigue when the asphalt layers are thin.

18
The susceptibility of asphalt layers to fatigue is If systems are to be overlaid with mixtures containing
dependent on the strains induced in the asphalt and its larger sized aggregates, the membrane must pass the tests
fatigue properties. For a given wheel load, the strains with the aggregate indentator heated to 125 °C. Not all of
induced are dependent on the thickness and stiffness the waterproofing systems currently registered for use on
properties of the waterproofing system, the stiffness Highways Agency bridges have passed this test at 125 °C
properties of each asphalt layer, the combined thickness of and can be overlaid with coarse mixtures.
the layers, and the bond of the asphalt layers to each other Allied to the aggregate indentation requirements are those
and to the waterproofing system. In addition, cracking of in Clause 901.9 of the SHW (MCHW 1), which states that:
the upper asphalt layers can be induced because of the
‘With the exception of sand asphalt carpet, bituminous
break-up of the lower layers by high hydrostatic pressures
materials with a temperature greater than 125 °C
when they become saturated.
shall not be deposited on a bridge deck waterproofing
The strains induced by wheel loading are enhanced by
system unless adequate precautions are taken to avoid
the local and global deformation of the substrate to which
heat damage in accordance with a good industry
the waterproofing system and asphalt are applied. The
practice. A maximum temperature of 145 °C is
deformation of the substrate of steel bridge decks (the deck
permitted for sand asphalt carpet.’
plate) is significant. However, concrete bridge decks have
a local stiffness that is considerably higher than that of The BBA Roads and Bridges Agrément Certificates for
steel decks, so strains induced by the local deformation of the different waterproofing systems currently registered
the substrate are relatively low. Similarly, strains induced include the following statements:
by the global deformation of decks are considered to be
‘Temperature of the APL or HRA surfacing when
lower that those due to the other factors listed above.
applied should exceed the minimum reactivation
The stiffness moduli and Poisson’s ratio of the
temperature of 100 °C required for the tack coat R
membranes of waterproofing systems have been measured
tack coat.’
at different temperatures. The membranes were found to
behave elastically or visco-elastically, with one type stiffer
‘The rolling temperature of the surfacing must not
at lower temperatures and the other type stiffer at higher
fall below the minimum reactivation temperature of
temperatures at the strains and strain rates encountered on
85 °C required for Tack coat P, and 90 °C for Tack
bridge decks. A series of finite element analyses is being
coat SA 1030.’
undertaken to determine the significance of the properties
of the membrane on the strains induced in the asphalt ‘Temperature of the surfacing when applied should
layers overlaying waterproofing systems on bridge decks. exceed the minimum reactivation temperature of 80 °C
The findings, yet to be published, indicate the required for Britdex MDP Tack Coat.’
susceptibility to fatigue of surfacing overlaying different
types of waterproofing system and the significance of the ‘Temperature of the APL when applied should be as
stiffness properties of the waterproofing membrane. specified in BS 594-1: 1992 and BS 594-2: 1992.’

4.3.3 Effect of laying and compaction temperatures on Clause 2005.5 of the SHW (MCHW 1) states that:
the waterproofing system
‘The additional protective layer of surfacing laid on
The DMRB and SHW specify requirements that are the waterproofing system shall be full bonded to the
intended to: system for the life of the system. The bond shall be
! prevent waterproofing systems from being damaged achieved by either: (i) the binder within the directly
when they are overlaid with hot asphalt; and applied additional protective layer of surfacing; or
! ensure that asphalt remains bonded to the waterproofing (ii) a separate tack coat … details of which are given
systems over their service life. on the BBA Roads and Bridges Agrément Certificate.
Where the tack coat is of the type activated by the heat
Clauses B4.2 (i) and (j) of BD 47/99 (DMRB 2.3.4)
of the succeeding bituminous layer the temperature of
require waterproofing systems to pass tests that simulate
this layer shall be sufficient to ensure adhesion.’
the conditions when they are overlaid with hot asphalt
materials. The test specified in Clause B4.2 (j) assesses the When a layer of asphalt is laid onto a substrate such as a
effects of high temperatures encountered during surfacing waterproofing system, the base of the layer cools rapidly
on the crack bridging ability of the waterproofing as heat is transferred from the layer into the waterproofing
membrane by overlaying it with hot material to achieve a system and concrete substrate. After a short period of time,
temperature of 145 °C on its surface. Clause B4.2 (i) the temperature of the waterproofing system will have
assesses the resistance to aggregate indentation during the risen so that it is similar to that of the base of the layer.
compaction of the asphalt. Currently, all membranes must However, the waterproofing system will then cool as heat
be permanently indented by no more than half their is lost by conduction to the concrete substrate and also by
thickness after a force of 500 N has been applied by an convection and radiation through the top of the layer.
aggregate indentor heated to a temperature of 80 °C. This The asphalt layer must be laid at a sufficiently high
test is designed to simulate the compaction of sand asphalt. temperature so that the rolling/compaction temperature is

19
high enough to activate the tack coat and form a dense when the mid-layer temperature is, say, 135 °C or less.
layer. Table 4.1 and Table 4.2 show, respectively, the Clearly, the contractor should provide details of the
times after laying for 20 mm and 40 mm thick layers to temperature at the tack/bond coat above which compaction
reach a given temperature at mid-layer that were estimated should be completed or, preferably, the wording in the
using the method by Nicholls and Daines (1993). The BBA Roads and Bridges Agrément Certificates should be
temperature at the base of a layer (i.e. at the tack coat) changed accordingly.
could be about 10 °C below the mid-layer temperature. The above may appear to be unnecessarily complicated,
Therefore, the current Specification gives little time to but it is important because the bond of asphalt to
complete compaction at a sufficiently high temperature waterproofing systems tends to increase with the rolling
when the activation temperature of the tack/bond coat is temperature.
100 °C, especially when the asphalt layer is only 20 mm An aggregate indentation test could be carried out at a
thick and is laid at 145 °C. There is more time to complete temperature up to a maximum of 145 °C, the temperature
the compaction of a 40 mm layer, even if it is laid at only at which the crack bridging ability of the waterproofing
125 °C. Much more time would be available if mixtures membrane is assessed, if it is necessary to compact
containing coarse aggregates were laid at temperatures asphalt when the waterproofing membrane is at such
higher than 125 °C, which probably already happens on a temperatures. If the surfacing is likely to produce higher
number of bridges. The implication from the aggregate temperatures at the waterproofing system, it would be
indentation test is that rolling should not occur when the
necessary to carry out a crack bridging test during which
temperature at the waterproofing system is greater than
such temperatures were induced during the thermal shock
125 °C. Therefore, the Specification could be changed to
preconditioning phase.
prevent ‘rolling’ when the mid-layer temperature is 125 °C
(rather than specify a laying temperature of 125 °C).
However, because the temperature at the waterproofing 4.4 Asphalt properties
system is lower than that at mid-layer, rolling could be 4.4.1 Deformation resistance
permitted when the temperature at the waterproofing Deformation resistance of the component materials of a
system is 125 °C or lower, or rolling could be permitted
pavement is more important for those materials nearest the
surface where the loads caused by traffic are highest. The
Table 4.1 Effect of environmental conditions on time loads are distributed by the pavement layers so that they
available to compact a 20 mm thick layer reduce with the thickness of overlying pavement. The
Time after laying to
approach used to be to consider only the surfacing layers,
Wind reach given temperature with a more stringent requirement on the surface course
Air speed at Laying at mid-layer (min) than on the binder course. However, with the introduction
temperature 2 m height temperature of proprietary thin surfacings and hence thinner surface
(°C) (km/h) (°C) 120 °C 110 °C 100 °C
courses, it becomes more logical to apply the stringent
5 0 145 2.5 3.3 4.3 requirement for deformation resistance to composite
5 10 145 2.0 2.7 3.5 samples for the top 50 mm and the less stringent
5 40 145 1.7 2.2 2.9 requirement to the next 50 mm. It is proposed to take this
20 0 145 2.9 3.9 5.1 approach for bridge deck surfacings.
20 10 145 2.3 3.1 4.1
Requirements for deformation resistance should become
20 40 145 1.9 2.6 3.3
more important as the overall thickness of the surfacing
layers is reduced. Using the conventional construction
Table 4.2 Effect of environmental conditions on time approach, if in thin layers, the red sand carpet would come
available to compact a 40 mm thick layer more within the critical depth. Sand carpet is not
particularly deformation resistant, being the worst of the
Wind materials tested (Section 3.1.5). Therefore, it is proposed
Air speed Laying Time after laying to reach given to apply the current surface course requirements (as
temper- at 2 m temper- temperature at mid-layer (min)
ature height ature defined in tables NG 9/28 and NG 9/29 of the Notes for
(°C) (km/h) (°C) 120 °C 110 °C 100 °C Guidance on the Specification for Highway Works
(MCHW 2)) for deformation resistance to (50 ± 10) mm
5 0 125 3.9 6.9 10.2
thick samples of each material type that occurs in the top
5 10 125 3.2 5.7 8.3
5 40 125 2.6 4.7 6.9 50 mm. In addition, it is proposed to apply the
20 0 125 4.6 8.1 12.0 requirements at the next level down to (50 ± 10) mm thick
20 10 125 3.7 6.5 9.6 samples of materials that occur in the next 50 mm of the
20 40 125 3.0 5.3 7.9
pavement. Where materials are nominally laid at a
5 0 145 8.6 11.6 15.1
5 10 145 7.0 9.5 12.4 thickness less than 40 mm, it is proposed to permit the
5 40 145 5.8 7.8 10.2 testing of composite samples made up of more than one
20 0 145 10.1 13.6 17.8 different layers to be used in the pavement provided no
20 10 145 8.0 10.9 14.2
part layer has a thickness less than twice its maximum
20 40 145 6.6 8.9 11.6
aggregate size.

20
4.4.2 Texture depth be impermeable if surfacing on bridge decks is to remain
Texture depth is a standard requirement for the surface durable. If the lower asphalt layers on bridge decks are
course of high-speed trunk roads, and the requirements of permeable, any water that passes through the surface
clause 921 of the Specification for Highway Works course through the air voids in the layer and any defects
(MCHW 1) should be applied when appropriate for the (cracks and/or improper sealing around joints) will
specification of surfacings to bridge decks. percolate down towards the waterproofing system while
flowing across the deck towards the low points. If the
4.4.3 Skid resistance water encounters a barrier on its way, such as an expansion
joint or a less permeable area, it will accumulate in the
Although it is only referenced explicitly in clause 918
asphalt until it is drained by a sub-surface drainage system.
(slurry surfacing), 919 (surface dressing), 922 (surface
Though-deck drains will be required at all low points on
dressing), 938 (porous asphalt) and 942 (thin surface
the deck, including large hollows and depressions that are
course systems) of the Specification for Highway Works
(MCHW 1), the skid resistance of the surface course is not free draining. Furthermore, the lower asphalt layer will
defined by the polished stone value of the coarse aggregate need to remain sufficiently permeable throughout its
as laid out in Appendix 7/1 of the job specification. Advice service life to drain the sub-surface water quickly in the
on values to specify in Appendix 7/1 is given in Advice heaviest periods of rain. However, premature failures have
Note HD 36/99 (DMRB 7.5.1). This approach should be occurred on a number of bridges when the lower asphalt
used for the specification of surfacings for bridge decks. layer has been permeable, even when there has been some
sub-surface drainage (Section 4.2.3), because high
hydrostatic pressures have been generated by the wheels of
4.4.4 Flexibility and fatigue heavy goods vehicles in the surfacing when saturated.
Flexibility has been identified as an important factor over Therefore, the lower asphalt layer should be impermeable
bridge decks (Section 4.3.2). However, there are doubts and sub-surface drainage should be provided to drain the
about the current fatigue tests for asphalt because they do small amounts of water that will inevitably enter the layer
not provide a consistent ranking. In particular, the choice and permeate down to the waterproofing system.
between constant stress and constant strain can reverse the In spite of the great importance of permeability in
order. For bridge decks, it is assumed that constant strain relation to the durability performance bridges, there is
will apply because the support is so much stiffer than the currently no permeability specification to control the
asphalt. The obvious fatigue tests are the indirect tensile, quality of asphalt surfacing. This deficiency may be
four-point bending and two-point bending, the first having attributed to the controversy regarding the repeatability
had some use in the UK whilst the latter two are the and reproducibility of permeability tests using different
methods that will be available for CE marking the property techniques and the relatively long time required to produce
when the European standards are implemented. However, results. A new European Standard, EN 12697-19 (CEN,
none of these tests are currently used for routine testing of 2004c), describes a method for measuring the permeability
asphalt in the UK. of porous asphalt to water, but this method would be time-
An alternative to fatigue is flexibility. The research into consuming for partially permeable and unsuitable for
a possible test for flexibility identified the semi-circular impermeable asphalt mixtures.
bending test from several put forward. This test was found Permeability testing has received more attention from the
to be practical, equivalent to a controlled stress fatigue test concrete industry and has been widely used within the
and ranks materials similarly to binder content, a known durability specifications of major concrete constructions,
component of flexibility if other things are constant. The such as the Jubilee Line Extension Project in the UK.
test is currently being considered for European Hassan and Cabrera (1997) demonstrated that the
standardisation as a measure of crack propagation. For the permeability of concrete is dependent on many factors that
BBA HAPAS scheme for thin surfacings, it was suggested are all related to the volume of open pores, as the
that initially the values of 3.0 N/mm² and 19 N/mm3/2 permeability increases with higher volume of open porosity.
could be set as the minimum values required for tensile A concrete with a volumetric proportion of open pores of
strength and fracture toughness, respectively. These limits less than 4 % exhibited low permeability values, whereas
were considered practical because the majority of the above 5 % there is a very rapid increase in permeability.
results obtained from trial mixtures exceeded them. This change indicates that an interconnected system of pores
Whilst it would be preferable to have fatigue and/or is established when the volume of open pores is greater than
flexibility requirements that get more severe as the asphalt about 3 % or 4 %. A similar trend of permeability results is
thickness gets less, there is currently no agreement on the obtained from the asphalt mixtures, as shown in Figure 3.3.
test to be used or the limits to be achieved. Therefore, it is Regardless of the mixture type, low permeability values are
proposed not to set any fatigue requirements at this time, obtained when the air voids content is 4 % or below and the
but to consider latter inclusion if subsequent research permeability increases rapidly for air voids contents above
identifies more precisely what is appropriate. 5 %. This hypothesis is supported from results of other
investigations (Zoorob et al., 1999).
4.4.5 Permeability and air voids content The permeability test, under a pressure gradient, is an
Previous research at TRL has shown that the layer of easy test and takes a short time to be conducted. The test
asphalt directly overlaying the waterproofing system must results are found reliable in ranking the permeability of

21
various asphalt mixtures and, therefore, could be adopted that can comply. The material itself and the interface
in the durability specifications of asphalt surfacing to between the material and waterproofing system would
bridge decks. However, there is still no agreement about need to be impermeable. Also, the resistance of the
the test method, the associated limits of permeability to be waterproofing system to aggregate indentation when
specified or the precision of the results. In the absence of a overlaid with the material would need to be demonstrated.
generally accepted standard permeability test, it appears The use of red oxide in the sand carpet could be
reasonable to use the air voids content as an indicator for reproduced in the alternative mixtures if required.
permeability. However, there have been hearsay reports that often the
Any specified maximum values for air voids contents, as colour has been lost by the time the material is exposed
a surrogate for permeability, should depend on the and, with a thinner overall surfacing, the protective layer
presence and effectiveness of sub-surface drainage. If there will be reached sooner so that greater care will be needed.
is adequate drainage that can be properly maintained, a It has been found that the red sand carpet cannot be used
design air voids contents of up to 6 % could be acceptable reliably as an indicator layer to enable a bridge to be
for asphalt layers other than that directly over the resurfaced without it being re-waterproofed, which was its
waterproofing system whereas, for the bottom layer and main purpose. Therefore, the requirement to use red oxide
where there is no adequate sub-surface drainage, an air should become optional.
voids contents of not more than 4 % should be required.
Both requirements can be met by all of the materials tested 4.5 Joints and interface between layers
(Section 3.1.2) other than the open SMAs and the 35 %
0/10 HRA (despite 35 % 0/14 HRA having been used 4.5.1 General
successfully on bridges in the past), the latter being the The use of impermeable asphalt layers will be inadequate
only one to fall between the two proposed limits. If those without ensuring proper sealing at any details, interfaces or
limits are used for the mean, a tolerance of 2 % should be joints between the rips, and any variations in permeability
included to give maxima on individual readings of 8 % and associated with defects such as cracks. The relevant details
6 %, respectively. It would be ideal to use the individual are kerbs, parapets and expansion joints. The relevant
maxima on site, but it is appreciated that it is best not to interfaces are between the asphalt layers and between the
take cores when impermeability is sought. lowest asphalt layer and the waterproofing system.

4.4.6 Protection for waterproofing systems 4.5.2 Joints


The current Specification requires waterproofing systems The joints between adjacent rips of asphalt should be
to be overlaid with a 20 mm thick additional protection sealed effectively in order to prevent an easy path for
layer of sand asphalt (0/2 HRA). Because the layer has low water to flow vertically through an asphalt layer, as set out
resistance to deformation, it performs satisfactorily only if in both BS 598-2 and BS 4987-2. However, joints that are
it is overlaid with sufficient thickness of deformation fully sealed will prevent the flow of water horizontally
resistant surfacing. Therefore, it should be omitted when across a bridge so provision must be made so there is no
the overlaying surfacing is less than 100 mm thick. danger that water can accumulate in a layer, which could
However, the benefits of laying sand asphalt, even when lead to high hydrostatic pressures being generated by the
it is overlaid with 100 mm or more of surfacing, have been wheels of heavy goods vehicles. However, given the
questioned by many in the industry. The original intention intention to ensure that the materials used are relatively
was to lay the sand asphalt immediately after impermeable (Section 4.4.5), it is proposed to require all
waterproofing in order to protect the waterproofing system cold vertical joints to be painted with bitumen before being
before and during surfacing operations. However, a laid against.
number of waterproofing systems have passed the
aggregate indentation test at 125 ºC (see Section 4.3.3) and 4.5.3 Bond between layers
do not require the protection of sand asphalt provided due The bond between layers has become more important with
care is taken before and during surfacing operations. the use of thin surface course layers because the absence of
Appendix B to BD 47/99 (DMRB 7.2.3) requires that all the weight supplied by a thick layer needs to be replaced by
permitted waterproofing systems are tested for aggregate positive adhesion with the layer below in order to minimise
indentation at 40 °C and 80 °C with a compliance the potential for de-bonding. However, there is no generally
requirement that the indentation after a recovery period accepted method for measuring the property in-situ on a
shall not exceed 50 % of the initial system. Waterproofing routine basis. There are two aspects to bond, the adhesion
systems may be overlaid with asphalt containing coarse achieved by the binder present and the aggregate interlock
aggregates of maximum aggregate size greater than that created by texture of the underlying material being filled by
found in sand asphalt if they have passed an aggregate particles from the overlay. Different test methods that have
indentation test at 125 °C. Not all of the registered systems been proposed to measure bond are affected differently by
have passed the test at 125 °C, which could disadvantage the two aspects, with pull-off tests (as used for bridge water-
them in the marketplace, particularly when a specification proofing and high-friction surfacing system) ignoring
requires that the asphalt directly overlaying the aggregate interlock whilst shear tests (as used for thin
waterproofing system should contain large sized coarse surfacing systems) incorporate that aspect. Provided they are
aggregate, reducing the number of waterproofing systems not saturated, the adhesion generally improves with

22
trafficking because the two layers are pressed together with aggregates directly overlays the waterproofing system.
each wheel-pass, so the timing of any measurements of the Also, higher limits are appropriate for thicknesses less than
property is also important. 120 mm, and the proposal is to increase them in steps of
Bond, based on a torque shear test, is a required 30 mm down to 60 mm. Any total depth of surfacing less
property for thin surfacing systems in order to obtain a than 60 mm is regarded as a special case that will require
Highway Authorities Product Approval Scheme (HAPAS) expert advice. The resultant limits are given in Table 4.3.
certificate from the British Board of Agrément (BBA).
However, there is no pass/fail rating under the current Table 4.3 Bonding limits for waterproofing systems
guidelines (BBA, 2000), only the need to report the result. when overlaid by coarse mixtures
The test, which is carried out after between 28 and 56 days
of trafficking, is intended as a type test because of the Surfacing thickness ≥120 mm <120 mm; ≥90 mm < 90 mm; ≥60 mm
disruption caused by closing the road to obtain the samples Tensile adhesion test
so soon after being re-opened. However, the results @ -10 °C 0.30 N/mm² 0.50 N/mm² 0.70 N/mm²
recorded to date typically vary between 400 and 1500 kPa, @ 23 °C 0.30 N/mm² 0.50 N/mm² 0.70 N/mm²
so that a specification requirement for 400 kPa would @ 40 °C 0.20 N/mm² 0.30 N/mm² 0.30 N/mm²
allow all currently certified thin surfacing systems to be
Shear adhesion test
used whereas a limit of, say, 700 kPa would exclude some @ -10 °C 0.30 N/mm² 0.30 N/mm² 0.40 N/mm²
of those that have performed least well in the test. If a limit @ 23 °C 0.30 N/mm² 0.30 N/mm² 0.40 N/mm²
is set, any non-proprietary surfacing would also need to be @ 40 °C 0.10 N/mm² 0.15 N/mm² 0.15 N/mm²
tested before it was permitted for use on bridge deck
Tensile bond test
overlays. The limits proposed will be dependent on the @ 23 °C 0.40 N/mm² 0.45 N/mm² 0.50 N/mm²
depth of the interface with 700 kPa if within 20 mm of the
surface, reducing to 400 kPa at 50 mm and no requirement
below that. The tests should be performed with the
relevant tack or bond coat to be used on site with the 4.5.5 Permeability of interface between asphalt and
appropriate materials as substrate and overlay. waterproofing system
Given the variable approach, it is proposed here to raise Most waterproofing systems comprise a primer to optimise
the general requirement, of either a tack or bond coat the bond of the system to the concrete substrate, a
between each layer, to a requirement for bond coats at all membrane to prevent water and chlorides from reaching
interfaces. This approach does not affect the current bond the concrete, and a tack/bond coat to optimise the bond of
requirements for waterproofing systems or thin surfacing the overlaying asphalt to the system.
systems in order to obtain their BBA-HAPAS certificates. When asphalt with coarse aggregates is compacted onto a
‘hard’ surface, the bulk of the material may have a low void
4.5.4 Bond to waterproofing system content but there may be large voids at the base of the layer.
According to clause 2003.2 of the Specification for The voids tend to be larger with larger aggregate sizes and
Highway Works (MCHW 1), a waterproofing system with higher proportions of coarse aggregate. Above a certain
needs to have a BBA Roads and Bridges certificate in proportion of coarse aggregate, the voids may be
order to be permitted for use on trunk road bridges. Part of interconnecting. When there are voids at the interface
the laboratory test procedure to gain certification is to between the lowest asphalt layer and the waterproofing
check the bond between the waterproofing system and the system, water can accumulate and there is a risk of premature
underlying concrete substrate by means of a tensile failure. Therefore, the void content at the interface should be
adhesion (pull-off) test and to check the bond between low and voids should not be interconnecting.
waterproofing system and the overlying asphalt by means The permeability of the interface between the
of shear-adhesion and tensile bond tests. These tests are waterproofing system and the asphalt is dependent on the
described in Appendix B to BD 47/99 (DMRB 7.2.3). properties of both the asphalt and the tack/bond coat.
There are limits for the tests, with minima of: Whereas a tack/bond coat for a waterproofing system may
! 0.3 N/mm² at -10 °C, 0.3 N/mm² at 23 °C and 0.2 N/mm² comprise more than one layer, the upper layer that is in
at 40 °C for the tensile adhesion test on the contact with the asphalt and which aggregates can
waterproofing system to concrete substrate interface; penetrate can be described as ‘thin’ (generally <0.2 mm) or
‘thick’ (generally >1.0 mm). When the tack/bond coat is
! 0.2 N/mm² at -10 °C, 0.2 N/mm² at 23 °C and 0.1 N/mm²
‘thin’, an asphalt layer with a small aggregate size and low
at 40 °C for the shear adhesion test on the surfacing to
proportion of coarse aggregate will yield no large
waterproofing system interface; and
interconnecting voids. However, any mixture that contains
! 0.1 N/mm² at 23 °C for the tensile bond test on the large aggregates will result in some voids where water may
surfacing to waterproofing system interface. accumulate and a reduction in contact area between the
These limits apply to overlaying asphalt of minimum mixture and the waterproofing system. As discussed
thickness 120 mm where the asphalt directly overlaying below, a reduction in the contact area may adversely affect
the waterproofing system is 0 % 0/2 mm HRA (sand the tensile bond of the asphalt to the waterproofing system.
asphalt) or 50 % 0/10 HRA. However, higher limits are A ‘thick’ tack/bond coat can (partially) fill the voids at
considered necessary when mixtures containing coarse the base of an asphalt layer with coarse aggregates and,

23
thereby, limit the accumulation of water and will not usually have equal ranking. Material optimisation
interconnecting voids and, potentially, improve the tensile is usually avoiding any excessively adverse property rather
adhesion. Therefore, a waterproofing system incorporating than getting the best performance in all.
a thick tack/bond coat should be specified for mixtures The initial results from the torque bond test varied by an
with coarse aggregates. The tack/bond coat must not be too order of magnitude, with the two waterproofing systems
thick otherwise ‘bleeding’ of the excess binder through the ranking the surfacing materials differently. In all cases, the
overlying asphalt layer may occur during its laying and failures were at the interface between the waterproofing
compaction. Also, the asphalt layer will be more membrane and the tack/bond coat, so the failure stresses
susceptible to fatigue if it ‘floats’ on a thick layer of ‘soft’ were more dependent on the properties of the
tack/bond coat. Ideally, the coarse aggregates should waterproofing systems than the asphalt mixtures. For
almost fully penetrate the tack/bond coat as the tack/bond Membrane B systems, the highest failure stresses were
coat material fills the voids at the base of the layer. measured with tack coat R and sand carpet. For Membrane
Therefore, the specification needs to ensure that the A systems, the failure stresses were higher for tack coat P
lower asphalt layer overlaying the waterproofing system is than for tack coat Q. The high values for mastic asphalt
impermeable. The mixture design and type of tack/bond were probably due to the high temperature at which the
coat should ensure that the air voids contents at the material was laid and compacted and at which the tack/
interface between the waterproofing system and the base bond coat was activated.
of the layer is low without interconnecting voids (see Based on these findings and other considerations,
Section 3.3.3). To minimise the amount of water that can various additions and changes to the Design Manual for
enter the lower asphalt layers, the full depth of permeable Roads and Bridges, Specification for Highway Works and
surface courses should be drained at the edges. If this is Notes for Guidance on the Specification for Highway
not possible, an impermeable surface course should be laid Works have been proposed. The main changes include:
that directs water to a suitable surface drainage system.
! sub-surface drainage is emphasised;
These requirements are covered in Section 4.4.5. However,
frequent changes in surface texture and road noise ! bond requirements are strengthened;
associated with a change in surface course materials from ! deformation requirements are specified for all mixtures
bridge to pavement should be avoided. within 100 mm of the surface; and
The total thickness of surfacings on bridges may range ! maximum air voids content limits on all asphalt mixtures.
from 40 mm to over 120 mm. Such surfacings may require
one, two or three asphalt layers, each with a different Aspects that were not fully covered are permeability
material. On most bridges, there are small variations in the testing of the asphalt and at the interfaces. Potential tests
level of the deck that can be accommodated without have been identified that could be developed for
varying the number of asphalt layers. However, it has been standardisation if these aspects are considered critical. The
necessary on a few bridges to vary the number of layers, asphalt permeability can be covered by the surrogate of air
requiring some degree of compromise in the regions where voids content, but this is more difficult at interfaces where
the number changes. Under these circumstances, the lower more than one material is involved.
layer directly overlaying the waterproofing system should For flexibility, the test selected from those put forward
be of reasonably uniform thickness and of the same is the semi-circular bending test. This test was found to be
impermeable material. Changes in thickness should be practical, equivalent to a controlled stress fatigue test and
accommodated in the upper layers, with tapered layers ranks materials similarly to binder content, a known
trimmed so material too thin to have been compacted component of flexibility if other things are constant.
sufficiently is removed. It is suggested that initially the values of 3.0 N/mm² and
19 N/mm3/2 should be set as the minimum values required
for tensile strength and fracture toughness, respectively.
5 Conclusions These limits are considered practical because the majority
of the results obtained exceeded them.
From the tests on a series of twelve asphalt mixtures,
mastic asphalt was found to have the most suitable air
voids content, permeability and stiffness modulus 6 Acknowledgements
properties and was fifth at wheel-tracking, making it the
best material overall for these tests. However, mastic The work described in this report was carried out in the
asphalt is a relatively expensive mixture, a factor that Infrastructure and Environment Division of TRL Limited.
cannot be excluded. The dense 0/10 SMA was the next The authors are grateful to Kevin Green, Mario Patchett
best despite having the 8th highest air voids content whilst and Jon Harper, who carried out all the laboratory testing
the open 0/10 SMA was the worst, showing that the that underpin this work, and to Val Atkinson, who carried
precise mixture design can be critical. The remaining out the quality review and auditing of this report.
mixtures showed relatively similar overall ratings, but with Particular thanks are given to the two manufacturers who
the sand carpet only being ranked 8th out of 12 overall. applied their waterproofing systems to concrete samples
Nevertheless, when considering the appropriate materials, and to the four positive responders to the questionnaire.
the choice is often a trade off between properties and they

24
7 References Comité Européen de Normalisation (2003b). Bituminous
mixtures – Test methods for hot mix asphalt – Part 8:
British Board of Agrément (2000). Guidelines document Determination of void characteristics of bituminous
for the assessment and certification of thin surfacing specimens. BS EN 12697-8: 2003. London: British
systems for highways. BBA-HAPAS SG3/98/169, Working Standards Institution.
Draft 3. Watfird: British Board of Agrément.
Comité Européen de Normalisation (2003c). Bituminous
British Standards Institution (1988). Specification for mixtures – Test methods for hot mix asphalt – Part 22:
mastic asphalt (limestone fine aggregate) for roads, Wheel tracking. BS EN 12697-22: 2003. London: British
footways and paving in building. BS 1447: 1988. London: Standards Institution.
British Standards Institution.
Comité Européen de Normalisation (2004a). Bituminous
British Standards Institution (1993). Methods for mixtures – Test methods for hot mix asphalt – Part 35:
determination of the indirect tensile stiffness modulus of Laboratory mixing. BS EN 12697-35: 2004. London:
bituminous mixtures. British Standard Draft for British Standards Institution.
Development DD 213: 1993. London: British Standards
Institution. Comité Européen de Normalisation (2004b). Bituminous
mixtures – Test methods for hot mix asphalt – Part 24:
British Standards Institution (1996). Testing concrete – Resistance to fatigue. BS EN 12697-24: 2004 British
Recommendations for the determination of the initial Standards Institution, London.
surface absorption of concrete. BS 1881-208: 1996.
London: British Standards Institution. Comité Européen de Normalisation (2004c). Bituminous
mixtures – Test methods. Part 19, Permeability of porous
British Standards Institution (2001). Coated macadam asphalt specimen. BS EN 12697-19: 2004. London: British
(asphalt concrete) for roads and other paved areas – Standards Institution.
Part 1: Specification for constituent materials and asphalt
mixtures. BS 4987-1: 2001. London: British Standards Daines M E (1994). Tests for voids and compaction in
Institution. rolled asphalt surfacings. Project Report PR78.
Wokingham: TRL.
British Standards Institution (2002). Hot rolled asphalt
for roads and other paved areas – Part 1: Specification for Grube H and C D Lawrence (1984). Permeability of
constituent materials and asphalt mixtures. BS 594-1: concrete to oxygen. Proceedings of the RILEM Seminar on
2002. London: British Standards Institution. ‘Durability of concrete structures under normal outdoor
exposure’. Hanover, pp. 68-79.
Comité Européen de Normalisation (2000a). Bituminous
mixtures – Material specification. Part 5, Stone mastic Hassan K E and J G Cabrera (1997). Controlling the
asphalt. Draft BS EN 13108-5, DPC No. 00/100954DC. quality of concrete by measuring its permeability.
London: British Standards Institution. Proceedings 13th International Conference on ‘Building
Materials’, Volume 3. Weimar-Germany, pp. 3/005-3/0017.
Comité Européen de Normalisation (2000b). Bituminous
mixtures – Test methods for hot mix asphalt – Part 27: The Highways Agency, Scottish Development
Sampling. BS EN 12697-27: 2000. London: British Department, The National Assembly for Wales and The
Standards Institution. Department for Regional Development Northern
Ireland. Manual of Contract Documents for Highway
Comité Européen de Normalisation (2002a). Bituminous Works. London: The Stationery Office:
mixtures – Test methods for hot mix asphalt – Part 7: Volume 1: Specification for Highway Works (MCHW 1).
Determination of bulk density of bituminous specimens by
Volume 2: Notes for Guidance on the Specification for
gamma rays. BS EN 12697-7: 2002. London: British
Highway Works (MCHW 2).
Standards Institution.
Highways Agency, Scottish Executive Development
Comité Européen de Normalisation (2002b). Bituminous
Department, National Assembly for Wales and
mixtures – Test methods for hot mix asphalt – Part 5:
Department for Regional Development, Northern
Determination of the maximum density. BS EN 12697-5:
Ireland. Design Manual for Roads and Bridges. London:
2002. London: British Standards Institution.
The Stationery Office:
Comité Européen de Normalisation (2003a). Bituminous BD 47/99: Waterproofing and surfacing of concrete
mixtures – Test methods for hot mix asphalt – Part 33: bridge decks (DMRB 2.3.4)
Specimen prepared by roller compactor. BS EN 12697-33: HD 36/99: Surfacing materials for new and maintenance
2003. London: British Standards Institution. construction (DMRB 7.5.1)

25
Nicholls J C and Carswell I (2004). Durability of thin
asphalt surfacing systems: Part 2: Findings after 3 years.
TRL Report TRL606. Wokingham: TRL.

Nicholls J C and Daines M E (1993). Acceptable weather


conditions for laying bituminous materials. Project Report
PR13. Wokingham: TRL.

Peoples Republic of China (1993). Complete ultimate


strain at a low temperature –10 ºC: > 6 x 10-3; Asphalt
Bending Test. PRC Requirement T0715-93.

Zoorob S E, Cabrera J G and Suparma L B (1999). A


gas permeability method for controlling quality of dense
bituminous composites. Proceedings of the 3rd European
Symposium on Performance and durability of bituminous
materials and hydraulic stabilised composites. pp. 549-572.
Leeds.

26
Appendix A: HA draft notes for bridge-deck overlays

A.1 Sub-surface drainage waterproofing. However they are resilient and although
All blacktop leaks to some extent, through joints etc – and they are designed to be applied in very thin layers,
the use of macadam bases with void contents that may be concrete bridge decks to which they are applied are often
(at worst) up to 7 or 8 % and more porous quiet surfacings quite rough. This can result in thicker areas of membrane,
may well exacerbate this. Advice is given in DMRB Vol 2 increasing the resilience. Premature failures have occurred
Section 3 Part 4 BD 47/99 Chapter 4 ‘Drainage’. In with both hot rolled asphalt and stone mastic asphalt due at
paragraph 4.2 on sub-surface drainage, to quote: least in part to fatigue of a sub-standard thickness overlay.
Such failures are always associated with debonding of the
‘Bituminous surfacing is porous and can retain surface
overlay and it has not been possible to determine the
water. Where the geometry of the deck or deck movement
joints prevents this water from draining naturally through primary cause of these failures – poor bond, water
surface drainage, sub-surface drains shall be provided. saturation and traffic generated pressure, or fatigue. It is
Advice on sub-surface water drainage is given in BA 47 likely that each plays some part.
(DMRB 2.3.5)’
A.5 Actions to be considered before applying for a
The term surfacing in this instance is referring to the
bituminous overlay above the waterproofing not just the departure from standards to use a sub-standard
surfacing course. bituminous overlay less than 120 mm in thickness
Further information is given in TRL Application Guide 33 i Re-assessment of the structure to maximise the
‘Water Management for Durable Bridges’ in the drainage thickness of the bituminous overlay.
section. Other useful references are BD 33/94 and BA 26/94
ii Specification of the highest modulus approved
both concerning expansion joints and covering their drainage.
proprietary waterproofing system obtainable. (note: the
Bridge-deck overlays (the surface course and base layers)
need for this will diminish with increasing thickness of
should always be regarded as porous and drainage should be
provided below the overlay at low points over the bridge overlay and where the thickness of the membrane can
deck waterproofing. Edge drainage should be provided at be kept to a minimum.)
joints – where compaction is likely to be least efficient – and iii Omission of the sand asphalt layer above the
sub-surface drainage installed where interstitial ponding waterproofing system.
may occur. There have been one or two instances of fretting iv Specification of a waterproofing system which includes
in thin surfacings at such locations, caused by poor the provision of a bond coat for the asphalt overlay.
compaction. It is important to note that this may not be on Where the waterproofing system offers alternative
the bridge deck side of the joint, depending on the gradient. bond promoting treatments, the specification of the
treatment claimed to provide the maximum bond shall
A.2 Bond to the waterproof membrane be specified. Any proprietary bond coat between the
Another issue is the bond of the overlay to the waterproofing system and the overlay shall be a ‘tack-
waterproofing membrane. This tends to be relatively low free’ material, such that it does not adhere to tyres of
with the modern waterproofing systems now in use and vehicles delivering asphalt to the paver.
may be disrupted by pressures generated under traffic if v Provision of sub-surface, edge and joint drainage as
the overlay becomes saturated. There have been instances appropriate, to reduce or eliminate water pressure
where this appears to have occurred; in particular where under traffic.
the blacktop overlay, usually in the past surfaced with
HRA, is less than the standard 120 mm in thickness. vi Specification of a paver-laid hot rolled asphalt layer or
layers, containing an elastomeric polymer and an
appropriate aggregate size, to form the overlay binder
A.3 The thin surface course
course, all in accordance with SHW Clause 943, (but
It is government policy to use quiet surfacings on all trunk omitting any coring over the bridge-deck!).
roads in England, including motorways. The capacity of
vii Specification of a thin surface course system which
these proprietary systems to waterproof the base layers
below often appears to depend more on the bond or tack incorporates a heavy elastomeric-polymer modified
coat applied than the apparent ‘porosity’ of the surface bond coat.
course. Anecdotal evidence suggests that thin surfacing The risk of premature failure of sub-standard thickness
systems with an open texture laid on a heavy polymer bituminous overlays on bridge decks is considerable. For
modified bond coat can be more effective at sealing and further advice or clarification, please contact James
waterproofing the base layers than thicker, less open Gallagher or John Williams of SSR RDS Highways
systems (SMAs) laid on a thin tack coat. Infrastructure Group.
James Gallagher: Tel: 0161 930 5527 (GTN 4315 5733)
A.4 Sub-standard overlay thickness
A bridge-deck overlay is more vulnerable the thinner it is. John Williams: Tel: 01234 796116 (GTN 3013 6116)
Modern waterproofing systems are very effective at or 01438 718487 (GTN 6492 0008)

27
Appendix B: Questionnaire

28
29
30
31
Appendix C: Permeability tests

C.1 TRL permeability test The intrinsic oxygen permeability can be determined
The TRL permeability cell (Figure C.1) is similar to that from the measurements of flow rate according to the
developed by the Cement and Concrete Association modified D’Arcy’s equation, as given in Equation (C.1).
(C&CA, since renamed the British Cement Association,
BCA) (Grube and Lawrence, 1984) and Leeds University 2ν η L P2
K=
(Hassan and Cabrera, 1997) for concrete under differential (
A P12 − P22 ) ….. (C.1)
pressure techniques. Measurement of air permeability was
conducted by sealing the curved surface of 100 mm where: K = intrinsic air permeability (m²)
diameter cores and applying a desired pressure on one side ν = flow rate (m³/sec)
of the specimen. The pressure gradient across the specimen
η = viscosity of air (1.82 × 10-5 N.s/m² at 20 °C)
results in a flow, which is measured at the other side using
a flowmeter. The test duration is quite short, less than L = length of the specimen (m)
30 min, depending on the air voids content and the A = cross-sectional area of the specimen (m²)
continuity of voids of the tested asphalt. P1 = inlet absolute applied (gauge) pressure (bar)
P2 = outlet pressure at which the flow rate is
measured (bar), usually 1 bar

C.2 Interface permeability test


The water permeability at the interface between the asphalt
surfacing and the waterproofing system was measured
using the apparatus described in BS 1881-208: 1996 for
measuring the initial surface absorption of concrete. The
principle of the test is to determine the time taken for a
quantity of water to flow through a calibrated glass tube
onto a given area of the test specimens.
The composite slabs (concrete, waterproofing and
asphalt) were sliced vertically down the centre to provide
cut faces specimens with the waterproofing system along
Figure C.1 The TRL permeability cell the centre. The test cap was sealed into the sliced
specimens using silicon rubber and a clamp mechanism, as
shown in Figure C.2. Water was then introduced through
the reservoir, funnel, to fill the test cap. The amount of
water leaking through the sliced specimens was recorded
periodically to determine the flow rate per unit area of the
test specimen.

Figure C.2 The initial surface absorption test for measuring permeability at the interface

32
Appendix D: Semi-circular bending test

D.1 Scope D.3.7


This annex describes the semi-circular bending (SCB) test Symbols
method to determine the tensile strength or fracture D Diameter in millimetres (mm).
toughness of an asphalt mixture. The results of the test can
W Height in millimetres (mm).
be used to calculate:
! The maximum load that the material containing a notch t Thickness in millimetres (mm).
(crack) can resist before failure. a Notch depth in millimetres (mm).
! When the presence of a notch is critical. F Force in newtons (N).
Fmax Maximum force in newtons (N).
D.2 Normative references ∆W Vertical displacement in millimetres (mm).
The following referenced documents are indispensable for Ki c Fracture toughness in newtons per metre to the
the application of this annex. For dated references, only
power of 1.5 (N/m1.5).
the edition cited applies. For undated references, the latest
edition of the referenced document (including any ε Strain.
amendments) applies. σ Stress in newtons per metre squared (N/mm²).
BS EN 12697-27, Bituminous mixtures – Test methods – σhor Horizontal stress in newtons per metre squared
Part 27: Sampling. (N/mm²).
BS EN 12697-33, Bituminous mixtures – Test methods – σmax Maximum stress at failure in newtons per metre
Part 33: Specimen preparation by slab compactor. squared (N/mm²).
BS EN 12697-35, Bituminous mixtures – Test methods – f(a/W) Geometric factor.
Part 35: Laboratory mixing.

D.4 Principle
D.3 Terms and definitions
A half-cylinder test piece is loaded in three-point bending in
For the purposes of this annex, the following terms and
such a way that the middle of the base of the test piece is
definitions apply.
subjected to a tensile stress. During the test, the deformation
increases at a constant rate of 0.085 mm/s. The corresponding
D.3.1
load increases to a maximum value, Fmax, that is directly
SCB test piece
related to the fracture toughness of the test sample.
sample obtained by sawing an asphalt cylinder through a
diameter
D.5 Apparatus
D.3.2 D.5.1 Universal test machine
strain The machine should have a range of at least 50 kN and
relative deformation of the SCB test piece have a drive with which a constant deformation rate of
(5 ± 0.5) mm/min is maintained during the test.
D.3.3
NOTE: Contact with moving parts should be avoided
stress
when the machine is in operation. In case of emergency,
the force per unit area
the machine should be capable of being stopped by
pressing an emergency button.
D.3.4
horizontal stress
D.5.2 Load cell or other force-measuring gauge
the tensile stress prevailing at the base of the SCB test piece
Capable of measuring loads up to 50 kN with an accuracy
of ±0.2 kN.
D.3.5
tensile strength
the maximum stress occurring during the test on the base D.5.3 Roller bearings
of the SCB test piece With a support length (centre to centre) of 120 mm and a
diameter of 35 mm, as shown in Figure D.1.
D.3.6
fracture toughness D.5.4 Metal loading strip
resistance to failure by breaking With a width of (10 ± 0.2) mm, as shown in Figure D.1.

33
Loading strip D.6.1.2
Prepare asphalt slabs by either:
! Manufacturing sufficient asphalt in the laboratory in
accordance with BS EN 12697-35 and compacting the
asphalt into slabs 50 mm thick by roller compactor in
accordance with BS EN 12697-33.
! Manufacturing sufficient asphalt from the asphalt plant
and compacting the asphalt into slabs 50 mm thick by
roller compactor in accordance with BS EN 12697-33.
Cut 150 mm diameter cores from the slabs in accordance
Roller
bearings with BS EN 12697-27.

120mm D.6.1.3
Cut 150 mm diameter cores from site-compacted asphalt in
Figure D.1 Schematic of semi-circular bend test accordance with BS EN 12697-27. Ensure that the top and
bottom surfaces of the SCB specimen are flat and parallel
by slicing if required.
D.5.5 Monitoring equipment
Capable of continuously logging the loading and the D.6.1.4
vertical deformation of the SCB test piece. Cut each core into two equal semi-circular SCB specimens
through the middle (Figure D.2).
D.5.6 Sliding callipers
With a reading accuracy of ±0.1 mm D.6.1.5
If required for determining the fracture toughness, cut a
notch nominally to a width of (5 ± 0.3) mm and a depth of
D.5.7 Climatic chamber
10 mm (Figure D.2).
With a temperature range of 0 °C to 50 °C and an accuracy
of ±2 °C. NOTE: Depending on the parameters to be
determined, either a notched cylinder (for fracture
toughness) or an un-notched cylinder (for pulling
D.6 Sample preparation
force) may be used.
D.6.1 Manufacture
D.6.1.1
D.6.2 Dimensional check
Prepare asphalt cylinders in accordance with either D.6.1.2
or D.6.1.3. D.6.2.1
Ensure that the SCB specimen is dry.
NOTE: If the cylinders are cored, it is important that
they are drilled as near perpendicular to the surface
D.6.2.2
of the asphalt as practicable.
Measure the diameter of the SCB specimen with the
callipers at two places along the longest side of the piece.
Record the average diameter, D, to an accuracy of 0.1 mm.
h2

h1
W

Notch
a

½D t

D
D = 150mm t = 50mm
W = 75mm a = 10mm

Figure D.2 SCB specimen dimensions (with the nominal dimensions for a standard test)

34
D.6.2.3 D.8 Calculations
Measure the height of the SCB specimen with the callipers D.8.1
on each side, h1 and h2. Discard any SCB specimen for
Determine the maximum load, Fmax, and the vertical
which (h1 – h2) > 0.5 mm. Record the average height, W,
deformation, ∆W, from the plot.
to an accuracy of 0.1 mm.
D.8.2
D.6.2.4 Calculate the strain, ε in accordance with Equation D.1.
Measure the thickness of the SCB specimen with the
callipers at the two ends of the base and once at the top. ∆W
ε= × 100 % …. (D.1)
Record the average thickness, t, to an accuracy of 0.1 mm. W
where: W = Height in millimetres (mm)
D.6.2.5
∆W = Vertical displacement in millimetres (mm)
If cut, measure the depth of the notch, a, to an accuracy of
0.1 mm.
D.8.3
Calculate the horizontal stress, σhor, in accordance with
D.6.3 Storage Equation D.2.
SCB specimens that are not to be tested directly shall be
stored on a flat surface in a cool area at (7 ± 3) °C. 4.263 × F
σ hor = N/mm 2 …. (D.2)
D×t
D.7 Procedure where: D = Diameter in millimetres (mm)
t = Thickness in millimetres (mm)
D.7.1 F = Force in newtons (N)
Place the SCB specimen in the climatic chamber at the test
temperature of 15 ºC for at least 4 h.
D.8.4
Calculate the maximum stress at failure, σmax, in accordance
D.7.2 with Equation D.3.
Position the load cell and loading strip on the upper beam
and the roller bearings on the lower beam of the universal 4.263 × Fmax
σ max = N/mm 2 …. (D.3)
test machine. D×t
where: D = Diameter in millimetres (mm)
D.7.3
t = Thickness in millimetres (mm)
Remove the SCB specimen from the climatic chamber and
install in the universal test machine between the loading Fmax = Maximum force in newtons (N)
strip and the roller bearings (Figure D.1) in as short a time
as practicable. The overhangs of the SCB specimen from D.8.5
the roller bearings shall be equal to ±2 mm. For un-notched SCB specimens, the tensile strength is
equal to the maximum stress.
D.7.4
D.8.6
As soon as the specimen is in place, raise the lower beam
For notched SCB specimens, calculate the fracture toughness,
of the universal test machine until the SCB specimen just
Kic, of the material in accordance with Equation D.4.
touches the top load strip. Set the vertical deformation at
zero and then apply a load to the specimen sufficient to ⎛a⎞
produce a deformation rate of 0.085 mm/s = 5.1 mm/min. Kic = σ max π × a × f ⎜ ⎟ N/mm3/2 …. (D.4)
Record the force and the vertical displacement at intervals ⎝W ⎠
of not more than 5 s until the specimen fails. where: W = Height in millimetres (mm)
NOTE: The time between the removal of the SCB a = Notch depth in millimetres (mm)
sample from the climatic chamber and the sample σmax = Stress at failure in newtons per
failing should be as short as possible (preferably less millimetre squared (N/mm²)
than 60 s). f(a/W) = Geometric factor in accordance with
Equation D.5

2
D.7.5 ⎛a⎞ ⎛a⎞ ⎛a⎞
f ⎜ ⎟ = −0.623 + 29.29 × ⎜ ⎟ − 171.2 × ⎜ ⎟
Plot both the force, F, and the vertical displacement ⎝W ⎠ ⎝W ⎠ ⎝W ⎠
against time from the start of the test. …. (D.5)
3 4 5
⎛a⎞ ⎛a⎞ ⎛a⎞
+457.1× ⎜ ⎟ − 561.2 × ⎜ ⎟ + 265.54 × ⎜ ⎟
⎝ ⎠
W ⎝ ⎠
W ⎝W ⎠

35
D.9 Test report
The test report shall include the following information:
a Identification number, type of mixture and
composition.
b Date and time of testing and name of operator.
c Reference to this document.
d Sample code.
e Date sample was taken.
f Test temperature.
g Average sample diameter, D, to an accuracy of 0.1 mm.
h Average sample height, W, to an accuracy of 0.1 mm.
i Average sample thickness, t, to an accuracy of 0.1 mm.
j Notch depth, a, to an accuracy of 0.1 mm.
k Maximum force, Fmax, to an accuracy of 1 N.
l Maximum stress at failure, σmax, to an accuracy of
0.1 N/mm².
m Fracture toughness, Kic, to an accuracy of 0.1 N/mm².
n Any observation which may have an influence on the
evaluation.

D.10 Precision data


No precision data is currently available for this method.

36
Appendix E: Asphalt bending test

E.1 Scope E.4 Principle


E.1.1 The test is conducted on a standard beam by exerting
This annex describes a method to determine the concentrated load on the midspan point of the specimen
mechanical properties of hot mixed asphalt when until failure occurs. The bending strength of the test piece
bending failure occurs at a specified temperature and in megapascals (MPa) and the bending strain of the asphalt
loading rate. Unless stated otherwise, the recommended at failure can be worked out are calculated from the
test temperature is (15.0 ± 0.5) ºC and loading rate is maximum load at failure and the midspan deflection,
(50 ± 1) mm/min. For the evaluation of the low respectively. The ratio of the bending strength to the
temperature tensile properties of the asphalt, the bending strain is the bending stiffness modulus at failure in
recommended test temperature is (–10.0 ± 0.5) ºC and megapascals (MPa).
loading rate is (1.0 ± 0.1) mm/min.
E.5 Equipment
E.1.2 E.5.1 Universal test machine
This annex applies to prismatic beams cut from slabs The load shall be measured by a transducer and the
compacted by roller-compactor. The dimensions of the measuring range shall be at not less than 125 % and not
specimen are (250 ± 2.0) mm in length, (30 ± 2.0) mm in greater than five times the maximum load. The machine
width and (35 ± 2.0) mm in height. The span shall be shall have a beam pedestal and the centre-to-centre spacing
(200 ± 2.0) mm. of the lower pedestal shall be 200 mm. The upper pressure
head shall be located in the middle and the upper pressure
E.2 Normative references head and pedestal shall be fixed steel cylindrical rods with
The following referenced documents are indispensable for a radius of (10 ± 1) mm. The upper pressure head shall be
the application of this annex. For dated references, only movable and capable of coming into close contact with the
the edition cited applies. For undated references, the latest test specimen. The strain rate shall be adjustable.
edition of the referenced document (including any Note 1: A loading range of 1 kN or 5 kN is
amendments) applies. desirable with divisions at 10 N.
BS EN 12697-5, Bituminous mixtures – Test methods –
Part 5: Maximum density. Note 2: It is desirable for the testing machine to be
servo system.
BS EN 12697-6, Bituminous mixtures – Test methods –
Part 6: Bulk density, measurement.
BS EN 12697-8, Bituminous mixtures – Test methods – E.5.2 Ambient temperature cabinet
Part 8: Air voids content. Capable of controlling the temperature to an accuracy of
BS EN 12697-27, Bituminous mixtures – Test methods – ± 0.5 ºC (optional).
Part 27: Sampling.
BS EN 12697-33, Bituminous mixtures – Test methods – E.5.3 Midspan displacement measuring device
Part 33: Specimen preparation by slab compactor.
Note: An LVDT, electronic dial gauge or similar
BS EN 12697-35, Bituminous mixtures – Test methods – displacement meter can be used
Part 35: Laboratory mixing.

E.3 Terms and definitions E.5.4 Data acquisition system or X-Y recording
instrument
For the purposes of this annex, the following terms and
definitions apply. Capable of automatically acquiring the electric signals
from the transducer and displacement meter and either
storing them in the data acquisition system or plotting the
E.3.1 curves of the load against the midspan torsion on the
bending stiffness modulus in the elastic range recording instrument.
ratio of the bending strength to the bending strain in the
elastic range in megapascals (MPa)
E.5.5 Water thermostat or refrigerator and dry oven
Having a temperature range meeting the test requirements
E.3.2 with an accuracy of ± 0.5 ºC. When the test temperature is
bending stiffness modulus at failure below 0 ºC, a 1:1 methyl alcohol water solution or
ratio of the bending strength to the bending strain at failure antifreeze solution may be adopted as the refrigerant media
in megapascals (MPa) for the water thermostat. The liquid in the water thermostat
shall be able to circulate.

37
E.5.6 Circular saw E.7 Test procedure
Capable of cutting asphalt. E.7.1
When the loading rate is less than 50 mm/min, raise the
E.5.7 Callipers temperature in the ambient temperature cabinet of the
universal test machine, if available, to the value required
E.5.8 Stopwatch by the test.

E.7.2
E.5.9 Thermometer
Position the beam pedestal of the universal test machine
With the division value being 0.5 ºC.
with (200 ± 0.5) mm between the bearing points. Make both
the balance between the upper and lower pressure heads and
E.5.10 Balance the distance on each side equal; then fix the position.
With the sensitivity being no greater than 0.1 g.
E.7.3
E.5.11 Miscellaneous plate glass Take the test beams out of the water-bath or air bath and
immediately put them symmetrically on the pedestal.
E.6 Sample preparation
E.6.1 E.7.4
Place the displacement-measuring device at the centre of the
Prepare asphalt slabs by either: bottom edge of the beam span and fix the pedestal on the
! Manufacturing sufficient asphalt in the laboratory in body of the universal test machine. Put the measuring head
accordance with BS EN 12697-35 and compacting the of the displacement meter at the centre of the top edge of the
asphalt into slabs by roller compactor in accordance midspan of the test beam or (in the case of two displacement
with BS EN 12697-33. meters) put the measuring heads on both sides. Select the
! Manufacturing sufficient asphalt from the asphalt plant measuring range such that the effective range is
and compacting the asphalt into slabs by roller approximately 1.2 times the expected maximum deflection.
compactor in accordance with BS EN 12697-33.
! Cutting slabs from site-compacted asphalt in accordance E.7.5
with BS EN 12697-27. Connect the transducer and displacement meter to the data
acquisition system or the X-Y recording instrument with
Cut the slabs into prismatic beams of length (250 ± 2) mm, the displacement on the X-axis and the load on the Y-axis.
width (30 ± 2) mm and height (35 ± 2) mm. Set the measuring range and zero the readings.
Note: Up to 8 testing piece may be cut down from a Note 1: The midspan torsion can be measured by
300 mm × 300 mm × 50 mm plate. means of LVDT, electric dial gauge or similar
displacement meter.
E.6.2
Measure the dimensions of the test beams at sections of the Note 2: When the displacement of the pressure head of
midspan and the two bearing points by means of callipers. the precision electro-hydraulic servo testing machine
A test beam shall be rejected when the two height or width is taken as the torsion of the beam, the torsion can be
measurements differ by more than 2 mm. The width, b, calculated from the loading rate arid the time recorded
and height, h, of the midspan section shall be the mean of by the X-Y recording instrument.
the relevant measured values with an accuracy of 0.1 mm.
Note 3: In order to record the midspan torsion curve
correctly, the paper moving speed (or scanning speed)
E.6.3 of the Axis X of the recording instrument should be
Measure the maximum density, bulk density and air voids chosen to be between 500 mm/min and 5000 mm/min,
content of the test beams in accordance with depending on the temperature, when the loading rate of
BS EN 12697-5 Procedure A, BS EN 12697-6 50 mm/min is chosen.
Procedure C and BS EN 12697-8, respectively.

E.7.6
E.6.4
Start the universal testing machine, exerting a concentrated
Place the test beams either in the water-bath at the
load on the centre of the span at the specified rate until the
specified temperature for at least 45 minutes or in the
testing piece fails. During the loading, the strain rate shall
constant temperature air bath for at least 3 hours until the
remain constant. At the same time, the recording
test beam reaches the required testing temperature
instrument shall record the load-midspan torsion curve.
± 0.5 ºC. During the heating, the test beams shall be on the
supported plate glass and the spacing between the test Note: Examples of the load-midspan torsion curve are
beams shall be not less than 10 mm. shown in Figure E.1.

38
L = span length of the testing piece in
millimetres (mm).
FB = the maximum load when the testing
Load F or stress S

piece fails in newtons (N).


d = the midspan torsion when the testing
piece fails in millimetres (mm).
Note: In the calculation, the self weight of the beam is
ignored and, therefore, this method does not apply
Brittle example when the test temperature is greater than 30 ºC.
Medium example
Ductile example
E.8.3
Torsion d or strain E Calculate the stress, strain and stiffness modulus at any
time during the loading process using the method in E.8.2.
Figure E.1 Bending test load – midspan torsion curve Note: It is only necessary to take readings of load and
deformation at the moment of failure and substitute
them for the maximum load and failure deformation in
E.7.7 the equations.
When the testing machine is not equipped with an ambient
temperature cabinet, the time period between taking the the E.8.4
test beams out of the water-bath or air bath and the
When the linear region of the load-deformation curve is in
completion of the test shall not be greater than 4 s.
the range of 0.1 to 0.4 times the maximum test load,
calculate the bending stiffness modulus in the elastic range
E.8 Calculations using E.8.1 to E.8.3.

E.8.1 E.8.5
As shown in Figure E.1, extend the linear region of the If the difference between a single measurement and the mean
load-deflection curve to cross the abscissa, taking the value is greater than ‘k’ times of the standard deviation, this
intersection as the origin of the curve. Measure the measured value should be rejected and the mean value of the
maximum load, Fm, and the midspan deflection, d0, at the remaining measured values is to be taken as the test result.
peak point from the graph. The value k shall be 1.15, 1.46, 1.67 and 1.82 when the
number of tests, n, is 3, 4, 5 and 6, respectively.
E.8.2
Calculate the bending strength, RB, the beam bottom E.9 Test report
maximum bending strain, EB, and the bending stiffness The test report shall include the following information:
modulus, SB, at the time of the failure in accordance with a Identification number, type of mixture and composition.
equations (E.1), (E.2) and (E.3), respectively.
b Date and time of testing and name of operator.
3 L FB c Reference to this document.
RB = …. (E.1)
2 b h2 d The method of manufacture of each test beam.
e The dimensions of each test beam.
6h d f Maximum density, bulk density and air voids content of
EB = 2 …. (E.2)
L each beam.
g The test temperature and loading rate.
R
SB = B …. (E.3) h The bending strength, maximum bending strain and
EB bending stiffness modulus at failure of each test beam.
where: RB = bending strength of the asphalt in i The bending stiffness modulus in the elastic range, if
megapascals (MPa). calculated, of each test beam.
EB = maximum bending strain of the asphalt. j Any observation which may have an influence on the
SB = bending stiffness modulus at failure of evaluation.
the asphalt in megapascals (MPa).
b = width of the midspan section testing E.10 Precision
piece in millimetres (mm). No precision data is currently available for this method.
h = height of the midspan section testing
piece in millimetres (mm).

39
Abstract
The current clause in the Specification for Highway Works requires waterproofing systems on concrete bridge
decks to be overlaid with a 20 mm thick sand asphalt protection layer and binder and surface courses so that the
total thickness of the three layers is 120 mm. However, the total thickness on some bridges has to be reduced and, in
such cases, a number of premature failures have occurred. The objective was to develop a specification for
surfacings on concrete bridges that will enhance the probability of achieving reasonable durability when they are
less than the standard thickness. The research has included a literature search, a questionnaire and a laboratory test
programme. The laboratory test programme identified some differences in their properties that have been used to
identify the basis for the specification. A secondary test programme was undertaken specifically to look at tests for
measuring the flexibility of asphalt materials. Based on the findings and other considerations, various additions and
changes to the Design Manual for Roads and Bridges, Specification for Highway Works and Notes for Guidance on
the Specification for Highway Works have been proposed.

Related publications
TRL606 Durability of thin asphalt surfacing systems: Part 2: Findings after 3 years by J C Nicholls and
I Carswell. 2004 (price £40, code EX)
TRL522 A history of the recent thin surfacing revolution in the United Kingdom by J C Nicholls.
2001 (price £35, code E)
TRL494 The behaviour of asphalt in adverse hot weather conditions by J C Nicholls and I G Carswell.
2001 (price £35, code E)
TRL250 Design of long-life flexible pavements for heavy traffic by M E Nunn, A Brown, D Weston and
J C Nicholls. 1997 (price £65, code L)
PR78 Tests for voids and compactin in rolled asphalt surfacings by M E Daines. 1994 (price £35, code E)
PR13 Acceptable weather conditions for laying bituminous materials by J C Nicholls and M E Daines.
1993 (price £35, code E)
LR1023 Trials of buried loints and surfacings on a composite motorway viaduct by A R Price. 1982 (price £30)
Prices current at August 2006

For further details of these and all other TRL publications, telephone Publication Sales on 01344 770783, email:
publications@trl.co.uk, or visit TRL on the Internet at www.trl.co.uk.

40

You might also like