1

Airfield and Highway Pavements 2015 © ASCE 2015 86

Moisture Susceptibility of Superpave Mixtures with Varying Binder Contents

Kiran Kumar Uppu1; M. Hossain, Ph.D., P.E.1; Lon S. Ingram, P.E.2; and Rick
Kreider, P.E.3
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

1
Department of Civil Engineering, Kansas State University, 2118 Fielder Hall,
Manhattan, KS 66506. E-mail: mustak@ksu.edu
2
Consultant. E-mail: lon.i@cox.net
3
Bureau of Research, Kansas Department of Transportation, 2300 Van Buren,
Topeka, KS 66611. E-mail: RickK@ksdot.org

Abstract
The process control for the hot-mix asphalt mixtures (HMA) by the Kansas
Department of Transportation (KDOT) requires each lot of HMA be produced within
±0.6% of the design binder content. Recent reviews indicate that HMA lots are being
produced consistently with lower binder contents. This situation has been exacerbated
by the use of higher proportion of Reclaimed Asphalt Pavement (RAP) materials in
the HMA mixtures. This project evaluated two moisture susceptibility test methods to
assess the effect of asphalt content on the moisture resistance of HMA. Two different
commonly used mixtures for overlaying and four varying asphalt contents, optimum
and lower, were selected. The Hamburg Wheel Tracking Device (HWTD) test was
used to predict moisture damage potential of these mixes. All specimens tested were
prepared with the Superpave gyratory compacter. Results indicate that within the
tested range, deformation decreases as the binder content decreases. This trend was
verified by studying the correlation of the number of wheel passes with the asphalt
binder film thickness.

© ASCE

Airfield and Highway Pavements 2015

 2
Airfield and Highway Pavements 2015 © ASCE 2015 87

INTRODUCTION
Binder content is an important mix design parameter for Superpave pavements.
The role of proper asphalt amount in determining long-term performance of hot-mix
asphalt (HMA) mixtures is well known. It affects mixture stiffness, strength, durability,
fatigue life, raveling, rutting, and moisture damage. Insufficient binder in the HMA mix
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

can lead to high permeability, high air voids, and thin asphalt coatings around the
aggregates which will cause durability problems (Kandhal et al. 1998). On the other
hand, excessive asphalt though durable and flexible, but may cause flushing and low mix
stability. A mixture is said to be durable when it offers long-term resistance to weathering
and aging, and provides good performance without abnormal raveling and cracking of the
paved surface (Kumar and Goetz 1977). The process control for the hot-mix asphalt
mixtures (HMA) by the Kansas Department of Transportation (KDOT) requires that each
lot of HMA be produced within ±0.6% of the design binder content in the job-mix
formula. Recent reviews indicate that HMA lots in Kansas are being produced
consistently with lower asphalt contents than that in the job-mix formula (Gedafa et al.
2011). This situation has been exacerbated by the use of higher proportion of Reclaimed
Asphalt Pavement (RAP) materials in the HMA mixtures. These drier mixtures are
thought to be moisture susceptible to moisture due to thinner asphalt films. KDOT is
considering use of the Hamburg Wheel Tracking Device (HWTD) test for evaluating
moisture susceptibility of Superpave mixtures with reclaimed asphalt pavement (RAP). It
is to be noted that HWTD test is not necessary a moisture susceptibility test. It is actually
a torture test that subjects the sample to extreme conditions not encountered in the field.
However, it is accepted that a mixture that passes the established test criterion (in terms
of number of passes and/or maximum depression or rut depth), it will be resist rutting
and/or stripping in the field.
OBJECTIVE
The main objective of this study was to investigate the moisture resistance of two
Superpave mixtures with reclaimed asphalt pavement (RAP) and varying asphalt contents
using the Hamburg Wheel Tracking Device (HWTD) test. The mixtures were sampled
from two different projects in Kansas.

HAMBURG WHEEL TRACKING DEVICE TEST

Hamburg wheel tracking device (HWTD) evaluates combined effects of rutting
and moisture susceptibility of HMA mixtures (Brown et al. 2009). The Hamburg wheel
tracking device, manufactured by PMW, Inc. of Salina, Kansas, was used in this study.
This device can test two specimens simultaneously. The device is operated by rolling a
pair of steel wheels across surface of specimens submerged in a water bath held at 50oC.
The wheels have a diameter of 204 mm (8 inches) and width of 47 mm (1.85 inches). The
device operates at approximately 50 wheel passes/min and the load applied by each
wheel is approximately 705±22 N (158±5 lbs). Specimens used in this test were
compacted to 7±1 percent air voids using a Superpave gyratory compactor. The
specimens were 150 mm (6 inches) in diameter and 62 mm (2.4 inches) in height. Rut

© ASCE

Airfield and Highway Pavements 2015

 3
Airfield and Highway Pavements 2015 © ASCE 2015 88

depth was measured automatically and continuously at 11 different points along the
wheel path of each sample with a linear variable differential transformer (LVDT) with an
accuracy of 0.01 mm (0.0004 inch). HWTD automatically ends the test if the preset
number of cycles is reached or if the rut depth measured by the LVDTs reaches a value of
20 mm (0.8 inch) for an individual specimen. The maximum rut depth versus number of
cycles is plotted to obtain a typical curve which is shown in Figure 1. The maximum rut
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

depth value anywhere in the specimen was taken following the Texas Department of
Transportation (TxDOT) practices (TxDOT 2009).
The main parameters obtained from the plot are rut depth, average number of
wheel passes, creep slope, stripping slope, stripping inflection point, and post-compaction
consolidation. Post-compaction consolidation is the deformation (mm) at 1,000 wheel
passes. Creep slope is the inverse rate of deformation (wheel passes per 1-mm rut depth)
in the linear region of the plot between the post-compaction consolidation and the
stripping inflection point. Creep slope is used to measure rutting susceptibility due to
mechanisms other than moisture damage. The stripping inflection point and stripping
slope are used to measure moisture damage. The stripping inflection point is the number
of wheel passes at the intersection of the creep slope and stripping slope. The stripping
slope is the inverse rate of deformation (wheel passes per 1-mm rut depth) after the
stripping inflection point (Brown et al. 2009).

Figure 1. Typical HWTD test plot showing test output parameters (after
Brown et al. 2009).

PROJECT SELECTION

Two different Superpave mixture types were selected for this study. The mixtures
represent typical 9.5 mm (SR-9.5A) and 12.5 mm (SR-12.5A) Nominal Maximum Aggregate
Size (NMAS) Superpave mixes with Reclaimed Asphalt Pavement (RAP). These mixtures had
been used in past construction seasons in Kansas.

© ASCE

Airfield and Highway Pavements 2015

 4
Airfield and Highway Pavements 2015 © ASCE 2015 89

Aggregates and Binder Sources

SR-12.5A and SR-9.5A mixtures were collected from two different projects in Kansas.
The binder used was PG 58-28 for all mix types. Liquid antistripping additive (Arr Maz HP+)
was also used in all mixtures.
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

Gradations and Blending

Figures 2(a) and 2(b) show the combined aggregate gradations of the mixtures, SR-9.5A
and SR-12.5A, respectively. The combined gradations pass over the maximum density line in the
sand sizes. Thus these gradations are finer. Both mixes have 25% RAP materials. SR-12.5A also
has 15% 12.5-mm crushed limestone, 20% sand, 5% limestone screening, and 35% crushed
gravel. SR-9.5A has 17% 9.5-mm crushed limestone, 12% limestone screening, 11% finer
limestone screening, and 35% sand in addition to RAP.

Experimental Design Matrix

The only variable that changed was the mixture asphalt content. The HWTD tests were
conducted on specimens prepared at four different asphalt contents, starting from the design
asphalt content and decrements of two tenths (0.2%) of a percent each time. In order to expedite
the statistical analysis of data, three sets of HWTD specimens (four plugs per set) were prepared
at each asphalt content. The design matrix is presented in Table 1.

© ASCE

Airfield and Highway Pavements 2015

 5
Airfield and Highway Pavements 2015 © ASCE 2015 90

( SR-9.5A
(a) A
100
90
80
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

70
Percent Retained

60
50
40
30
20
10
0
0.075 0.3 0.6 1.18 2..36 4.755 9.5 12.5 19
S
Sieve Size in mm
m
Maax Density Line Co
ombined Gradaation

b) SR-12.5A
(b A
10
00

9
90

8
80

7
70
Percent Retained

6
60

5
50

4
40

3
30

2
20

1
10

0
0.075
5 0.3 0.6 1.18 2.36 4.75 9.5
5 12.5
Sieve
e size in mm
M Density
Max Combined…

Fig
gure 2. 0.45--power charrt of Superp
pave mixturres studied.

© ASCE

Airfield and Highway Pavements 2015

 6
Airfield and Highway Pavements 2015 © ASCE 2015 91

Table 1 Experimental design matrix

SR-9.5A SR-12.5A
Design binder content=5.54% Design binder content=5.3%
Additive: Arr Maz HP+ (0.6%) Additive: Arr MazHP+ (0.45%)
5.54% 5.34% 5.14% 4.94% 5.3% 5.1% 4.9% 4.7%
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

3 sets for each binder content 3 sets for each binder content
(3×4=12 plugs) (3×4=12 plugs)
Air voids=7±1% Air voids=7±1%

RESULTS AND ANALYSIS

All specimens were compacted to 7±1 air voids and tested in wet condition. In general,
the HWTD specimens are subjected to 20,000 wheel passes or a maximum rut depth of 12.5 mm
(TxDOT) or 20 mm (Colorado DOT), whichever comes first. However, in this study, specimens
were subjected to unlimited wheel passes or 20-mm rut depth, whichever came first. The only
variable in the study was binder content (%). For each binder content, three replicate sets (each
set with four plugs) were fabricated and tested. Average number of wheel passes and
corresponding rut depths are tabulated in Table 2 for the SR-9.5A mixture.

Table 2 shows the average number of wheel passes is lower for the design binder content
when compared to the number of wheel passes for the drier mixes. The highest average number
of wheel passes recorded was 11,861 when 4.94% of asphalt (0.6 % below design binder
content) was used. The lowest number of wheel passes recorded was 5,087 when 5.54% of
asphalt (design binder content) was used. For specimens SB-2,3 and SB-4,5, the average number
of wheel passes was too high when compared to the other specimens in the same subset (samples
with 5.34 % AC), thus they were discarded and not taken into consideration while calculating the
average number of wheel passes since these were proven to be outliers by conducting influence
statistics such as Cook’s Distance, Dffits and Rstudent were calculated (SAS 9.3 User Guide,
2011). Details can be found in somewhere else (Uppu, 2012).

From Table 3, we can see that performance of SR-12.5A was better at lower binder
content i.e., below the design binder content. As the binder content decreased, the average
number of wheel passes increased. The average number of wheel passes increased from 5,380 to
40,244 when the binder content was decreased from 5.3% to 4.7%.

The trends in the results of decreasing wheel passes in the HWDT test with higher asphalt
content were also studied using the asphalt film thickness of both mixtures (Uppu, 2012). The
results of linear regressions show that the number of wheel passes decreases as the asphalt film
thickness increases. Trend line showing the relationship between these two parameterd has a
coefficient of determination (R2) value of 0.74 and 0.81 for SR-9.5A and SR-12A, respectively.

© ASCE

Airfield and Highway Pavements 2015

 7
Airfield and Highway Pavements 2015 © ASCE 2015 92

Table 2 Summary of Hamburg Wheel Tracking Device test results for SR-9.5A mixture
Asphalt Virgin asphalt No of Rut depth
Sample ID
content (%) added (%) passes in mm
SA-1,SA-4 5,759 20
SA-2,SA-3 4,789 20
5.54 4.11
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

SA-6,SA-7 2,250 20
(Design Asphalt
SA-9,SA-10 3,700 20
Content)
SA-11,SA-12 7,433 20
SA-13,SA-15 6,589 20
Average 5.54 4.11 5,087 20
SB-2,SB-3 20,855* 20
SB-4,SB-5 24,187* 20
SB-6,SB-7 3.91 8,367 20
5.34
SB-8,SB-11 9,091 20
SB-12,SB-13 9,450 20
SB-14,SB-15 3.91 11,621 20
Average 5.34 3.91 9,632 20
SC-3,SC-4 8,867 20
SC-5,SC-6 11,689 20
SC-7,SC-8 3.71 12,679 20
5.14
SC-9,SC-10 13,033 20
SC-11,SC-14 9,217 20
SC-13,SC-15 9,649 20
Average 5.14 3.71 10,856 20
SD-2,SD-4 11,547 20
SD-3,SD-5 11,049 20
SD-6,SD-14 3.51 9,550 20
4.94
SD-7,SD-12 10,903 20
SD-8,SD-15 12,091 20
SD-9,SD-10 16,023 20
Average 4.94 3.51 11,861 20
* statistically identified outliers

© ASCE

Airfield and Highway Pavements 2015

 8
Airfield and Highway Pavements 2015 © ASCE 2015 93

Table 3 Summary of Hamburg Wheel Tracking Device test results for SR-12.5A mixture
Asphalt % Virgin No of Rut depth
Sample ID
content (%) Asphalt added passes in mm
RA-2,RA-3 15,723* 20
RA-4,RA-5 5.3 26,211* 20
RA-7,RA-6 (Design 4,113 20
4.14
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

RA-9,RA-10 asphalt 4,583 20

RA-12,RA-13 content) 5,291 20
RA-14,RA-15 7,533 20
Average 5.3 4.14 5,380 20
RB-1,RB-3 7,127 20
RB-4,RB-5 11,347 20
RB-6,RB-9 14,653 20
5.1 3.94
RB-8,RB-10 12,621 20
RB-11,RB-13 11,967 20
RB-14,RB-15 25,563* 20
Average 5.1 3.94 11,543 20
RC-1,RC-4 8,373* 20
RC-2,RC-3 15,401 20
RC-6,RC-10 29,541* 20
4.9 3.74
RC-8,RC-9 26,893* 20
RC-11,RC-13 16,637 20
RC-12,RC-14 18,519 20
Average 4.9 3.74 16,852 20
RD-1,RD-4 19,125* 20
RD-2,RD-3 18,355* 20
RD-6,RD-9 42,335 20
4.7 3.54
RD-7,RD-8 38,153 20
RD-11,RD-13 25,650* 13.3*
RD-14,RD-15 25,650* 6.1*
Average 4.7 3.54 40,244 20
*statistically identified outliers

When testing specimens RD-11, 13 and RD-14 and 15, the test stopped due to technical
error in the HWTD machine. Thus the final number of wheel passes could not be determined but
the test still yielded valuable information in the form of creep slope, stripping slope, and
stripping inflection points, if any.

From Figure 3, we can clearly see that for both mixtures the lowest number of wheel
passes was recorded when design asphalt content (Pb) was used, and the highest number of wheel
passes was recorded when the lowest binder content was used in the mixture (Pb-0.6%), where
Pb is the design asphalt content. There was a large variation in the average number of wheel
passes for the SR-12.5A mixture when compared to the SR-9.5A mixture. The SR-12.5A
mixture seems to be affected by variation in asphalt content more than the SR-9.5A mixture.

© ASCE

Airfield and Highway Pavements 2015

 9
Airfield and Highway Pavements 2015 © ASCE 2015 94

SR-9.5A NMAS SR-12.5A NMAS

50000

No of Wheel Passes
40000
30000
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

20000
10000
0
Pb Pb-0.2 Pb-0.4 Pb-0.6
Asphalt content (%)

Figure 3. Effect of asphalt content on average number of wheel passes.

Hamburg Wheel Tracking Device Test Output Parameters (Creep slope, Stripping Slope
and Stripping Inflection Point)

The performance of the mixtures can be better studied with the HWTD output
parameters. Figures 4(a) and 4(b) show creep slopes, stripping slopes, and stripping inflection
points for the SR-9.5A and SR-12.5A mixture, respectively.

From the number of wheel passes data, we concluded that mixtures performed better in
HWTD tests at asphalt contents lower than the design asphalt contents. This also can be affirmed
from the HWTD parameters for Figures 4(a) and 4(b). The creep slope and stripping inflection
points also increased with a decrease in asphalt content indicating the performance of the mixture
was better for mixes with lower asphalt content. HWTD tests are known to be affected by the
binder grade. The increase may have happened because of the aged RAP binders in the mixtures
and use of a liquid anti-stripping agent. The anti-stripping agent degrades the aged binder less.

CONCLUSIONS
The main objective of this research study was to investigate the moisture resistance of Superpave
HMA mixtures with varying asphalt content especially with RAP in it with the Hamburg Wheel
Tracking Device (HWTD) tests. Based on the results obtained from the two mixtures, the
following conclusions can be drawn:
• Binder content significantly affects the response of HMA mixtures in the HWTD
tests.
• For two mixtures tested, the number of wheel passes, creep slope, and stripping
inflection point was higher on the dry side of the design asphalt content.
• For both mixtures, the number of wheel passes, creep slope, and stripping inflection
point increased as the asphalt content decreased.

© ASCE

Airfield and Highway Pavements 2015

 10
Airfield and Highway Pavements 2015 © ASCE 2015 95

• These trends in results were verified by studying the correlation of the number of
wheel passes with the asphalt binder film thickness.

(a) SR-9.5A
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

14000
12000
No of Wheel Passes

10000
8000
6000
4000
2000
0
5.54 5.34 5.14 4.94
Asphalt Content (%)
Average No of wheel passes Average Stripping Inflection Point
Average Creep Slope Average Stripping Slope

(b) SR-12.5 A
40000
35000
No of Wheel Passes

30000
25000
20000
15000
10000
5000
0
5.3 5.1 4.9 4.7
Asphalt Content (%)
Average No of wheel passes Average Stripping Inflection Point
Average Creep Slope Average Stripping Slope

Figure 4. Effect of varying asphalt content on HWTD test parameters.

ACKNOWLEDGEMENTS
The authors would like to acknowledge the Kansas Department of Transportation for sponsoring
this study under its Kansas Transportation and New Developments (K-TRAN) program.

© ASCE

Airfield and Highway Pavements 2015

 11
Airfield and Highway Pavements 2015 © ASCE 2015 96

REFERENCES
Brown, E., P. S. Kandhal, F.L. Roberts, Y.R. Kim, D. Lee, and T.W. Kennedy (2009). Hot Mix
Asphalt Materials, Mixture Design, and Construction. Third Edition, NAPA Research
and Education Foundation, Lanham, Maryland.
Downloaded from ascelibrary.org by Universidad Nacional Autonoma de Mexico on 08/08/16. Copyright ASCE. For personal use only; all rights reserved.

Gedafa, D.S., M. Hossain, L.S. Ingram, and R. Kreider (2011) Performance-Related

Specification for Superpave Pavements. In Transportation Research Record: Journal of
Transportation Research Board, No. 2228, Washington, D.C., pp. 78-86.

Kandhal, P. S., and S. Cross (1993). Effect of Aggregate Gradation on Measured Asphalt
Content. In Transportation Research Record: Journal of the Transportation Research
Board, No. 1417, Washington, D.C., 1993, pp. 21-28.

Kandhal, P. S., K. Y. Foo, and R. B. Mallick (1998). A Critical Review of VMA Requirements
in Superpave. In Transportation Research Record: Journal of the Transportation Research
Board, No. 1609, Washington, D.C.

Kumar, A., and W. H. Goetz (1977). Asphalt Hardening as Affected by Film Thickness, Voids
and Permeability in Asphaltic Mixtures. Proceedings of Association of Asphalt Paving
Technologists, San Antonio, Texas, pp. 571-605.

SAS Institute Inc. (2011). SAS user guide for Windows, Release 9.3. Cary, North Carolina.

TxDOT (2009). TEX-242-F: Test Procedure for Hamburg Wheel-Tracking Test. Texas
Department of Transportation, Austin. Retrieved from ftp://ftp.dot.state.tx.us/pub/txdot-
info/cst/TMS/200-F_series/pdfs/bit242.pdf.

Uppu, Kiran (2012). Durable Superpave Hot-Mix Asphalt Mixes in Kansas. Master’s Thesis.
Dept. of Civil Engineering, Kansas State University, Manhattan, July.

© ASCE

Airfield and Highway Pavements 2015