EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY

METHOD DESIGN

Xiao-Lin Chou 1, A. Chegenizadeh2* and H. Nikraz3

1
Graduate School of Civil and Mechanical Engineering, Curtin University, Australia
2
* Research Fellow, School of Civil and Mechanical Engineering, Curtin University, Australia
Corresponding author: [email protected]
3
Professor, School of Civil and Mechanical Engineering, Curtin University, Australia

ABSTRACT
A series of wheel tracking tests are performed in this study. Mixtures are designed based on a designer’s experience and
grading curves developed from past practice. This has led to many issues with inconsistent and poor performing SMA
on the focus of rutting resistance. An investigation to evaluate the Bailey Method as a tool for designing SMA mix built
on aggregate packing was undertaken. The Bailey Method was used to design a SMA blend which was then
manufactured, tested and compared against conventional SMA used in Western Australia. Results show that the degree
of aggregate packing directly influences the rutting resistance performance of Stone Mastic Asphalt using similar
aggregates. The Bailey Method provided a useful tool in improving the behaviour of aggregate packing characteristics
within SMA.

1 INTRODUCTION
The properties of the aggregate greatly influence the performance of asphalt mixtures. This is because aggregates
commonly make up the largest portion of an asphalt mixture by weight or volume. It is for this reason that one of the
most important properties of an asphalt blend is the aggregate gradation (Manjunath and Poornachandra, 2014). The
effect of aggregate gradation is generally the most important design parameter as it directly impacts performance
properties such as deformation resistance, fatigue resistance, stiffness, durability and workability of asphalt. SMA is
known for its high rut resistance properties due to the structural skeleton (stone-on-stone contact) of aggregates
predominant within the mix. The interlock between aggregates provides better load carrying capacity and durability
properties, which is why SMA can be used for heavily trafficked roads that require asphalt surfacing with high
resistance to deformation. Aggregate packing is the orientation and placement of the aggregates within a mixture. This
is applicable for SMA as the mix requires the contact between particles to transfer loads (Qiu and Lum, 2006). There is
no performance design method for optimizing aggregate packing for design blends to ensure stone-on-stone contact
between particles (Brown and Haddock, 1997). Commercially available SMA mixtures in Western Australia have
shown a need for attention to terms of performance in the Wheel Tracking Test for rutting. It is believed that designers
must undergo trial and error in SMA design procedures to determine a design blend based on the aggregates available.
There is some research in the literature which have tried to describe aggregate matrix in asphalt (Feng, Ye and Hao,
2013; Chen et al., 2015; Li and Li, 2012; Underwood and Kim, 2013) but still there is a need to investigate further
especially in Western Australia. Current practice typically relies on trial and error procedures or previous experience of
designers to design SMA mixes. Therefore, the aim of this research is to propose a method that can be available to
designers to ensure that stone-on-stone contact will be achieved based on the aggregate(s) properties. This paper is
extracted from the first author’s thesis and is part of an ongoing research in Curtin University.

2 MATERIALS AND METHODS

2.1 BAILEY METHOD

In the early 1980’s, Robert Bailey developed the Bailey Method based on his personal experiences with asphalt mix
designs. The purpose for developing this method was to improve the permanent deformation due to rutting of asphalt
while maintaining durability (Vavrik et al., 2002). The Bailey Method provides the tools for evaluating gradation
parameters that directly influence a mixture’s performance properties (Al Shamsi et al., 2006).
The Bailey Method goes through a series of steps to evaluate the aggregate packing characteristics within a blend. The
method firstly defines what particles make the up the coarse and fine aggregates of a blend. This is known as the break
point sieve or the Primary Control Sieve (PCS). This is defined as 0.22 times the Nominal Mineral Particle Size
(NMPS). The portion above the break is considered to be the coarse fraction and the portion below as the fine fraction.
The aggregates within the coarse fraction of the blend are considered to form an aggregate structure that creates voids
and transfers the loads. The fine fraction is then designed to fill in the created voids to ensure that the aggregates are

AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016 105

AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016 105
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

further packed together. For a NMPS of 10 mm the PCS = 0.22 x 10, which equals to 2.2. As there is no sieve of this
size, the closest sieve is chosen as the PCS which is the 2.36 mm sieve. Shen and Yu (2011) found that aggregates of
2.36 mm or higher contributed to the formation of an aggregate skeleton within a mixture to transfer loads. Aggregates
smaller than 2.36 mm within the blend also provided transmission of stresses, however these are considered too weak to
be considered as part of the aggregate structure of a blend.
Traditionally, aggregates for a blend are combined by weight method which is also represented by a gradation curve
that shows the percentage passing of each sieve. However, this lacks in determining the degree of interlock possible
within a mixture due to varying particle densities of each aggregate used. To evaluate the degree of particle interlock
desired to ensure the aggregates are packed together, the Bailey Method goes through combining aggregates by volume
of each aggregate’s unit weight. The unit weight of bulk density is obtained by determining the Loose Unit Weight
(LUW) and Rodded Unit Weight (RUW) of each individual aggregate stockpile. The LUW is the contact between
particles with no compactive effort, whereas the RUW is the contact between particles with compactive effort. As SMA
requires the stone-on-stone contact between particles, the RUW is used to design the desired aggregate packing of a
SMA blend.
To design the degree of aggregate packing desired within the blend The Design Unit Weight (DUW) sets the degree of
particle interlock within the mixture. Typical design values or the Chosen Unit Weight (CUW) generally ranges from
110% to 130% of the RUW for SMA. By increasing the design RUW, the contact between particles is increased, thus
ensuring a stone-on-stone contact between the particles is prominent within the mixture

2.2 SPECIMEN DESIGN

Commonly used coarse aggregates used for Western Australian roads are crushed aggregate with a NMPS of 10 mm
suitable for SMA mixture designs. 7/10 mm BGC Dust was used for fine aggregate with Baghouse Fines widely
acceptable for mineral filler use. The standard class of bitumen used in Western Australia is Class 320. Cellulose fibres
are commonly used in SMA mixtures in Australia. Cellulose fibres usually make up to 0.3% of the total blend by mass
and this was allocated into the design (Alderson, 2013).
Each individual aggregate’s gradation, particle density and unit weights were obtained to design two desired aggregate
structures: 120% RUW and 130% RUW. Using the Bailey Method, the design outputs are shown in Tables 1 and 3 of
both designs respectively. The Bailey Ratios are also shown in Table 2 and 4. The desired mineral filler content was
kept at 11%.

106 AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016

106 AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

Table 1: 120% RUW Design Blend

Main Main
10mm 7/10 mm Roads Roads
Baghouse Combined
Sieve Crushed BGC
Fines Gradation
WA WA
Granite Dust SMA10 SMA10
Max Min
26.5 100 100 100 100.00 100.00 100.00
19 100 100 100 100.00 100.00 100.00
13.2 100 100 100 100.00 100.00 100.00
9.5 88.72 100 100 91.25 100.00 90.00
6.7 22.84 100 100 40.12 35.00 25.00
4.75 4.4 99.7 100 25.76 30.00 18.00
2.36 1.65 89.7 100 22.28 28.00 15.00
1.18 1.44 68.8 100 19.30 24.00 13.00
0.6 1.33 50.29 100 16.72 21.00 12.00
0.3 1.19 32.96 99.87 14.26 18.00 10.00
0.15 1 20.42 96.58 12.13 14.00 9.00
0.075 0.7 12.91 84.44 9.81 12.00 8.00
Particle Density (t/m3) 2.72 2.58 2.74
Water Absorption (%) 0.6 1.17
LUW (t/m3) 1.42 1.58
RUW (t/m3) 1.56 1.78
CUW (%) 120
DUW (kg) 1864.80
Binder Content (%) 5.5
Cellulose Fibres (%) 0.3
Design (%) 78.1 12.87 9.03 = 100% of the aggregate
blend

Table 2: 120% RUW Bailey Ratios

Bailey SMA Min Max
% CA RUW 120 100.0 130.0
CA Ratio 0.05 0.15 0.30
FAc Ratio 0.76 0.60 0.85
FAf Ratio 0.73 0.65 0.90

AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016 107

AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016 107
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

Table 3: 130% RUW Design Blend

Main
Main
10mm Roads
Baghouse Combined Roads WA
Sieve Crushed BGC Dust
Fines Gradation
WA
Granite SMA10
SMA10
Max
Min
26.5 100 100 100 100.00 100.00 100.00
19 100 100 100 100.00 100.00 100.00
13.2 100 100 100 90.68 100.00 100.00
9.5 88.72 100 100 36.25 100.00 90.00
6.7 22.84 100 100 20.99 35.00 25.00
4.75 4.4 99.7 100 17.94 30.00 18.00
2.36 1.65 89.7 100 16.15 28.00 15.00
1.18 1.44 68.8 100 14.62 24.00 13.00
0.6 1.33 50.29 100 13.15 21.00 12.00
0.3 1.19 32.96 99.87 11.70 18.00 10.00
0.15 1 20.42 96.58 9.70 14.00 9.00
0.075 0.7 12.91 84.44 12.00 8.00
Particle Density (t/m3) 2.72 2.58 2.74
Water Absorption (%) 0.6 1.17
LUW (t/m3) 1.42 1.58
RUW (t/m3) 1.56 1.78
CUW (%RUW) 130
DUW (kg) 2020.20
Binder Content (%) 7
Cellulose Fibres (%) 0.3
Design % 82.90 7.38 9.72 = 100% of the aggregate
blend

Table 4: 130% RUW Bailey Ratios

Bailey SMA Min Max

% CA RUW 130 100.0 130.0
CA Ratio 0.04 0.15 0.30
FAc Ratio 0.82 0.60 0.85
FAf Ratio 0.81 0.65 0.90

108 AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016

108 AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

The combined gradation curve of both mix designs is shown in Figure 1, which demonstrates that by increasing the
desired CUW from 100% to 130%, the design mixture blend becomes coarser. This is because the increase in RUW of
the aggregate increases the coarse aggregate content by volume.

Figure 1: Combined Gradation Curves

From the above design outputs, three slabs were prepared as shown in Table 5. These slabs were then tested on the
Wheel Tracking Testing apparatus. Specimens are noted as Slab #1, Slab #2 and Slab #3 respectively. It should be
noted that due to the nature of SMA mixes consisting of predominately coarse aggregates (creating voids), volumetric
requirements are generally higher as a result of a coarser blend. Main Roads WA Specification 502 Stone Mastic
Asphalt, specifies a minimum VMA of 18% for a nominal aggregate size of 10mm. This criteria ensures that the
minimum air void requirement will be met and that adequate asphalt binder content is present within the mixture to
eliminate durability issues. It can be seen that Slab #1 and Slab #2, yielded a VMA of 17.6%. This is a result of the
design method of optimizing aggregate packing, where fine aggregates are designed to fill in the voids, thus reducing
the overall VMA. The minimum and maximum air void content of 3.5% and 5.5%, was also satisfied.
Table 5: Volumetric Properties of Test Specimens
Slab #1 Slab #2 Slab #3
Design Blend 120% RUW 120% RUW 130% RUW
Bulk Density (t/m3) 2.372 2.374 2.358
Maximum Density (t/m3) 2.464 2.464 2.489
Air Voids (%) 4.70 4.62 4.30
VMA (%) 17.6 17.6 20.8
VFB (%) 75.6 75.6 76.4

2.3 TEST PROCEDURE

The deformation resistance or rutting performance of the tested specimens was tested using the Wheel Tracking Test.
The procedure follows placing the specimen under a wheel load (700 ± 20 N) for a minimum of 10,000 passes under a
conditioning temperature of 60 degrees Celsius. Testing was carried out in accordance with AG:PT/T231Deformation
Resistance of Asphalt Mixtures by the Wheel Tracking Test.

AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016 109

AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016 109
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

Figure 2: Cooper Wheel Tracking Testing Machine.

Figure 3: (Left) Temperature Probe Inserted in the Slab, (Right) Wheel Load Acting on Slab.

3 RESULTS AND DISCUSSION

Three specimens were prepared and tested on the Wheel Tracking Test. A commercially available SMA 10 blend was
also supplied from industry and subjected to the test. A comparison of the rut (tracking depth) progression results can be
seen in Table 6. The rutting performance from each specimen showed varied results. This was due to several issues
arising within the course of manufacturing SMA. However, comparing the rutting performance of commercially
available SMA, the 120% RUW design showed far superior performance in the rut depths analysed. A rut depth of 5.5
mm and 8.8 mm was observed for Slab #1 and Slab #2 respectively. The difference in results can be attributed to how
sensitive SMA mixes are to volumetric changes. Although both slabs are the same design, with a difference of 0.08%
air void content, SMA mixes require higher binder content due to the coarse aggregate nature of the mix. This results in
producing a very sticky bituminous mixture, difficult to manufacture compared to other conventional HMA mixtures.
Maintaining conditioning temperatures and compaction of test slabs also proved to be difficult. Other manufacturing
discrepancies may have also contributed to the varied results.
Slab #1 has shown almost half of the rut depth of the commercially available SMA of 10.9 mm observed under the
same testing conditions. This shows an improved rutting performance of the SMA blends that were designed using the
design analysis tool. This has yielded a good result, indicating that the designs using the Bailey Method developed for
this research project achieved better rutting resistance performance. However, it can be seen that although Slab #3 had a
coarser design blend, the results showed that the specimen yielded a final rut depth of 12.8 mm as shown in Figure 4. It
should be noted that the desired mineral filler content was kept at 11% as previously mentioned under Section 2.2. This

110 AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016

110 AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

then resulted in a higher binder content for a very coarse 130% RUW mixture, causing the mixture to become highly
viscoelastic (i.e. easily deformed under loading). However, if the mineral filler content were increased, it is presumed
that Slab #3 would have performed better (i.e. additional mineral filler, thus reducing voids, thus reducing binder
content required, resulting in a less viscoelastic mixture)
Table 6: Final Rut Depths of Tested Specimens
Slab #1 Slab #2 Slab #3 Commercial SMA
Final Rut Depth (mm) 5.5 8.8 12.8 10.9

Figure 4: Combined Rut Progression of Tested Specimens.

4 CONCLUSION
This research hoped to improve resistance to the rutting deformation of existing commercially available SMA mix
through design of aggregate packing characteristics. An investigation into the Bailey Method for improving aggregate
packing characteristics of SMA was undertaken. Results show that the SMA design blends by the Bailey Method for
120% RUW design showed far superior rutting resistance performance. However, by increasing to 130% RUW design
the mixture did not perform as well as expected. It can be decided that, although increasing the coarse aggregate content
of a mixture does increase the particle contact of the aggregates, this does not necessarily mean the deformation
resistance of the mixture increases. In fact, as a result of a higher coarse aggregate content within a blend, more bitumen
is required within the mixture, resulting in high binder content and a weaker blend. SMA is a gap graded mixture that
consists of predominately aggregates. As a result the slightest changes within the aggregate skeleton (e.g. adjusting the
mineral filler content), directly influences the performance properties of SMA. As there is currently no performance
design method that asphalt mix designers should conform to, the Bailey Method can be used as a tool for a starting
point in designing aggregate packing properties of SMA.

5 REFERENCES
Al Shamsi, Khalid, LN Mohammad, Zhong Wu, Sam Cooper, and Chris Abadie (2006). Compactability and
Performance of Superpave Mixtures with Aggregate Structures Designed Using the Bailey Method., Journal of
the Association of Asphalt Paving Technologists, 75, 91-131, Louisiana Transportation Research Center.
Alderson, Allan (2013). Ap-T221-13 Austroads Technical Report: Development of Aspects of a Design Procedure for
Stone Mastic Asphalt, Austroads Ltd. Sydney, NSW. Source agency: ARRB

AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016 111

AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016 111
EVALUATION OF STONE MASTIC ASPHALT MIX BY THE BAILEY METHOD DESIGN CHOU ET AL.

Brown, Ray E, and John E Haddock (1997). A Method to Ensure Stone-on-Stone Contact in Stone Matrix Asphalt
Paving Mixtures., NCAT Report No. 97, 2, 3, National Centre for Asphalt Technology.
Chen, Jun, Hui Li, Linbing Wang, Jiantao Wu, and Xiaoming Huang. (2015). Micromechanical Characteristics of
Aggregate Particles in Asphalt Mixtures. Construction and Building Materials 91: 80-85. doi:
http://dx.doi.org/10.1016/j.conbuildmat.2015.05.076.
Feng, Xinjun, Song Ye, and Peiwen Hao. (2013). A New Laboratory Method to Characterize Gradation Segregation of
Large Stone Asphalt Mixtures. Construction and Building Materials 38: 1199-1203. doi:
http://dx.doi.org/10.1016/j.conbuildmat.2012.10.003.
Li, Chang, and Linglin Li. (2012). Criteria for Controlling Rutting of Asphalt Concrete Materials in Sloped Pavement.
Construction and Building Materials 35: 330-339. doi: http://dx.doi.org/10.1016/j.conbuildmat.2012.04.003.
Manjunath, K. R, and N. B Poornachandra Dev (2014). Design of Hot Mix Asphalt Using Bailey Method of Gradation.,
International Journal of Research in Engineering and Technology, 03 (06), 386-393, eSAT Publishing House.
Main Roads Western Australia (2015). Specification 502 Stone Mastic Asphalt. Perth, Western Australia.
Qiu, Y., and K. Lum (2006). Design and Performance of Stone Mastic Asphalt., Journal of Transportation Engineering,
132 (12), 956-963, American Society of Civil Engineers.
Shen, Shihui, and Huanan Yu (2011). Characterize Packing of Aggregate Particles for Paving Materials: Particle Size
Impact., Construction and Building Materials, 25 (3), 1362-1368, Elsevier Ltd.
Underwood, B. Shane, and Y. Richard Kim. (2013). Effect of Volumetric Factors on the Mechanical Behavior of
Asphalt Fine Aggregate Matrix and the Relationship to Asphalt Mixture Properties. Construction and Building
Materials 49: 672-681. doi: http://dx.doi.org/10.1016/j.conbuildmat.2013.08.045.
Vavrik, R. William , Gerald Huber, J. William Pine, H. Samuel Carpenter, and Robert Bailey (2002). Bailey Method for
Gradation Selection in Hot-Mix Asphalt Mixture Design., Transportation Research E-Circular No. E-C044,
Transportation Research Board, National Research Council, Washington, D.C.

112 AUSTRALIAN GEOMECHANICS VOLUME 51 NO. 2 JUNE 2016

112 AUSTRALIAN GEOMECHANICS | VOLUME 51: NO.2 JUNE 2016