This article was downloaded by: [University of Waterloo]

On: 22 October 2014, At: 18:56

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41
Mortimer Street, London W1T 3JH, UK

International Journal of Pavement Engineering

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/gpav20

An Alternative Approach for the Determination of Bulk

Specific Gravity and Permeability of Hot Mix Asphalt (HMA)
a a
Sudip Bhattacharjee & Rajib B. Mallick
a
Civil and Environmental Engineering Department , Worcester Polytechnic Institute (WPI) , Institute
Road, Worcester, MA, 01609, USA
Published online: 27 Oct 2010.

Bulk SG Permeability
To cite this article: Sudip Bhattacharjee & Rajib B. Mallick (2002) An Alternative Approach for the Determination of Bulk
Specific Gravity and Permeability of Hot Mix Asphalt (HMA), International Journal of Pavement Engineering, 3:3, 143-152, DOI:
10.1080/1029843021000067782

To link to this article: http://dx.doi.org/10.1080/1029843021000067782

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the
publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or
warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed
by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings,
demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly
in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction,
redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly
forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-
conditions
 The International Journal of Pavement Engineering, 2002 Vol. 3 (3), pp. 143–152

An Alternative Approach for the Determination of Bulk

Specific Gravity and Permeability of Hot Mix Asphalt (HMA)
SUDIP BHATTACHARJEE* and RAJIB B. MALLICK†

Civil and Environmental Engineering Department, Worcester Polytechnic Institute (WPI), Institute Road, Worcester, MA 01609, USA
Downloaded by [University of Waterloo] at 18:56 22 October 2014

(Received 19 November 2001; Received 1 July 2002)

The objective of this study was to investigate the use of an alternative method for determination of bulk
specific gravity and estimation of water permeable voids of dense graded HMA mixes. As a part of an
ongoing study on evaluation of permeability of HMA, several dense graded mixes with coarse and fine
gradations from three New England states were compacted to produce mixes at different air voids. The
bulk specific gravities of these mixes were determined using both the saturated surface dry method and
the vacuum seal method. Theoretical maximum densities of these mixes were also determined. Air
voids of the different compacted mixes were estimated from the bulk specific gravity and theoretical
maximum density values. Porosity and permeability tests were also conducted on the mixes. The results
showed that for coarse graded mixes and fine graded mixes with high air voids, the vacuum seal method
provided a better estimation of air voids in a compacted HMA mix. For coarse graded mixes, porosity
can provide a good indication of inter-connectivity of air voids in a compacted HMA mix. Analysis of
the data showed that porosity could provide an excellent indication of permeability of HMA. A porosity
of 7% seemed to be critical—mixes with porosity greater than 7% showed significantly higher
permeability than mixes with porosity lower than 7%. Charts have been provided for selection of proper
gradation and air voids to avoid mixes with excessive permeability.

Keywords: Air voids; Bulk specific gravity; Vacuum seal; Porosity; Permeability

INTRODUCTION results in percolation of large amounts of air and water, is

Air void content is the single most important property that
is used for design and construction quality control of Hot
Mix Asphalt (HMA). Generally, air void content is
determined from bulk specific gravity and theoretical
maximum density of HMA mixes. Among the different
available methods of determination of bulk specific
gravity, the saturated surface dry method, as indicated in
AASHTO T166-88 (1990) is the most widely used.
Although a good test, the method does not produce
accurate results if the air voids are very high or if very
coarse graded mixes are used. Since high air voids are
frequently encountered during design, and the use of
coarse graded mixes have increased significantly since the
introduction of Superpave mix design method, mix
designers are searching for alternative techniques for
determination of bulk specific gravity.
Excessive permeability of HMA can cause durability
problems in pavements. A high permeability, which
therefore detrimental to pavement performance. Since the
introduction of coarse graded Superpave mixes, per-
meability of HMA has become a major concern. Since
coarse graded mixes contain more interconnected voids
and hence higher percentage of water accessible voids,
there is a general concern (Brown et al., 1999) regarding
high permeability and permeability testing of mixes. Even
though in-place permeameters have been developed by
many researchers, (for example, Mallick et al., 2001) there
is a need for the development of a simple and accurate
method for predicting the permeability of mixes during
mixture design.
OBJECTIVE
The objective of this study was to investigate the use of an
alternative method for determination of bulk specific
gravity and permeability of dense graded HMA mixes.

*E-mail: ciesudip@wpi.edu
†
Corresponding author. E-mail: rajib@wpi.edu

ISSN 1029-8436 print/ISSN 1477-268X online q 2002 Taylor & Francis Ltd
DOI: 10.1080/1029843021000067782
 144 S. BHATTACHARJEE AND R.B. MALLICK
CURRENT PROCEDURES FOR DETERMINATION
OF BULK SPECIFIC GRAVITY OF COMPACTED
HMA
Bulk specific gravity is defined as the ratio of the oven dry
weight in air of a unit volume of permeable material
(including both permeable and impermeable voids normal
for the material) at a stated temperature to the weight of an
equal volume of gas free distilled water at a stated
temperature (Roberts et al., 1996). Currently used
American Association of State Transportation and High-
way Officials (AASHTO) procedures for determination of
bulk specific gravity of compacted HMA mixes are
Downloaded by [University of Waterloo] at 18:56 22 October 2014
AASHTO T166-88 (1990) (Bulk Specific Gravity of Non-
Absorptive Compacted Bituminous Mixes) and AASHTO
T275-89 (1990) (Bulk Specific Gravity of Compacted
Bituminous Mixes using Paraffin Coated Specimens).
AASHTO T166-88 is valid for measuring the bulk
specific gravity and density of non absorptive (samples,
which absorb less than 2% of water by volume)
compacted bituminous mixture (AASHTO T166-88).
The procedure involves determination of dry mass,
mass under water at 258C (778F), and the mass of a
sample at the saturated surface dry condition. The
difference between the last two masses is used to
measure the mass of an equal volume of water at 258C.
The following expression is used to compute the bulk
specific gravity:
Bulk Specific Gravity at 258C ¼ A=ðB 2 CÞ
where A ¼ mass of the dry specimen in air, B ¼ mass
of the saturated surface dry specimen in air and
C ¼ mass of the specimen in water.
The water absorption of the specimen must also be
checked in order to verify the applicability of the method.
AASHTO T275-89 should be used with samples that
contain open or interconnecting voids or absorb more than
2% of water by volume or both (AASHTO T275-89). In
this method, the mass of the HMA sample is determined
before coating it with liquid paraffin wax. The sample is
then weighed in air and under water. The bulk specific
gravity of the film-coated specimen is then computed
using the following formula:
Bulk Specific Gravity ¼ A={D 2 E 2 ðD 2 AÞ=F}
where A ¼ Mass of the dry specimen in air, D ¼ Mass of
dry coated specimen, E ¼ Mass of coated specimen under
water and F ¼ Specific Gravity of the coating as
determined at 258C.
Unfortunately, the AASHTO T275-89 test method used
for sealing of compacted asphalt samples can have poor
repeatability, high sensitivity to operator involvement and
training (Buchanan, 2000). Furthermore, there are
currently no specifications for sealing 150 mm diameter
samples. Consequently, few agencies require the use of
this method.
Drawbacks of Current Test Procedures
For open graded and coarse mixes the density results
obtained by these methods are higher than the actual
density of the asphalt specimen. Problems related to
inaccurate specific gravity measurements can have serious
and detrimental effects on design and quality control of
asphalt mixes.
Inaccurate air void contents based on erroneous specific
gravity can seriously affect the performance of the
roadway and its quality. Field cores are generally different
from laboratory prepared cores in surface texture and
thickness. In the laboratory gyratory compactor, the
mixture is confined on all surfaces. The difference in
surface roughness of these two different sampling
methods in effect produces a different degree of water
absorption and drainage during water displacement tests
which in effect reduces the reliability of the saturated
surface dry weight of the samples. This causes the
calculated densities for laboratory and field samples with
the same air content, asphalt content and density, to be
different when tested with water displacement method.
Use of Vacuum Seal Method
The vacuum seal method (Operator’s guide, 2000) can be
used to avoid these problems. The method requires the use
of a vacuum chamber capable of evacuating and sealing
asphalt cores up to 150 mm thick. This device can, in less
than two minutes, completely seal laboratory and/or field
compacted samples. The method involves limited operator
involvement and protects the samples against water
infiltration. The sample can later be removed from the seal
and used for other tests.
This method utilizes an automatic vacuum chamber
with specially designed, puncture resistant, resilient bags.
Using a 1.25 hp vacuum pump, the unit automatically
evacuates and seals the bag during the vacuum operation.
The vacuum pump is capable of pulling up to 117.91 mm
Hg (1 TORR). The bags are designed in two different sizes
to accommodate different HMA sample sizes.
The steps involved in sealing and analyzing the samples
are specified in ASTM standard D130-01 and are outlined
as follows:
. Step 1: Obtain the bag weight.
. Step 2: Obtain the dry sample weight.
. Step 3: Place the bag containing the asphalt sample
inside the vacuum chamber.
. Step 4: Close the chamber door. The vacuum pump will
start automatically and evacuate the chamber to
approximately 117.91 mm Hg.
. Step 5: In approximately two minutes, the chamber
door will open automatically. The sample is completely
sealed, ready for water displacement analysis.
. Step 6: Obtain the sealed weight of the sample.
. Step 7: Water displace the sample using a basket that is
connected to a standard scale.
 VOIDS AND PERMEABILITY OF HMA 145

. Step 8: Use the following equation to calculate the bulk TABLE I Mix information
specific gravity of the sample.
State Mix Sieve size (mm) % Passing
A
Bulk Specific Gravity ¼ CT TD 12.5 19 100
B 2 E 2 B2A
FT 12.5 93
9.5 74
4.75 43
where: A ¼ mass of dry specimen in air, g; B ¼ mass of 2.36 30
dry, sealed specimen, g; E ¼ mass of sealed specimen 1.18 22
underwater, g and FT ¼ Apparent specific gravity of 0.6 15
0.3 9
plastic sealing material at 258C (778F), provided by the 0.15 6
manufacturer. 0.075 3.6
TP 12.5 19 100
12.5 96
9.5 83
Downloaded by [University of Waterloo] at 18:56 22 October 2014

SCOPE OF THIS STUDY 4.75 48

2.36 33
1.18 23
As part of an ongoing study on evaluation of permeability 0.6 18
of HMA, several dense graded mixes with coarse and fine 0.3 12
gradations from three New England states were com- 0.15 7
0.075 3.7
pacted to produce mixes at different air voids. The bulk TW 9.5 19
specific gravities of these mixes were determined using 12.5 100
both the Saturated Surface Dry (SSD) method and the 9.5 99
4.75 71
vacuum seal method. Theoretical maximum densities of 2.36 46
these mixes were also determined. Air voids of the 1.18 33
different compacted mixes were estimated from the bulk 0.6 24
0.3 15
specific gravity and theoretical maximum density values. 0.15 8
Porosity of the mixes was also determined. Permeability 0.075 4.5
tests were conducted in the laboratory with the laboratory LW 9.5 19
12.5 100
compacted samples. Air voids and porosity data were then 9.5 99
analyzed to investigate the use of the vacuum seal method 4.75 62
for estimation of interconnectivity of air voids of mixes. 2.36 44
1.18 30
Air voids, porosity and permeability data were then 0.6 20
analyzed to investigate the use of porosity for evaluation 0.3 13
of permeability of HMA. 0.15 8
0.075 4.3
CT LW 12.5 19 100
12.5 97
TEST PLAN 9.5 75
4.75 43
2.36 31
Loose mixes from the states of New Hampshire, 1.18 21
Connecticut and Massachusetts were obtained from 0.6 14
0.3 9
HMA plants. Information for each mix is provided in 0.15 6
Table I. Both fine and coarse graded mixes were used. 0.075 3.9
Each mix was compacted to three air voids levels 5, 7 and MA MA 19.0 25 100
19 95
10%, using a Superpave gyratory compactor (AASHTO 12.5 70
TP4-97). Three samples were compacted for each target 9.5 56
air void content. The theoretical maximum density of each 4.75 39
2.36 29
mix was determined using the AASHTO T 209-99 1.18 21
procedure. The compacted samples were then tested for 0.6 16
bulk specific gravity, using both AASHTO T 166-88 and 0.3 10
0.15 7
the vacuum seal method. Void in Total Mix (VTM), which 0.075 4.3
is defined as the ratio of volume of voids to total volume of NH CPI 1 19.0 25 100
the compacted mix (Roberts et al., 1996) was determined 19 99
using the bulk specific gravity (determined by two 12.5 75
9.5 64
methods) and theoretical maximum density. 4.75 39
Effective porosity was also determined with the vacuum 2.36 26
seal method. Effective porosity is defined as the ratio of 1.18 18
0.6 12
interconnected pore volume to the total volume of the 0.3 7
medium (Bear, 1972). In this method a sample is vacuum 0.15 4
sealed inside a bag and a density, r1 is calculated by using 0.075 3
 146 S. BHATTACHARJEE AND R.B. MALLICK

TABLE I – continued

State Mix Sieve size (mm) % Passing

CPI 2 12.5 19 100

12.5 99
9.5 90
4.75 61
2.36 45
1.18 33
0.6 21
0.3 10
0.15 4
0.075 2.6
ME ME 214 19
12.5 100
9.5 96
Downloaded by [University of Waterloo] at 18:56 22 October 2014

4.75 60
2.36 41
1.18 27
0.6 17
0.3 10
0.15 7
0.075 5
VT RT 4 19 100
12.5 99
9.5 82
4.75 57
2.36 37
1.18 24
0.6 16
0.3 10
0.15 6
0.075 3.5

the method outlined in ASTM D130-01. The same sample

while under water is opened and a second density, r2 is
determined. Since the sample is under complete vacuum
prior to opening the bag, r2 will yield an apparent or
maximum density of the compacted sample. The density
r2 includes the volume due to inaccessible air voids.
 
r2 2 r1
% Effective porosity ¼ %P ¼ £ 100
r2 FIGURE 1 Test plan.

where: r1 ¼ the vacuum sealed density of compacted

sample and r2 ¼ density of the vacuum sealed sample
after opening under water.
Permeability tests were conducted for the different
mixes, using a falling head permeameter, according to
Florida DOT specification (FM 5-565). The cores and
laboratory samples were saturated before testing by
applying a vacuum under water for 15 min. The air voids,
porosity and permeability data were analyzed. The entire
test plan is shown in Fig. 1.
RESULTS
Table II shows the results of bulk specific gravity testing
with the SSD and the vacuum seal method, the estimated
air voids, porosity and permeability. In general, the SSD
bulk specific gravity of the coarse graded mixes were
difficult to obtain since the water drained out too fast from
the samples before any meaningful data can be obtained.
In the case of the 19 mm coarse graded mix from NH, it
was not possible to obtain any bulk specific gravity data
using the SSD method at 10% VTM.
A comparison of VTM obtained from SSD and vacuum
seal methods reveals that in most cases the SSD voids are
lower than the vacuum seal method voids. An explanation
of this observation is that since some water always runs
out of the sample during the SSD test prior to obtaining the
SSD mass, the volume of the sample is underestimated,
resulting in overestimation of density and underestimation
of air voids. This error becomes more significant at higher
air voids, since the potential of the water to run out should
increase with an increase in air voids. This is shown in
Fig. 2, which shows that for each mix the difference
between the SSD and the vacuum seal method air voids
increases with an increase in air voids (as obtained from
vacuum seal method). Figure 2 proves that use of the
vacuum seal method is capable of giving a better
indication of “true” air voids, compared to the SSD
 VOIDS AND PERMEABILITY OF HMA 147

TABLE II Voids and permeability (laboratory testing) of mixes

State Mix Target VTM, (%) Sample No. VTM, SSD (%) VTM, vacuum seal (%) Porosity (%) Permeability (cm/s)
CT TD 12.5 5 1 4.47 4.74 7.33 1.83E 2 03
2 5.48 6.26 8.89 2.05E 2 03
3 4.56 5.42 8.14 2.60E 2 03
7 3 6.85 8.10 10.72 6.65E 2 03
12 6.79 8.28 10.73 6.85E 2 03
13 6.53 8.37 10.95 7.47E 2 03
10 1 11.95 11.99 14.01 2.61E 2 02
2 9.06 11.33 14.04 2.54E 2 02
3 8.89 10.98 13.53 2.20E 2 02
TW 9.5 5 1 4.48 4.68 3.48 1.05E 2 04
2 5.30 5.76 3.97 1.73E 2 04
3 4.60 5.05 3.06 1.28E 2 04
7 2 6.66 7.10 5.18 3.04E 2 04
Downloaded by [University of Waterloo] at 18:56 22 October 2014

6 6.63 7.05 5.09 3.71E 2 04

9 6.97 8.94 7.52 4.38E 2 04
10 2 9.75 10.30 9.26 1.81E 2 03
5 9.93 10.51 8.94 2.02E 2 03
LW9.5 10 6 9.86 10.44 9.89 2.27E 2 03
LW 12.5 5 1 4.74 6.26 5.30 3.99E 2 04
2 4.65 5.60 3.45 6.43E 2 04
3 4.83 5.63 4.36 6.32E 2 04
7 1 7.16 8.11 8.33 2.04E 2 03
5 7.39 8.73 8.87 4.10E 2 03
6 7.46 8.29 8.40 1.84E 2 03
10 4 9.71 12.18 12.70 1.30E 2 02
6 10.39 13.17 14.01 2.50E 2 02
7 9.78 11.24 11.91 1.29E 2 02
TP12.5 5 2 4.98 5.26 3.88 2.95E 2 04
3 4.95 5.12 3.17 3.08E 2 04
4 4.99 5.45 4.77 5.05E 2 04
7 1 7.53 7.79 6.66 1.180E 2 04
2 7.54 8.12 6.42 8.70E 2 04
7 6.95 7.27 5.93 9.08E 2 04
10 1 9.92 9.86 8.38 3.180E 2 03
2 9.72 10.86 8.03 1.830E 2 03
3 10.17 10.88 10.52 1.560E 2 03
LW9.5 5 1 5.40 5.76 4.08 3.66E 2 04
3 4.79 5.28 4.26 4.76E 2 04
4 4.60 5.00 3.74 3.69E 2 04
7 1 7.05 7.61 6.42 9.040E 2 04
2 7.15 7.70 6.90 7.480E 2 04
3 6.88 7.46 6.40 6.170E 2 04
10 7 10.39 11.53 11.95 4.770E 2 03
8 10.76 11.53 12.09 7.760E 2 03
9 10.46 11.51 11.85 6.830E 2 03
MA MA 19.0 5 2 4.97 5.78 4.87 3.18E 2 04
3 5.35 5.95 5.39 2.52E 2 03
4 5.43 5.99 4.71 5.00E 2 04
7 2 7.27 8.19 8.38 8.196E 2 04
3 7.16 8.22 8.70 1.659E 2 03
4 7.29 8.41 5.93 1.994E 2 03
10 2 10.17 11.31 9.87 4.389E 2 03
3 10.19 11.21 12.14 6.011E 2 03
4 9.98 10.88 11.84 6.684E 2 03
NH CPI 19.0 5 1 5.28 6.56 7.26 2.43E 2 03
3 5.37 6.68 7.64 2.96E 2 03
5 5.25 6.63 7.89 5.26E 2 03
7 1 7.14 9.60 10.88 2.03E 2 02
2 6.84 9.20 10.80 9.47E 2 03
10%_01 7.19 11.44 11.46 2.43E 2 02
10 6 Not available 12.18 Not available 3.15E 2 02
8 Not available 12.11 12.26 2.25E 2 02
9 Not available 12.40 12.01 2.51E 2 02
CPI 12.5 5 1 5.13 5.14 5.23 3.65E 2 05
3 5.12 5.00 5.16 7.41E 2 05
4 5.01 4.61 4.97 7.41E 2 05
7 1 7.25 7.18 7.28 2.89E 2 04
5 7.17 6.94 6.90 3.12E 2 04
6 7.33 8.45 8.49 3.79E 2 04
10 2 9.91 10.22 10.67 3.82E 2 03
3 9.90 10.66 11.51 4.50E 2 03
5 10.15 10.44 10.54 3.99E 2 03
 148 S. BHATTACHARJEE AND R.B. MALLICK

TABLE II – continued

State Mix Target VTM, (%) Sample No. VTM, SSD (%) VTM, vacuum seal (%) Porosity (%) Permeability (cm/s)

ME ME 214 5 1 5.32 Not available Not available 2.784E 2 04

2 4.94 7.64 5.071 2.227E 2 04
3 5.06 5.75 4.005 1.371E 2 04
7 1 6.58 6.78 5.127 6.786E 2 04
2 7.25 8.00 -126.773 7.281E 2 04
3 6.65 7.46 6.658 8.513E 2 04
10 6 10.31 Not available Not available 4.273E 2 03
8 9.63 10.56 10.326 4.185E 2 03
9 9.95 12.16 11.781 3.997E 2 03
VT Rt 4 5 1 4.63 5.66 5.299 6.057E 2 04
3 5.23 5.05 4.716 4.067E 2 04
5 5.03 6.03 4.227 5.770E 2 04
7 1 6.99 7.99 7.570 1.489E 2 03
Downloaded by [University of Waterloo] at 18:56 22 October 2014

2 7.06 7.90 7.504 1.153E 2 03

3 7.17 8.26 7.386 1.912E 2 03
10 6 9.79 11.63 11.411 7.034E 2 03
8 9.64 11.22 11.005 6.210E 2 03
9 9.09 11.22 11.208 6.960E 2 03

VTM—voids in total mix, SSD—saturated surface dry method.

Permeability – Coefficient of permeability.

method, particularly with compacted mixes at relatively
high air voids (such as 10%).
Figure 3a – c, shows the difference between vacuum seal
method VTM and porosity data for 9.5,12.5, 19 mm
NMAS mixes, respectively. The plots show that the
difference is positive and very low at lower air voids. At
high air voids (10% VTM), for the fine mixes, the
difference is positive, indicating the presence of
inaccessible air voids. For the coarse mixes, at high air
voids, the difference is either very small or negative,
indicating the presence of highly inter-connected air
voids, which are accessible to water. Negative differences
can also indicate a less than optimum asphalt coating that
would ultimately allow water to penetrate into the
aggregate grain. Therefore, the vacuum seal method
seems to be a good method for getting an indication of
FIGURE 2 Plot of VTM versus difference of vacuum seal method VTM
and SSD VTM.
percent of water accessible air voids in HMA, especially
for coarse graded mixes.
Analysis of Laboratory Permeability Testing
(on Laboratory Compacted Samples) Data
Water flows through accessible voids or pore spaces in a
pavement. Hence, the rate of flow must be related to the
amount of water accessible voids, or effective porosity, in
some way. Therefore, the coefficient of permeability must
be a function of porosity. Table II shows the effective
porosity versus permeability data, as obtained from
laboratory testing. The best model for defining the
relationship between effective porosity and permeability
seems to be one of exponential in nature, as shown in Fig. 4
(considering data from 9.5 to 12.5 mm NMAS mixes
only). Figure 4 shows two relationships: VTM versus
permeability ðR 2 ¼ 0:55Þ and effective porosity
versus permeability ðR 2 ¼ 0:80Þ: The effective porosity
versus permeability model is significantly better than the
VTM versus permeability model. If one considers
the effective porosity versus permeability model, then the
critical porosity corresponding to a critical permeability
can be determined. Since, in general, permeability data for
HMA is highly variable, and the nature of the
permeability –porosity plot is exponential, it is logical to
consider the value of permeability in terms of 10X, rather
than in terms of exact value. As recommended by Florida
Department of Transportation (DOT) researchers (Chou-
bane et al., 1998), considering a critical permeability of
1023 cm/s, a critical porosity of 7% is determined (from
Fig. 4). Therefore, mix design samples can be checked for
permeability potential by conducting porosity tests on
samples compacted to construction voids (as determined
by SSD method), and a maximum porosity value of 7%
can be recommended.
 VOIDS AND PERMEABILITY OF HMA 149
Downloaded by [University of Waterloo] at 18:56 22 October 2014

FIGURE 3 (a) Plot of vacuum seal VTM versus difference between vacuum seal VTM and porosity for 9.5 mm mixes; (b) Plot of vacuum seal VTM
versus difference between vacuum seal VTM and porosity for 12.5 mm mixes; (c) Plot of vacuum seal VTM versus difference between vacuum seal
VTM and porosity for 19 mm mixes.

An important question is that how can one estimate the

FIGURE 4 Effective porosity versus permeability model (from results
of laboratory permeability testing) (a weaker air voids versus
permeability model is also shown).
porosity from knowledge of mix gradation and air voids?
To answer this question multiple regression analysis was
conducted with porosity and aggregate gradation data. The
aggregate gradation was characterized with the help of
several parameters, such as percent passing the 2.36 mm
sieve (PP2.36), fineness modulus and coefficient of
curvature. The results of multiple regression analysis,
presented in Table III, show that an excellent model can be
defined for relating porosity to air voids and percent
passing the 2.36 mm sieve (PP2.36).
Next, considering the defined model relating percent air
voids, percent passing the 2.36 mm sieve and porosity, a
set of curves were plotted for different PP2.36 and
different air voids, as shown in Fig. 5. Considering a
critical porosity of 7% (as determined earlier, correspond-
ing to a critical permeability of 1023 cm/s), critical air
voids (as determined by SSD bulk specific gravity and
 150 S. BHATTACHARJEE AND R.B. MALLICK

TABLE III Results of multiple regression analysis with mix gradation, air voids and porosity

Regression summary
Porosity vs. 2 independents
Row exclusion: perm #1_12.5_9.5.svd
Count 68
Num. Missing 3
jRj 0.897
R squared 0.804
Adjusted R squared 0.798
RMS residual 1.406
ANOVA table
Porosity vs. 2 independents
Row exclusion: perm #1_12.5_9.5.s.vd
DF Sum of squares Mean square F-value P-value
Regression 2 526.694 263.347 133.281 ,0.0001
Downloaded by [University of Waterloo] at 18:56 22 October 2014

Residual 65 128.432 1.976

Total 67 655.126
Regression Coefficients
Porosity vs. 2 independents
Row exclusion: perm #1_12.5_9.5.svd
Coefficient Std. error Std. coefficient t-value P-value
Intercept 3.472 1.225 3.472 2.834 0.0061
Air voids, SSD 1.260 0.080 0.861 15.686 ,0.0001
PP2.36 20.127 0.028 20.251 24.575 ,0.0001

theoretical maximum density) can be determined. A list of
critical air voids, for mixes with different percent passing
the 2.36 mm sieve, is shown in Table IV. The critical air
voids range from 5 to 7.5 corresponding to PP2.36 of
25 to 45. One can use this table to specify maximum
construction air voids for a specific mix.
From Fig. 5 it can be noted that for a mix with
approximately 45% passing the 2.36 mm sieve, a porosity
of 7 per percent corresponds to VTM of 7%. Most
experiences from the past (in the pre-Superpave era) has
been with mixes with 40 –45% passing the 2.36 mm sieve,
and a VTM of 7% has often been recommended and used,
without any significant permeability or durability (related
to excessive aging) problem. Therefore, the selection of
7% porosity as a critical porosity seems to be well
justified.
From Fig. 6 one can determine the critical PP2.36 value
for any specific air voids. Table V shows a list of critical
PP2.36 for a range of air voids. One can use the values
from this table to select suitable gradation(s) during mix
design for a specific construction air voids. For example, if
it is known that realistically air voids below 6% cannot be
achieved, and then one can select a specific PP2.36

FIGURE 5 Plot of air voids versus porosity for different percent passing
the 2.36 mm sieve.

TABLE IV Critical air voids for mixes with different gradations

Percent passing the 2.36 mm sieve Critical air voids (%)

25 5
30 6
35 6.5
40 7
45 7.5 FIGURE 6 Plot of percent passing the 2.36 mm sieve versus porosity
for different air voids.
 VOIDS AND PERMEABILITY OF HMA 151

TABLE V List of critical percent passing the 2.36 mm sieve for specific
air voids
Air Voids Allowable percent passing the 2.36 mm sieve
5 .25
6 .31
7 .41
8 .45
(greater than 31%) such that the porosity of the resulting
mix is below 7% air voids.
A question that arises from the analysis of the data is
that which parameter should be used during mix design to
Downloaded by [University of Waterloo] at 18:56 22 October 2014
relationship with permeability, perhaps porosity is a good
mix design parameter candidate also. This argument
becomes stronger when one considers the coefficient of
variation of the permeability and the porosity tests.
Table VI shows the coefficient of variation (CV) for each
set of samples as well as average CV for porosity and
permeability data. The average CV for permeability
testing is about three times the average CV of porosity.
Hence, as a regular test procedure, because of better
repeatability, porosity seems to be more appropriate than
permeability. Also, determination of porosity is simple
and rapid as compared to laboratory permeability test.
Since asphalt laboratories can use the vacuum sealing

prevent mixes from being excessively permeable—is it device to determine the bulk specific gravity of
the permeability value or is it the porosity value? compacted HMA, the determination of porosity will be
Obviously, measuring permeability is a more direct just an additional step when testing for bulk specific
approach. However, since porosity shows a very good gravity.

TABLE VI Coefficient of variation of porosity and permeability testing

Target VTM Porosity Coefficient of variation, porosity Coefficient of permeability Coefficient of variation, coefficient of permeability
(%) (%) (%) (cm/s) (%)
5 7.3 9.59 0.001826 18.52
8.8 0.002047
8.13 0.002602
7 10.7 1.21 0.006647 6.14
10.7 0.006849
10.9 0.007471
10 14.0 2.09 0.026135 9.05
14.0 0.025357
13.5 0.021963
5 3.8 20.38 0.000294 31.96
3.1 0.000307
4.7 0.000505
7 6.6 5.89 0.000118 70.49
6.4 0.00087
5.9 0.000908
10 8.3 15.02 0.00318 39.63
8.0 0.00183
10.5 0.00156
5 3.4 12.97 0.000104 25.61
3.9 0.000172
3.0 0.000128
7 5.1 23.20 0.000304 18.02
5.0 0.000371
7.5 0.000438
10 9.2 5.13 0.001808 11.43
8.9 0.002024
9.8 0.002273
5 4.0 6.60 0.000366 15.46
4.2 0.000475
3.7 0.000368
7 6.4 4.29 0.000904 18.99
6.8 0.000748
6.4 0.000617
10 11.9 1.03 0.00477 23.71
12.0 0.00776
11.8 0.00683
5 5.3 21.15 0.000399 24.70
3.4 0.000643
4.3 0.000632
7 8.3 3.46 0.002037 46.95
8.8 0.004096
8.3 0.001842
10 12.7 8.26 0.013027 41.15
14.0 0.025045
11.90 0.012860
Average 9.35 26.79
 152 S. BHATTACHARJEE AND R.B. MALLICK
CONCLUSIONS AND RECOMMENDATIONS
Based on the results obtained in this study the following
conclusions can be made:
1. For coarse graded mixes and fine graded mixes with
high air voids, the vacuum seal method provides a
better estimation of air voids in a compacted HMA
mix.
2. Porosity provides a fast and accurate method of
determining permeability of dense graded mixes.
3. A critical porosity of 7% corresponds to a critical
permeability of 1023 cm/s for coarse grade HMA.
Downloaded by [University of Waterloo] at 18:56 22 October 2014
4. Porosity is significantly affected by percent of material
passing 2.36 mm sieve.
Based on the above conclusions the following
recommendations are made:
1. Use of the vacuum seal method should be further
explored for routine bulk specific gravity testing of
HMA mixes.
2. Since the vacuum seal method leaves the samples
intact after testing, this method should be used for bulk
specific gravity testing of cold recycled mixes.
3. Use porosity as an indicator of mix permeability.
4. Estimate porosity from mix gradation at specific air
voids and use the estimated porosity to select
gradation or desirable construction air voids.
5. Conduct porosity test on mix design samples at
construction air voids and make sure that the critical
porosity is not exceeded.
DISCLAIMER
This paper, prepared in cooperation with the New England
Transportation Consortium, does not constitute a standard,
specification or regulation. The contents of this paper
reflect the views of the authors who are responsible for the
facts and the accuracy of the data presented herein. The
contents do not necessarily reflect the views of the New
England Transportation Consortium or the Federal High-
way Administration.
Acknowledgements
The data reported in this paper has been obtained as part of
a study, which is funded by the New England
Transportation Consortium. The authors are grateful to
Mr P.S. Kandhal, Dr Elton Ray Brown and Dr Ali
Regimand for their helpful comments and Matthew Teto
and John Gould for helping in running the tests.
References
AASHTO T166-88 (1990). Standard Method of Test for Bulk Specific
Gravity of Compacted Bituminous Mixes, Using Saturated Surface-
Dry Specimens. Standard Specifications for Transportation Materials
and Methods of Sampling and Testing, Part II Tests, 15th Ed.,
Adopted by the American Association of State Highway and
Transportation Officials.
AASHTO T275-89 (1990). Standard Method of Test for Bulk Specific
Gravity of Compacted Bituminous Mixes. Using Paraffin Coated
Specimens, Standard Specifications for Transportation Materials and
Methods of Sampling and Testing, Part II Tests. 15th Ed., Adopted by
the American Association of State Highway and Transportation
Officials.
AASHTO T209-99. Theoretical maximum specific gravity and density of
bituminous paving mixtures. Standard Specifications for Transpor-
tation Materials and Methods of Sampling and Testing, Part II Tests,
15th Ed., Adopted by the American Association of State Highway
and Transportation Officials.
AASHTO TP4-97. Standard method for preparing and determining the
density of Hot Mix Asphalt specimens by means of SHRP Gyratory
Compactor. AASHTO provisional standards. Approved for publi-
cation by the AASHTO sub committee on materials.
ASTM D130-01. Standard Test Method for Bulk Specific Gravity and
Density of Compacted Bituminous Mixes Using Automatic Vacuum
Sealing Method. American Society for Testing and Materials.
Bear, J. (1972) Dynamics of Fluid in Porous Media (Dover Publication,
Inc., New York).
Brown, R.E., Decker, D., Mallick, R.B. and Bukowski, J. (1999)
Superpave Construction Issues and Early Performance Evaluations.
Journal of Asphalt Paving Technologists 68, 613 –660.
Buchanan, M. Shane (2000) An evaluation of selected methods for
measuring the bulk specific gravity of compacted hot mix asphalt
(HMA) mixes. Journal of Association of Asphalt Paving Technol-
ogists 69, 2000.
Choubane, B., Page, G.C. and Musselman, J.A. (1998) Investigation of
water permeability of coarse graded superpave pavements. Journal of
the Asphalt Paving Technologists 67, 254 –276.
Florida DOT Specification (FM 5-565).
Mallick, R.B., Cooley, L.A., Teto, M. and Bradbury, R. (2001)
Development of a simple test for evaluation of in-place permeability
of asphalt mixes. International Journal of Pavement Engineering.
2(2), 67– 83.
Operator’s guide, Copyright December 2000, version 8, CoreLoke,
Instrotek, Incorporated.
Roberts, F.L., Kandhal, P.S., Brown, E.R., Lee, D. and Kennedy, T.
(1996) Hot Mix Asphalt Materials, Mixture Design and Construction
(NAPA Education Foundation, Lanham, MA).