Available online at www.sciencedirect.

com

Construction
and Building

MATERIALS

Construction and Building Materials 22 (2008) 20152024

www.elsevier.com/locate/conbuildmat

The inuence of aggregate, ller and bitumen on asphalt

mixture moisture damage
G.D. Airey
a

a,*

, A.C. Collop a, S.E. Zoorob a, R.C. Elliott

Nottingham Transportation Engineering Centre, University of Nottingham, Nottingham, NG7 2RD, UK

b
Scott Wilson Ltd, 12 Regan Way, Chetwynd Business Park, Chilwell, Nottingham, NG9 6RZ, UK
Received 19 January 2007; received in revised form 4 July 2007; accepted 4 July 2007
Available online 27 August 2007

Abstract
Moisture damage in an asphalt mixture can be dened as the loss of strength, stiness and durability due to the presence of moisture
leading to adhesive failure at the binderaggregate interface and/or cohesive failure within the binder or binderller mastic. Various test
methods exist to identify the susceptibility of asphalt mixtures to moisture damage, such as the AASHTO T283 procedure. This paper
describes a new combined ageing/moisture damage laboratory test known as the Saturation Ageing Tensile Stiness (SATS) test that has
been successfully used to quantify the moisture damage of a range of UK asphalt mixtures. The test consists of initial saturation prior to
placing compacted asphalt mixture cylindrical specimens in a moist, high temperature and pressure environment for an extended period
of time. The stiness modulus measured after the test divided by the stiness modulus measured before the test (retained stiness modulus), and the specimen saturation after the test (retained saturation), are used as an indication of the sensitivity of the compacted mixture to moisture damage. In this paper, the sensitivity of the SATS test to dierent aggregates, llers, binders and volumetric proportions
as well as mixture types has been assessed. The results show that the SATS test is able to discriminate between dierent asphalt mixture
combinations in terms of their moisture damage resistance. Compared to AASHTO T283, the SATS test was found to be a more aggressive conditioning protocol, although both tests ranked mixtures in a similar order with respect to moisture damage.
 2007 Elsevier Ltd. All rights reserved.
Keywords: Moisture damage; Asphalt mixtures; Aggregates; Fillers; Bitumen; Durability

1. Introduction
Moisture damage is an extremely complicated mode of
asphalt mixture distress that leads to the loss of stiness
and structural strength of the bound pavement layers of
a road and eventually the costly failure of the road structure. Essentially the damage is caused by a loss of adhesion
between aggregate and bitumen and/or a loss of cohesion
strength in the bitumen and/or bitumenller mastic due
to the presence of moisture in the asphalt mixture. Various
test methods have been developed in an attempt to identify
the susceptibility of asphalt mixtures to moisture damage
and can generally be divided into those conducted on loose
*

Corresponding author. Tel.: +44 115 9513913; fax: +44 115 9513909.
E-mail address: gordon.airey@nottingham.ac.uk (G.D. Airey).

0950-0618/$ - see front matter  2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2007.07.009

coated aggregate and those conducted on compacted

asphalt mixtures [1]. Tests on compacted mixtures generally use samples either prepared in the laboratory or cored
from existing pavements. Typically, the samples are conditioned in water to simulate in-service conditions and assessment of moisture damage is made by dividing the
conditioned stiness modulus or strength by the unconditioned stiness modulus or strength. Tests of this nature
include the accelerated water conditioning and freezethaw
AASHTO T283 procedure [2]. In addition, immersion
wheel tracking tests, such as the Hamburg wheel tracking
device [3], can be used to assess the moisture damage of
asphalt mixtures.
However, none of these tests has been found to accurately predict the magnitude of moisture damage (strength
and/or stiness reduction) of dierent asphalt mixtures in

2016

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

the eld. Researchers at the Nottingham Transportation

Engineering Centre (NTEC) have therefore recently developed a combined ageing/moisture damage laboratory test
that has been shown to correctly predict the performance
of asphalt mixtures in the eld and replicate the magnitude
of this moisture damage distress [4]. The test, known as the
Saturation Ageing Tensile Stiness (SATS) test [5], consists
of initial saturation under vacuum prior to placing compacted asphalt core samples in a high temperature and
pressure environment in the presence of moisture for an
extended period of time. The stiness modulus measured
after the test divided by the stiness modulus measured
before the test (retained stiness modulus), and the specimen saturation after the test (retained saturation), are used
as an indication of the sensitivity of the compacted mixture
to the combined eects of ageing and moisture.
This paper investigates the sensitivity of the SATS test
to assess the moisture damage performance of various
asphalt mixtures incorporating dierent aggregates, llers,
binders and volumetric proportions as well as dierent
mixture types. The results achieved with the SATS test
have been compared to observed moisture damage eld
performance and the assessment of moisture damage possible with the AASHTO T283 procedure.
2. Test description and procedure
The SATS test is based around the principle of combining ageing with moisture conditioning by conditioning
pre-saturated asphalt mixture specimens at an elevated
temperature (85 C) and pressure (2.1 MPa) in the presence
of moisture for a duration of 65 h [5]. A pressure vessel is
used to hold ve nominally identical specimens (100 mm
in diameter and 60 mm in thickness) in a custom-made
specimen tray as shown in Fig. 1. The dimensions and spec-

ications of the SATS testing equipment, including the size

and spacing of the holes in the perforated trays are detailed
in Clause 953 of Volume 1 of the UK Manual of Contract
Documents for Highway Works, 2004 [6]. The test conditions used with the SATS test were selected in order to
reproduce in the laboratory, the eld observed 60%
decrease in stiness modulus for a high modulus base
asphalt mixture as detailed by Airey et al. [7]. The SATS
test procedure as specied in Clause 953 is as follows:
1. The unconditioned (initial) indirect tensile stiness
modulus of each asphalt mixture specimen is determined at 20 C using the Nottingham Asphalt Tester
(NAT) [8] in accordance with BS DD213 [9]
(124 msec rise time, 5 lm peak transient horizontal
diametral deformation).
2. The dry mass of each specimen is next determined by
weighing.
3. The specimens are subsequently immersed in distilled
water at 20 C and saturated using a residual pressure
of between 40 and 70 kPa for 30 min.
4. The wet mass of each specimen is next determined by
weighing, and the percentage saturation of each specimen calculated, referred to as initial saturation.
5. The SATS pressure vessel is partly lled with a predetermined amount of distilled water (water level
between sample positions P4 and P6 in Fig. 2). The
pressure vessel and water are maintained at the target
temperature of 85 C for at least 2 h prior to introducing the specimens.
6. The saturated asphalt specimens are then placed into
the pressure vessel, the vessel is sealed and the air
pressure is gradually raised to 2.1 MPa.
7. The specimens are maintained at the testing conditions, i.e., 2.1 MPa and 85 C for 65 h.

Fig. 1. SATS pressure vessel and specimen tray.

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

Fig. 2. Specimen arrangement with the SATS test apparatus.

8. After 65 h, the vessel (and hence the specimens) is

allowed to cool to 30 C before the air pressure in
the vessel is gradually released. When the vessel has
achieved atmospheric pressure, the vessel is opened
and the specimens removed. Each specimen is then
surface dried and weighed in air. The percentage saturation calculated at this stage is referred to as the
retained saturation.
9. The specimens are nally brought back to 20 C and
the conditioned (nal) stiness modulus determined
once more using a NAT.
10. The ratio of the nal stiness modulus/initial stiness
modulus can thus be calculated, and is referred to as
the retained stiness modulus.
During the test there is a continuous cycling of moisture
within the pressure vessel, which causes condensation on
the underside of the top lid and dripping onto the top
specimen [5]. There is then a cascading eect where progressively smaller amounts of water drip onto the specimens below, resulting in a decrease in retained saturation
level for specimens that are located deeper inside the pressure vessel. This variation in moisture content (range of
saturation levels) is considered an advantage and allows
the sensitivity of dierent mixtures to a range of moisture
conditions to be investigated in a single test.

(added at 2% by mass of total aggregate) was selected as

the control ller together with a crushed granite ller and
hydrated lime. A total of seven bitumens were used as
the binder for the dierent asphalt mixtures. The binders
diered in terms of their penetration grade and source
(crude oil source as well as production process) [11]. In
addition, a standard French 14 mm Enrobe a` Module
leve (EME) base material was produced using both aggreE
gate types and one binder type. The gradations of the
DBM and EME mixtures are given in Table 1.
A target binder content of 4% by total mixture mass was
selected for the control (standard) 28 mm DBM asphalt
mixtures, while the EME2 mixture had a design binder
content of 5.75%. All the DBM and EME2 asphalt mixtures were manufactured using a slab compactor to produce slabs of dimensions 305 mm 305 mm 100 mm.
Four cylindrical specimens were then cored (100 mm diameter) from the laboratory compacted slabs and trimmed to
produce 60 mm height specimens with a target air void
content for the standard 28 mm DBM specimens of 8
10% (based on Rice Density determinations) and 5% for
the 14 mm EME2 specimens. In total 20 DBM mixtures
and 3 EME mixtures were produced. The mixture design
details of the various asphalt mixtures are presented in
Tables 2 and 3.
Fig. 3 shows a typical graph of retained stiness modulus versus retained saturation for the 28 mm DBM base
material, incorporating the acidic aggregate and the 15B
penetration grade bitumen (15B Acidic Standard). It can
be seen from this gure that there is a range of saturation
levels depending on the position of the specimens in the
pressure vessel as described above and that, for the specimens containing the acidic aggregate, the general trend is
for the retained stiness modulus to decrease as the
retained saturation level increases. At high retained saturation levels (>60%) the retained stiness moduli are lower
than 0.2, indicating that signicant damage is being
induced in the specimen by the conditioning procedure.
In addition, at saturation levels between 20% and 30%,
the retained stiness values are in the order of 0.4 which
compares favourably with the reported 60% drop in

Table 1
Aggregate grading of DBM + EME2 asphalt mixtures
Sieve size (mm)

3. Materials
A large testing matrix of asphalt mixture combinations
were incorporated in the moisture damage study. A standard continuously graded 28 mm Dense Bitumen Macadam (DBM) base material was selected as the control
asphalt mixture [10] and produced using either an acidic
(granite aggregate with known poor eld moisture damage
performance) or a basic (limestone aggregate with good
moisture damage resistance) aggregate. Limestone ller

2017

37.5
28
20
14
10
6.3
3.35
0.3
0.075

28 mm DBM

14 mm EME2

Specication limits
percentage passing (%)

Target specication
percentage passing (%)

100
90100
7195
5882
4460
3246
721
29

100
94
80
64
45
13
7

2018

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

Table 2
Asphalt mixture design details for basic aggregate specimens
Mixture

Gradation

Aggregate

Bitumena

Binder content (%)

Air void content (%)

Added ller

15B Basic standard

15B Basic source #2
15D Basic standard
15F Basic standard
15E Basic EME2
15B Basic high binder
15B Basic low air voids
25P Basic standard
35P Basic standard
50C Basic standard

DBM
DBM
DBM
DBM
EME2
DBM
DBM
DBM
DBM
DBM

Basic
Basic #2
Basic
Basic
Basic
Basic
Basic
Basic
Basic
Basic

15B
15B
15D
15F
15E
15B
15B
15B
15B
15B

4.0
4.0
4.0
4.0
5.75
5.0
4.0
4.0
4.0
4.0

810
810
810
810
5
810
4
810
810
810

2%
2%
2%
2%
2%
2%
2%
2%
2%
2%

limestone
limestone
limestone
limestone
limestone
limestone
limestone
limestone
limestone
limestone

15, 25, 35 and 50 refer to the penetration grade of the bitumen.

Table 3
Asphalt mixture design details for acidic aggregate specimens
Mixture

Gradation

Aggregate

Binder

Binder content (%)

Air void content (%)

Added ller

15B Acidic standard

15B Acidic source #2
15D Acidic standard
15B Acidic granite ller
15B Acidic ller #2
15B Acidic HL
15E Acidic EME2
15E Acidic EME2 HL
15B Acidic high binder
15B Acidic 4% air voids
15B Acidic 6% air voids
15B Acidic 8% air voids
25P Acidic standard
35P Acidic standard
50C Acidic standard

DBM
DBM
DBM
DBM
DBM
DBM
EME2
EME2
DBM
DBM
DBM
DBM
DBM
DBM
DBM

Acidic
Acidic #2
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic
Acidic

15B
15B
15D
15B
15B
15B
15E
15E
15B
15B
15B
15B
25P
35P
50C

4.0
4.0
4.0
4.0
4.0
4.0
5.75
5.75
5.0
4.0
4.0
4.0
4.0
4.0
4.0

810
810
810
810
810
810
5
5
810
4
6
8
810
810
810

2%
2%
2%
2%
2%
2%
2%
2%
2%
2%
2%
2%
2%
2%
2%

1.0

1.2

Basic aggregate
(Blue)

Results from a single test

0.8

4th Position
0.8

0.6

1st Position
(Top)

Immersed
Specimens
(Bottom)

0.4

Retained Stiffness

1.0

Retained Stiffness

limestone
limestone
limestone
granite
limestone #2
hydrated lime
limestone
hydrated lime
limestone
limestone
limestone
limestone
limestone
limestone
limestone

0.6

60%

0.4

0.2

3rd Position

0.2

15B

15D

15E

50C

Acidic aggregate
(Red)

0.0
0

20

40

60

80

100

Retained Saturation (%)

0.0
0

20

40

60

80

100

120

Retained Saturation (%)

Fig. 4. Retained stiness modulus versus retained saturation for asphalt

mixtures containing acidic and basic aggregate types.

Fig. 3. Typical results from SATS test repeats.

stiness modulus from the eld trial using nominally the

same mixture in both the SATS test and trial.
Fig. 4 shows some initial results generated by the SATS
test for asphalt mixtures containing three dierent 15 penetration grade bitumens and one 50 penetration grade bitumen for the two aggregates types (basic versus acidic) [5].
Results from 10 specimens (duplicate SATS tests) are

shown for each asphalt mixture except for mixtures containing the 15B binder where results from 15 specimens
are shown. It can be seen from this gure that, for the
specimens containing the acidic aggregate, the general
trend is for the retained stiness modulus to decrease
as the retained saturation level increases. At high retained
saturation levels, the retained stiness modulus is as low
as 0.2, indicating that signicant damage is being induced
in the specimen by the conditioning procedure. Results

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

for the mixtures containing the basic aggregate show a smaller reduction in retained stiness modulus that levels o
at approximately 0.7 as the retained saturation level is
increased. This indicates that signicantly less moisture
damage occurs in the mixture containing basic aggregate
compared to the mixture containing acidic aggregate. Also
shown in the gure (solid line) is the 60% reduction in stiness found for moisture damaged 15 pen 28 mm DBM base
material (same acidic aggregate) in the eld.
4. Testing parameter investigation
The initial SATS test protocol, used to produce the
results in Fig. 4, allowed the test pressure of 2.1 MPa to
be released as soon as the 65 h test duration was completed,
while the specimens were still at 85 C. This was felt to be
too severe on the test specimens and there were concerns
that additional damage was being done due to the eect
of compressed air expanding relatively rapidly within the
air void network of relatively low stiness specimens at
85 C. The protocol was therefore changed to allow the
specimens to cool to 30 C before the pressure was released.
The eect of releasing the pressure at 85 C using two
specimen tray designs based on the size of the tray holes
(protocol #1 small 2 mm diameter holes and protocol
#2 Clause 953 specied 5 mm diameter holes) compared
to releasing the pressure at 30 C (protocol #3 Clause
953) for the acidic and basic aggregate specimens is shown
in Fig. 5. It can be seen from this gure that the retained
stiness increased by approximately 30% for the basic
aggregate DBM specimens when the Clause 953 protocol
(#3) was used compared to protocols #1 and 2. The extra
reduction in stiness modulus (lower retained stiness)
for protocols #1 and 2 can be attributed to the damage
done to the specimens by reducing the pressure at high
temperatures.
Fig. 5 shows that in the case of DBM mixtures made
with acidic aggregate, depressurising at 30 C (compared
to 85 C) seems to have little eect on the results. However,

2019

it should be noted that the performance for this highly

moisture damage sensitive material combination is already
classed as very poor.
5. Results for basic aggregate materials
Based on the results shown in Fig. 4, the 28 mm DBM
material with basic aggregate was classed as having good
moisture damage resistance when produced with bitumen
from a range of dierent sources and the same binder consistency (based on the penetration grading system). However, research using surface free energy measurements to
quantify adhesive bond energies for bitumenaggregate
combinations with and without water has shown that binders from dierent sources can have signicantly dierent
bond energies with the same aggregate [12]. The sensitivity
of the SATS test to a range of material parameters within
the generic basic aggregate materials group was therefore
investigated.
5.1. Inuence of binder grade
The performance of standard 28 mm DBM asphalt mixtures (4% binder content and 810% air voids) containing
dierent penetration grades of bitumen (15 pen, 25 pen,
35 pen and 50 pen) are compared in Fig. 6.
It can be seen that increasing the binder grade from 15
pen to 25 pen produced very similar SATS results with only
a minor decrease in retained stiness for the softer (25 pen)
bitumen. Reducing the stiness of the bitumen further to
35 pen and 50 pen resulted in a further decrease in moisture
damage resistance with retained stiness values for the 50
pen bitumen 28 mm DBM asphalt mixture approaching
0.5 compared to the retained stiness values of 0.9 for
the harder 15 pen bitumen.
5.2. Inuence of aggregate source and mixture type
The eect of changing aggregate source as well as
asphalt mixture type was investigated by replacing the

1.2

1.2
1.0

0.8

Retained Stiffness

Retained Stiffness

1.0
15B Basic Protocol #1
15B Basic Protocol #2
0.6

15B Basic Protocol #3

15B Acidic Protocol #2
15B Acidic Protocol #3

0.4

0.8

0.6

0.4

0.2

0.2
15B Basic

0.0
0

20

40

60

80

100

120

140

Retained Saturation (%)

Fig. 5. Eect of testing parameters on SATS results for acidic and basic
aggregate 28 mm DBM asphalt mixtures.

25P Basic

35P Basic

50C Basic

0.0
0

20

40

60

80

Retained Saturation (%)

Fig. 6. SATS performance of dierent bitumen grades.

100

2020

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

standard basic aggregate with a second basic aggregate

source commonly used in the UK, using the standard basic
aggregate in a 14 mm EME2 asphalt mixture and nally
using an alternative 15 pen bitumen (15F) in place of the
15B binder. The results presented in Fig. 7 show that the
performance of the EME2 mixture (air void content of
5% and binder content of 5.75% by mass) compares
favourably with conventional basic aggregate DBM mixtures. The only signicant dierence between the two mixtures is that the increase in binder content and decrease in
air voids for the 14 mm EME2 has narrowed the range of
retained saturation levels compared to the standard 28 mm
DBM mixture.
The results in Fig. 7 further show that changing the
basic aggregate source (#2) whilst keeping all other mixture
variables (bitumen grade and source, binder content, gradation and air void content) constant causes no signicant
change in moisture damage resistance. This is not surprising as both basic aggregate sources have long histories of
very good moisture damage performance in the eld.
Finally, the same comments apply when changing the
source of the 15 penetration grade bitumen from 15B to
15F for the standard 28 mm DBM mixture and from 15B
to 15E for the 14 mm EME2 mixture. These results support
the ndings in Fig. 4 for the initial range of 15 pen binder
sources (15B, 15D and 15E).

Using the standard 28 mm DBM material with 15B

bitumen, the eect of altering the volumetrics of the asphalt
mixtures was investigated. The SATS results in Fig. 8 show
that there are minimal eects caused by either reducing the
void content (from approximately 810% to 4%) or by
increasing the bitumen content (from 4% to 5% by mass)
on retained stiness. The retained stiness results over a
range of retained saturation levels are all clustered between
0.8 and 1.2 in a region that has been classed as having very
good moisture damage performance.

5.3. Inuence of mixture volumetrics

6. Results for acidic aggregate materials

In addition to altering the constituent materials, two

means of improving the resistance of an asphalt mixture
to moisture damage is to either increase the binder content
of the mixture (increased binder lm thickness) or reduce
the air void content of the compacted material. The rst
approach theoretically provides a greater barrier (binder
cover) to the eect of water induced bitumenaggregate
adhesive failure, while the second approach limits the
movement of water within the asphalt mixture thereby
reducing the potential for moisture damage.

6.1. Inuence of binder grade

1.2

Retained Stiffness

1.0

0.8

0.6

1.2

Retained Stiffness

1.0

0.8

0.6

0.4

15B Basic Standard

15B Basic High Binder Content

0.2

15B Basic Low Air Void Content

0.0
0

20

40

60

80

100

120

Retained Saturation (%)

Fig. 8. Eect of increasing binder content and reducing air void content.

The eect of binder grade on moisture damage resistance for the standard acidic aggregate 28 mm DBM mixture was studied using the same four binders used in the
basic aggregate study. The SATS results show a slightly
dierent trend to that seen for the basic aggregate mixtures
in Fig. 6.
The results for the acidic aggregate mixtures are shown
in Fig. 9 where the use of the slightly softer 25 pen bitumen
produced very similar results to that of the 15 pen bitumen
samples. However, unlike the results previously seen for the
basic aggregate mixtures, there does not appear to be a
reduction in retained stiness for the softer 35 pen and
50 pen bitumens. This is probably due to the retained stiness values for the acidic aggregate mixtures already being
so low.
6.2. Inuence of ller type

0.4
15B Basic Standard

15B Basic Source #2

15F Basic Standard

15E Basic EME2

0.2

0.0
0

20

40

60

80

100

120

Retained Saturation (%)

Fig. 7. Eect of basic aggregate source, binder source and mixture type on
asphalt mixture moisture damage in the SATS test.

The eect of ller type on moisture damage performance

is shown in Fig. 10. It can be seen from this gure that
substituting the conventional 2% by mass limestone ller
with granite ller resulted in slightly poorer performance,
whilst the replacement of 2% limestone ller with hydrated
lime resulted in a signicantly improved performance.
The poorer performance of the granite ller is expected,
although the magnitude of this decrease is relatively small

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

#2 results in an approximately 50% increase in retained

stiness.

1.2
15B Acidic Standard

Retained Stiffness

1.0

2021

25P Acidic Standard

6.3. Inuence of aggregate source and mixture type

35P Acidic Standard

0.8

50C Acidic Standard

0.6

0.4

0.2

0.0
0

20

40

60

80

100

120

140

Retained Saturation (%)

Fig. 9. Eect of bitumen grade on the SATS results for acidic aggregate
mixtures.

and is probably a consequence of the high amount of natural acidic aggregate ller (6% by mass of total aggregate)
already present in the 2% added limestone ller mixture.
The performance of the hydrated lime ller mixture is also
not surprising as it is well recognised that adding an antistripping agent such as hydrated lime will improve the performance of moisture susceptible asphalt mixtures. For
example, Kennedy and Ping [13] investigated the eectiveness of hydrated lime and other anti-stripping agents using
two moisture susceptibility tests. They tested a range of
plant-prepared and laboratory-prepared asphalt mixtures
containing dierent amounts of the additives and concluded that hydrated lime was most eective in improving
the performance of moisture sensitive asphalt mixtures.
Similar results on the eect of hydrated lime have been
found by Lesueur and Little [14] and Kim et al. [15].
The considerable inuence of ller type and source on
moisture damage can also be seen in Fig. 10 when comparing the SATS results for the standard 15B acidic aggregate
mixture with an identical mixture using a second limestone
ller source. The replacement of the 2% limestone ller in
the standard mixture with 2% limestone ller from source

In addition to testing the 28 mm DBM mixture with the

control acidic aggregate, an alternative acidic aggregate
(source # 2) was tested using the same 15B control bitumen. The results are presented in Fig. 11 where it can be
seen that the source #2 acidic aggregate showed improved
performance and also a noticeable clustering of results
around a retained saturation level of 40% for those specimens placed above the water reservoir in the SATS
apparatus.
The eect of changing asphalt mixture type was also
investigated by using the standard acidic aggregate in a
14 mm EME2 asphalt mixture. The results show that the
performance of the EME2 mixture, with its reduced air
void content of 5% and increased binder content of
5.75% by mass, leads to a marginal increase in retained
stiness as a function of retained saturation, although the
moisture damage resistance of the mixture is still considered poor based on the threshold of a 60% decrease in
stiness shown in Fig. 4. However, the moisture damage
performance of the acidic aggregate EME2 was signicantly improved once the limestone ller, used in the standard mixture, was replaced with hydrated lime. Similar to
the eect seen in Fig. 10, the hydrated lime modied
14 mm EME2 showed results comparable with those
obtained for the basic aggregate mixtures presented in Figs.
68.
6.4. Inuence of volumetrics
The inuence of air void content on the moisture damage performance of the 28 mm DBM control asphalt mixture is shown in Fig. 12. Samples at an average air void
content of 4%, 6% and 8% were compared to the standard
15B acidic aggregate DBM mixture with an air void

1.2

1.2
1.0

Retained Stiffness

Retained Stiffness

1.0
15B Acidic Standard

0.8

15B Acidic Limestone Filler #2

15B Acidic Hydrated Lime

0.6

15B Acidic Granite Filler

15B Acidic Standard

0.8

15B Acidic Source #2

15E EME2 Acidic Limestone Filler

0.6

15E EME2 Acidic Hydrated Lime Filler

0.4

0.4
0.2

0.2
0.0
0

0.0
0

20

40

60

80

100

120

140

Retained Saturation (%)

Fig. 10. Eect of ller type and source on SATS moisture damage.

20

40

60

80

100

120

140

Retained Saturation (%)

Fig. 11. Eect of aggregate source and mixture type on moisture damage
performance.

2022

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

1.2
15B Acidic Standard

Retained Stiffness

1.0

15B Acidic 4% Air Voids

15B Acidic 6% Air Voids

0.8

15B Acidic 8% Air Voids

0.6

0.4

0.2

0.0
0

40

80

120

160

200

Retained Saturation (%)

Fig. 12. Eect of air void content on retained stiness versus retained
saturation.

content ranging between 8% and 10%. Similar to the results

for the 5% air void EME2 mixture in Fig. 11, the eect of
reducing the air void content of the acidic aggregate mixtures had no signicant eect on retained stiness,
although the 4% DBM mixture did have two results above
a retained stiness of 0.8. However, these two results were
obtained at a relatively low retained saturation level of less
than 20%.
7. Comparison with AASHTO T283
A set of the results from the SATS test were also compared
with moisture damage assessment results produced with the
accelerated water conditioning and freezethaw AASHTO
T283 procedure [2]. A total of eight asphalt mixtures were
selected incorporating both asphalt mixture types, both
aggregate types, hydrated lime ller, dierent sources of
bitumen and nally dierent grades of bitumen.
Although the principal steps of the AASHTO T283 procedure were followed, specic aspects of the method were
altered to allow the results to be compared to those obtained
from the SATS protocol. These modications are described
in detail in Airey et al. [7] and include the determination of a
stiness modulus ratio in addition to the required indirect

tensile strength ratio. The nal (retained) saturation level

in the asphalt mixture specimens was also determined after
the AASHTO T283 moisture conditioning procedure similar
to that used with the SATS test.
The results in terms of nal saturation level, AASHTO
T283 strength and stiness ratio, equivalent SATS stiness
ratio and nally eld performance are presented in Table 4.
As the SATS test produces a range of saturation levels and
related stiness ratios for each mixture type, a procedure
was derived to obtain a single SATS stiness ratio to compare with the AASHTO T283 stiness ratio. The equivalent
SATS retained stiness value for each of the asphalt mixtures was obtained by estimating the retained stiness value
that corresponded to the nal saturation level obtained
from the AASHTO T283 test using the range of retained
stiness versus retained saturation results. The asphalt mixtures in Table 4 were ranked in descending order in terms
of their measured moisture damage performance based
on the AASHTO T283 stiness ratio values. A stiness
ratio of 0.6 can be used as the threshold to distinguish
between good and poor asphalt mixture eld moisture
damage performance.
In general, the moisture damage ranking order of the
eight asphalt mixtures is similar irrespective of whether
the two AASHTO T283 parameters or the SATS test
parameter are used. Values that do not t the ranking
order determined by the AASHTO T283 stiness ratio
have been highlighted in bold in the table and their new
position indicated. Of more signicance are the actual values of the stiness ratios for the two moisture damage
methods. These results have been plotted against each
other in Fig. 13 and show that the accelerated water conditioning procedures associated with AASHTO T283,
including water saturation and a freezethaw cycle, are
not as severe as the combined ageing/moisture conditioning SATS procedure when assessing the performance of
the eight asphalt mixtures. Only the 15B Acidic Hydrated
Lime mixture lies on the retained stiness equivalency line
with the results for the other seven mixtures showing considerably higher retained stiness results after the
AASHTO T283 test than after the SATS test.

Table 4
AASHTO T283 versus SATS test results
Mixture

Final saturation level

(%)

AASHTO T283 strength

ratio

AASHTO T283 stiness

ratio

Equivalent SATS stiness

ratio

Field
performance

15B Acidic hydrated

lime
15B Basic standard
50C Basic standard
15E Basic EME2
15B Acidic standard
15B Acidic 4% air
voids
15B Acidic 5% b/c
50C Acidic standard

50

0.86 (4th)

1.04

1.05

Good

47
63
42
48
37

0.91 (3rd)
1.18 (1st)
0.97 (2nd)
0.71
0.62

0.99
0.98
0.95
0.77
0.77

0.90
0.55 (4th)
0.90 (3rd)
0.25
0.25

Good
Good
Good
Poor
Poor

62
65

0.58
0.46

0.63
0.53

0.25
0.10

Poor
Poor

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

AASHTO T283 Stiffness Ratio

1.2

0.8

15B Acidic Hydrated Lime

15B Basic Standard
50C Basic Standard
15E Basic EME2
15B Acidic Standard
15B Acidic 4% Air Voids
15B Acidic 5% b/c
50C Acidic Standard

0.6

0.4

0.2

0
0

0.2

0.4

0.6

0.8

1.2

SATS Retained Stiffness

Fig. 13. SATS retained stiness versus AASHTO T283 retained stiness.

8. Discussion
The principal objective of the SATS test was to develop
a combined ageing/moisture sensitivity laboratory test to
evaluate and replicate the moisture damage performance
of high modulus base materials containing relatively low
binder contents and high air void contents. This was
achieved and the test has been shown in addition to be able
to assess the relative moisture damage performance of a
range of asphalt mixtures. Both the SATS procedure and
AASHTO T283 are able to rank the moisture damage performance of asphalt mixtures with the SATS test, based on
comparison with eld performance of asphalt mixtures in
the UK, providing a more accurate prediction of the degree
of moisture damage.
However, both these current laboratory procedures for
assessing moisture damage of asphalt mixtures rely on
comparing mechanical properties of unconditioned specimens with moisture conditioned specimens. Although this
approach is helpful in comparative analysis of the moisture
susceptibility of various mixtures, it does not focus on measuring the fundamental material properties related to the
mechanisms described above. As such, the results cannot
be used to explain causes for poor or good performance,
and do not provide feedback into the process of redesigning better performing mixtures. It is therefore necessary
to supplement the normally measured mechanical properties with fundamental properties that aect physical adhesion between the bitumen and aggregate and internal
mastic cohesion and the propensity to lose these bonds in
the presence of water.
Moisture damage of an asphalt mixture can be attributed to a reduction in the adhesive and cohesive bonds in
the material due to the action of water leading to a softening or weakening of the mixture. Recent studies have
shown that a comprehensive characterisation of moisture
damage should include measurements of fracture, healing,
and viscoelastic properties [1618]. These mechanical properties are inuenced by the binderaggregate bond energy,

2023

and other physical characteristics of the mixture such as

volumetrics, bitumen lm thickness, air void distribution,
aggregate gradation and aggregate shape. These adhesive
and cohesive bonds are directly related to the surface
energy characteristics of bitumen and aggregate allowing
adhesive bond parameters with and without the presence
of water to be calculated for various bitumenaggregate
ller combinations [12,19,20].
A future approach to moisture damage assessment
would be to combine the mechanical assessment of asphalt
mixtures provided by the SATS test with a fundamental
surface energy approach. This approach would evaluate
the susceptibility of aggregates and bitumens to moisture
damage by understanding the mechanisms that inuence
the adhesive bond between aggregates and bitumen and
the cohesive strength and durability of the bitumen and/
or bitumenller mastic. The principles of surface energy
and the calculation of cohesive and adhesive bond energies
would allow the moisture sensitivity of dierent bitumen
aggregate combinations to be quantied and correlated
with moisture damage performance of the asphalt mixtures
provided by the SATS test.
9. Conclusions
The SATS test has been developed to evaluate the combined eects of ageing and moisture damage on asphalt mixtures. The test consists of ageing compacted asphalt mixture
specimens (100 mm in diameter and 60 mm in thickness) at
an elevated temperature and pressure in the presence of
moisture. The pressure vessel used in the test is capable of
holding ve nominally identical specimens with each specimen being conditioned at a dierent saturation level ranging
from high saturation levels (6090%) at the top of the vessel
to low levels (1040%) at the bottom of the vessel (above the
submerged fth specimen). Retained stiness modulus values are then determined for the specimens as a function of
their nal (retained) saturation level. The test has been able
to successfully reproduce in the laboratory the loss of stiness modulus observed in the eld for moisture damaged
high modulus base materials.
In this paper, the SATS test has been used to determine the resistance to moisture damage of a range of
asphalt mixtures incorporating an acidic aggregate (poor
eld performance) as well as a basic aggregate (good eld
performance). In addition to the two aggregates, dierent
ller types, binder types (source and grade) and mixture
gradations have been assessed. The results from the SATS
test have also been compared with those obtained on the
same mixtures using the AASHTO T283 procedure. The
following specic conclusions have been drawn from this
study:
 In terms of the SATS testing protocol, for both the basic
and acidic DBM asphalt mixtures, reducing the temperature to 30 C prior to pressure release is essential for
maintaining specimen integrity.

2024

G.D. Airey et al. / Construction and Building Materials 22 (2008) 20152024

 In terms of mixture volumetrics, reducing the air voids

content from approximately 104% or increasing the
bitumen content from 4% to 5% was found to have a
negligible eect on the retained stiness values for both
the basic and acidic DBM mixtures when tested in the
SATS test.
 The performance of the EME2 mixtures compared
favourably with conventional basic and acidic DBM
mixtures.
 In general, reducing the stiness of the asphalt mixture
binder for the basic DBM mixtures resulted in a reduction in resistance to moisture damage.
 The eect of ller type was found to aect the moisture
damage performance of the acidic aggregate asphalt
mixtures. Substitution of 2% added limestone ller with
hydrated lime dramatically improved the SATS results
for the acidic DBM and EME2 mixtures, while substitution of the added limestone ller by granite ller resulted
in lower retained stiness values.
 Compared to AASHTO T283, the SATS test was found to
be a more aggressive conditioning protocol, although
both tests ranked mixtures in a similar order with respect
to moisture damage. The increased severity of the SATS
conditioning procedure results in larger reductions in stiness modulus and therefore lower stiness ratios for materials known to have poor eld moisture damage
performance compared to the AASHTO T283 procedure.
The SATS test is therefore better at discriminating
between good and poor performing asphalt mixtures.

Acknowledgements
The work reported herein was carried out under a contract placed with Scott Wilson Pavement Engineering by
the UK Highways Agency. The views expressed in this
paper are not necessarily those of the Highways Agency
or the Department for Transport.

[5]

[6]

[7]

[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

[16]

[17]

References
[18]
[1] Airey GD, Choi Y-K. State of the art report on moisture sensitivity
test methods for bituminous pavement materials. Int J Road Mater
Pavem Des 2002;3(4):35572.
[2] American Association of State Highways and Transportation Ocials. Resistance of compacted bituminous mixture to moisture
induced damage AASHTO T283-99, USA; 2000.
[3] Aschenbrener T. Evaluation of hamburg wheel-tracking device to predict
moisture damage in hot-mix asphalt. Transport Res Record
1995;1492:193201. TRB, National Research Council, Washington DC.
[4] Airey GD, Choi YK, Collop AC, Elliott RC. Development of an
accelerated durability assessment procedure for high modulus base
(HMB) materials. In: Proceedings of the 6th international RILEM

[19]

[20]

symposium on performance testing and evaluation of bituminous

materials, PTEBM03, Zurich; 2003. p. 1606.
Collop AC, Choi Y, Airey GD, Elliott RC. Development of the
Saturation Ageing Tensile Stiness (SATS) Test. ICE J Transport
2004;157(Issue TR3):16371.
The Highways Agency. Manual of contract documents for highway
works, vol. 1, Specication for highway works, Clause 953: durability
of bituminous materials Saturation Ageing Tensile Stiness (SATS)
Test, Highways Agency, Series 900; November 2004.
Airey GD, Choi Y, Collop AC, Moore AJV, Elliott RC. Combined
laboratory ageing/moisture sensitivity assessment of high modulus
base asphalt mixtures. J Assoc Asphalt Paving Technol
2005;74:30745.
Cooper KE, Brown SF. Developments of a simple apparatus for
the measurement of the mechanical properties of asphalt mixes.
In: Proceedings of the eurobitume symposium, Madrid; 1989. p.
4948.
British Standards Institution, Method for determination of the
indirect tensile stiness modulus of bituminous mixtures, BS DD
213; 1993.
British Standards Institution, Coated macadam for roads and other
paved areas, BS 4987: Part 1. Specication for constituent materials
and for mixtures; 1993.
Airey GD, Choi Y, Collop AC, Elliott RC. Rheological and fracture
characteristics of low penetration grade bitumen. Int J Road Mater
Pavem Des 2004;5:10731. Special Issue.
Bhasin A, Masad E, Little D, Lytton R. Limits on adhesive bond
energy for improved resistance of hot mix asphalt to moisture
damage. Presented at 85th annual meeting of the Transportation
Research Board, Washington DC; 2006.
Kennedy TW, Ping WV. An evaluation of the eectiveness of
antistripping additives in protecting asphalt mixtures from moisture
damage. J Assoc Asphalt Paving Technol 1991;60:23063.
Lesueur D, Little DN. Eect of hydrated lime on rheology, fracture
and aging of bitumen. Transport Res Record: J Transport Res Board
1999;1661:93105. TRB, National Research Council. Washington,
DC.
Kim Y-R, Little DN, Song I. Mechanistic evaluation of mineral llers
on fatigue resistance and fundamental material characteristics.
Transport Res Record: J Transport Res Board 2003;1832:18.
TRB, National Research Council. Washington, DC.
Cheng D, Little DN, Lytton RL, Holste JC. Use of surface free
energy of asphaltaggregate system to predict moisture damage
potential. J Assoc Asphalt Paving Technol 2002;71:5988.
Hefer AW, Little DN, Lytton RLA. Synthesis of theories and
mechanisms of bitumenaggregate adhesion including recent advances in quantifying the eects of water. J Assoc Asphalt Paving
Technol 2005:74.
Masad E, Zollinger C, Bulut R, Little D, Lytton R. Characterization
of HMA moisture damage using surface energy and fracture
properties. J Assoc Asphalt Paving Technol 2006;75.
Cheng D, Little DN, Lytton RL, Holste JC. Moisture damage
evaluation of asphalt mixtures by considering both moisture diusion
and repeated-load conditions. Transport Res Record: J Transport
Res Board 2002;1832:429. TRB, National Research Council.
Washington, DC.
Cheng D, Little DN, Lytton RL, Holste JC. Surface energy
measurement of asphalt and its application to predicting fatigue
and healing in asphalt mixtures. Transport Res Record: J Transport
Res Board 2002;1810:4453. TRB, National Research Council.
Washington, DC.