Construction and Building Materials 75 (2015) 415420

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Strengthening of hot-mix asphalt mixtures reinforced by

polypropylene-impregnated multilament glass bres and scraps
Pyeong Jun Yoo , Tae Woo Kim
Korea Institute of Civil Engineering and Building Technology, 283 Goyang Dae Ro, Daehwa-Dong, Ilsan-Su Gu, Goyang-Si, Gyeonggi-Do 411-712, Republic of Korea

h i g h l i g h t s

 The indirect strength of brous mix is about two times higher than the plain mix.
 The rut-depth of brous mixtures is about three times lower than the plain mix.
 The glass bre is utilised as a strengthening material of asphalt concrete.

a r t i c l e i n f o a b s t r a c t

Article history: This study presents an experimental overview of thermoplastic polymer-coated glass bre and scrap-
Received 26 August 2014 reinforced hot-mix asphalt (HMA) mixtures. The toughening effects of the reinforced HMA mixtures were
Received in revised form 4 November 2014 characterised using indirect tensile and Hamburg wheel-tracking tests. The indirect tensile loading tests
Accepted 7 November 2014
were used to calculate the relative indirect tensile strengths of samples and to compare their moisture
Available online 5 December 2014
susceptibilities.
The bres relative contributions to increasing the indirect tensile strength and the resistance to rutting
Keywords:
are quantied by comparing the maximum indirect peak tensile stress and tensile strength ratio (TSR).
HMA
Glass bre
The indirect tensile strength and TSR of the brous mixtures are nearly two times higher than those of
Indirect tensile strength the plain mixtures, demonstrating the lower moisture susceptibility and superior eld performance of
Hamburg wheel test the brous asphalt mixture. Accelerated rutting tests using the Hamburg wheel test setup for the brous
mixtures caused at least three times more loading passes to be required to reach the rut-depth criteria
compared to the number of loading passes required for the plain mixtures.
2014 Elsevier Ltd. All rights reserved.

1. Introduction should be overcome to ensure the reinforcing effect on the asphalt

Recently, non-synthetic and synthetic bres have been utilised
to improve the performance of hot-mix asphalt (HMA) mixtures
against permanent deformation and fatigue cracking. A few studies
in the literature have reported on experiments using various bres
in asphalt concrete mixtures.
Kaloush et al. found that the use of polypropylene (PP) and ara-
mid bres mixed in an asphalt mixture improves the resistance of
the asphalt mixture to shear deformation, as demonstrated by tri-
axial shear strength tests. The brous mixture increases the tensile
strength by 2550% [1].
Lee et al. demonstrated that the increase in fracture energy due
to the addition of recycled nylon bres from carpet represents a
potential avenue for improving the fatigue life of asphalt mixtures.
Their study also suggested that the bres balling during mixing
Corresponding author. Tel./fax: +82 31 910 0175.
E-mail addresses: pjyoo@kict.re.kr (P.J. Yoo), twkim@kict.re.kr (T.W. Kim).
concrete mixture [2].
Mahrez and Karim reported that the addition of 20-mm-long
glass bres resulted in higher resistance to fatigue cracking and
permanent deformation during repeated indirect load tensile tests.
They concluded that chopped glass bres distributed in random
directions in a mixture resist shear displacement and effectively
prevent the internal dislocation of aggregates [3].
HMA mixtures reinforced with various bres, such as carbon
bres or polyethylene terephthalate bres, have been reported to
exhibit a superior mechanical behaviour compared to general
HMA in terms of toughness, indirect tensile strength, shear
strength, and fracture energy. The improved toughness and frac-
ture energy, which can increase the fatigue life of HMA, were the
most signicant effects in the use of those bres with HMA [4,5].
Li et al. emphasised that not all bres mixed in a composite are
equally effective in their toughening effects. The bres random
distribution and orientation features require the use of assump-
tions regarding the probability density functions for calculating

http://dx.doi.org/10.1016/j.conbuildmat.2014.11.009
0950-0618/ 2014 Elsevier Ltd. All rights reserved.
416 P.J. Yoo, T.W. Kim / Construction and Building Materials 75 (2015) 415420
the effective composite stresses along the failure plane by account-
ing for the bres bridging forces. The randomness may be over-
come by the acceptable dispersion of bres without bre balling
by determining the effective dimension or the optimum content
of bres in the HMA. A promising cause of toughening may be that
the bres in an HMA can increase the shear strength at the inter-
face between the HMA matrix and the bres and that this
increased shear strength can delay the initiation and propagation
of damage [6].
Yoo and Kim proposed a reinforcing mechanism as well as a
method of optimising the bre content utilising a direct tensile
loading test that they developed. However, the bres dimension
and aggregate gradation in their study were set as constant values;
therefore, the applicability of the study is limited [5].
Although the various effects of bres in asphalt concrete can
differ considerably depending on the dimensions and contents of
the bres, this study only utilises a constant bre dimension and
a xed mixture design to verify the marginal strengthening effect
of glass bres in asphalt concrete. In addition to the brous mix-
ture, bres can also affect the ductility or stiffness of the bitumen.
However, this study only addresses the experimental results
obtained when including glass bres in asphalt concrete mixtures.
The objective of this study was to evaluate the mechanical char-
acteristics of glass bre-reinforced asphalt concrete, which was
coated in a PP resin, using indirect tensile loading and Hamburg
wheel tests. The manufacturing process for the PP-coated glass
bre was proposed to obtain an effective bre dispersion in HMA
by increasing the specic gravity and dimension of the bres.
2. Multilament glass bre-reinforced HMA
The electrical grade glass (E-glass) bres used in this study
were from a chopped rod containing 800 to 1000 monolaments
of glass bre. The roved multilament glass bre was coated using
the PP resin through an impregnation process. In addition to the
physical characteristics, as shown in Table 1, the mechanical prop-
erties of the bres, such as the tensile strength and Youngs mod-
ulus, are referred from the study of Wallenberger et al. and are at
Table 1
Physical characteristics of E-glass bres [7].
Physical properties Characteristic value
3
Density (g/cm ) 2.55
Tensile strength (MPa) 3400
Elongation (%) 4.55.0
Number of laments 8001000
Length (mm) 1012
Fig. 1. (a) Glass bre scraps and (b) aggregate-like PP-coated scrap (PPGS).
least 1001000 times higher than the typical values of an HMA
at room temperature [7].
The specic gravity of the glass bre is comparable to a general
aggregate greater than 2.0, whereby the PP-coated multilament
bre strand may behave as an aggregate without any noticeable
balling of bres during the mixing process. The effective dispersion
without bre balling may be expected due to the aggregate-like
behaviour of the bre.
Two different concentrations of PP-coated multilament bres,
1% and 2% of the weight of the mixture, were used to reinforce the
asphalt concrete. Only one mixture contains the bre scraps in the
shape of aggregate; the bre scraps were coated using the PP resin
through an extrusion process, as shown in Fig. 1. The glass bre
scraps are a by-product obtained from producing a roving bre.
The nominal maximum sieve size of the scraps was 100 lm, which
is one grade higher than a standard sieve size for mineral ller. The
scraps passing the 75-lm sieve were utilised in making the
aggregate-like scraps in this study, which are coated using the PP
resin (PPGS) through the extrusion process, as shown in Fig. 1.
The application rate of PP for coating the scraps (PPGS) and rov-
ing bres (PPGF) was controlled by the relative weight ratio
between the weight of scraps or roving bres and the weight of
PP by 1:1.
The gradation of scraps was as shown in Table 2. Because of the
difculty in controlling the content of dust or mineral ller when
performing a mixture design in an asphalt plant, this study utilised
the PPGS to more easily perform the mixing with the aggregate
compared to the powder type and to replace all mineral llers
and dust in the HMA concrete.
In addition to the PPGS, PP-coated multilament glass bres
(PPGFs) were developed for use in HMA applications as reinforcing
media, as shown in Fig. 2a and b. The extrusion process with an
impregnation step for the PP coating shown in Fig. 2a can give
the PPGFs an elliptical cross-section. The PPGFs were cut into 10-
mm-long pieces that were 1 mm wide on the short side and
2 mm wide on the long side, as shown in Fig. 2b.
The circular PP particles in Fig. 2b were added to the HMA
according to the equivalent weight ratio between the glass bre
before coating and the PP resin. The PP particles were completely
melted during the wet-mixing process with a hot binder in a plant,
Table 2
Gradation of glass bre scraps.
Sieve size (lm) % Passing
100 100.0
75 83.0
45 47.0
 P.J. Yoo, T.W. Kim / Construction and Building Materials 75 (2015) 415420 417

Fig. 2. (a) Extrusion and impregnation process and (b) PP-coated glass bre and PP particles.

Table 3 between large aggregates greater than 4.75 mm may the cause of
Fibre contents and aggregate gradation. increasing volume of mixture. The increased volume may result
PPGF (%) 0 1 2 2 in relatively greater air void of the glass bre reinforced mixture
PPGS (%) 0 0 0 2 in this study.
Asphalt content (%) 5.45 5.45 5.45 5.45 The xed asphalt content of 5.45% was utilised as a predened
Sieve size (mm) % Passing value for the coarse aggregate gradation within the context of this
19 100.0 100.0 100.0 100.0 study through the Marshall mix-design process described in
13 93.0 93.0 93.0 93.0 another study [5].
10 56.0 56.0 56.0 56.0 This study attempted to have a xed binder content that was
4.75 31.0 32.0 33.0 33.0
independent of the mixtures behaviour, aiming to observe only
2.36 17.0 18.0 18.0 18.0
1.18 10.0 11.0 12.0 12.0 variations due to the bre contents. The asphalt binders for all of
0.6 7.0 8.0 9.0 9.0 the mixtures used in this study were Superpave PG 64-22 grade
0.3 4.0 5.0 6.0 6.0 with a penetration of 65 and a softening point of 49 C.
0.15 2.0 3.0 4.0 4.0 Before adding the asphalt binder, dry mixing at a temperature
Filler 2.0 1.0 2.0 0.0
of approximately 170 C, as shown in Fig. 3a, was performed to
determine whether the bres were entangled. All of the PP, such
as the additional PP particles and the impregnated PP in the PPGF,
and the particles may play a role as an adhesion improver. Fig. 2b was melted during the dry mixing for a duration of 2 min because
shows the PPGF including the PP particles. of the relatively low melting temperature of the PP of approxi-
Table 3 shows the bre contents and coarse-side aggregate gra- mately 150 C. The impregnated bres were released in each
dations used in this study. Except for the control mixtures without monolament and dispersed throughout the mixtures, as shown
bres in Table 3, the multilament bre contents varied from 1% to in Fig. 3b.
2% of the total weight of the HMA, and only one mixture contained Without any balling of the bres in the mixture, all of the mix-
the PPGS. According to the mixture design, one control mixture tures were compacted in a Marshall mould at approximately
and three brous mixtures were tested to address the relative 160 C until the target 75 tamps on both sides for each mixture
toughening effect due to the inclusion of bres through the indirect were achieved to perform the indirect tensile loading tests.
tensile strength and Hamburg wheel-tracking tests. In contrast, all of the mixtures for the Hamburg wheel tests
The aggregate gradations were designed to include 1013% air were compacted in a 150-mm-diameter mould at approximately
voids. Air voids comprised approximately 10% of the plain mixture 160 C using a gyratory compacter, with 100110 gyrations
and 13% of the brous mixtures; the interference of bres in applied to each mixture until the target height for each mixture

Fig. 3. (a) Dry-mixing bres and aggregate and (b) dispersed bres in the HMA after wet mixing.
418 P.J. Yoo, T.W. Kim / Construction and Building Materials 75 (2015) 415420

(i.e. 60 mm) was achieved. Two mixtures with dimensions of which is almost twice the tensile strength of the plain mixture.
approximately 150 (diameter)  60 mm (thickness) were placed In addition, the GF-2-2% was selected for use in experiments on
into the Hamburg test mould. the freezethaw cycle to estimate the moisture susceptibility.
Test mixtures of GF-0% (no bre) and GF-2-2% (PPGF and PPGS)
3. Laboratory tests in Fig. 4, were conditioned according to the AASHTO T-283-07.
Fig. 5a and b illustrate the indirect tensile strength test setup
3.1. Indirect tensile strength test and the conditioning for GF-0% and GF-2-2%, with six mixtures
for each test, half of which were to be tested under dry conditions
The toughening effect of glassbre-reinforced HMA was char- and the other half of which were to be tested after partial satura-
acterised by considering mixtures subjected to indirect tensile tion and moisture conditioning with a freezethaw cycle [8].
loading according to AASHTO T-283-07 with a freezethaw cycle. The indirect tensile strength and tensile strength ratio (TSR) are
Before performing the freezethaw cycle, indirect tensile calculated as follows:
strength tests without the freezethaw cycle were performed to 2000P
estimate the indirect tensile strengths for all of the mixtures in St 1
ptD
Table 3. The GF-2-2% (PPGF 2% and PPGS 2%) in Fig. 4 exhibited
the highest indirect tensile strength of approximately 1.0 MPa, where St is the indirect tensile strength (kPa), P is the maximum
Load (N), t is the mixture thickness (mm), and D is the mixture
diameter (mm).
S2
Indirect Tensile Strength (MPa)

1.2
TSR 2
S1
1.0
where S1 is the average tensile strength of the dry subset (kPa) and
0.8 S2 is the average tensile strength of the conditioned subset (kPa).
The indirect tensile strengths of the glass brous mixture and
0.6
the mixture without bres under dry conditions, as shown in
0.4 Fig. 6a, were measured to nearest 1.2 and 0.6 MPa, respectively.
The indirect tensile strengths of the glass brous mixture and the
0.2 mixture without bres under wet conditions, as shown in Fig. 6b,
were nearest 1.0 and 0.4 MPa, respectively.
0.0
GF-0% GF-1% GF-2% GF-2-2% The average TSR of the glass brous mixture and the mixture
without bres, according to Eq. (2) and as shown in Fig. 7, were
Fig. 4. Indirect tensile strength of dry specimens. approximately 0.73 and 0.44, respectively. The indirect tensile

Fig. 5. (a) Indirect tensile loading test and and (b) specimens in conditioning.
Indirec Tensile Strength (MPa)
Indirect tensile Strength (MPa)

1.4 1.4
1.2 No Fiber 1.2 No Fiber

1 1
0.8 Glass Fiberous Specimen 0.8
Glass Fibrous Specimen
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Displacement (mm) Displacement (mm)
(a) (b)
Fig. 6. (a) Dry condition and (b) freezethaw condition.
 P.J. Yoo, T.W. Kim / Construction and Building Materials 75 (2015) 415420
80
No Fiber
60 Glass Fibrous Specimen
TSR (%)
40
20
0
Coarse Aggregate gradation
Fig. 7. TSR comparison.
strength and average TSR for the brous mixture were nearly two
times higher than those from the plain mixture, indicating the
lower moisture susceptibility and potentially superior eld perfor-
mance of the brous asphalt mixtures.
3.2. Hamburg wheel-tracking test
The Hamburg wheel test is believed to be the best tool for esti-
mating the resistance of a mixture to rutting and stripping. In a
dual-wheel track system, two loaded steel wheels track over the
samples in a heated water bath, generally held at 50 C, and the
deformations are recorded versus the number of loading passes.
The Hamburg wheel test requires some allowance for variability
due to the difculties in maintaining consistencies, such as those
related to materials, sampling, and test conditions. However, it is
an effective test for simultaneously comparing the resistance to
rutting or stripping of two different mixtures [9].
The dimensions of the cylindrical mixtures are 150 2 mm in
diameter and 62 2 mm in height, as shown in Fig. 8ac. The test
criteria for the allowable maximum rut depth of 12.5 mm
depended on the types of binders, such as PG 64, PG 70, and PG
76 require the number of wheel passes to be 10,000, 15,000 and
20,000, respectively [10].
After completing the tests, the total rut depth of the brous
mixture reinforced by the 2% of PPGS and 2% of PPGF to the mix-
ture weight, shown in Fig. 8a, was only about 4.6 mm after
20,000 wheel passes. In contrast, all of the control mixtures with-
out bres, as shown in Fig. 8b and c, were completely destroyed
before 5000 wheel passes.
The rut depth versus the number of wheel passes graph (see
Fig. 9) shows that the brous mixture could support the wheel
loading until 20,000 loading repetitions were completed, with an
average terminal rut depth of 3.8 mm, which is extremely low,
indicating that there is no damage or stripping. In contrast, all of
the normal mixtures were completely destroyed before completing
5000 loading repetitions.
Fig. 8. Hamburg wheel-track test specimens (a) brous Specimens, (b) normal control 1, and (c) normal control 2.
419
0
-2
-4
Rut Depth (mm)
-6
-8
Fibrous Mix 1
-10
-12 Fibrous Mix 2
-14 Normal Mix 3
-16 Normal Mix 4
-18
-20
0 5000 10000 15000 20000
Number of Wheel Passes
Fig. 9. Rut depth vs. number of wheel passes.
This result indicates that the shear strength of the bre-rein-
forced mixture was likely increased by the bridging effect of the
bres along the vertical failure plane of the mixture [5].
4. Conclusion
This paper has presented the experimental results of a study on
PPGF- and PPGS-reinforced HMA mixtures.
The glass bres used in this study were from a chopped strand
containing 800 to 1000 monolaments of glass bre. The multil-
ament glass bres were coated using a PP resin through an impreg-
nation process and then cut to the proper length (10 mm in this
study). The glass bre scraps are the by-product obtained from
producing the roving glass bres.
The maximum sieve size of the scraps was 75 lm, which is a
typical sieve size of mineral ller. The scraps passing through the
75-lm sieve were utilised in making the aggregate-like scraps,
which were coated with a PP resin through an extrusion process.
The toughening effects of the glassbre-reinforced HMA mix-
tures were characterised using indirect tensile strength and Ham-
burg wheel-tracking tests.
Indirect tensile strength tests were performed to calculate the
relative indirect strengths of the samples and to compare their
moisture susceptibilities. The indirect tensile strength and TSR of
the brous mixtures were nearly twice as high as those of the plain
mixture, demonstrating the lower moisture susceptibility and
potentially superior eld performance of the brous asphalt
mixture.
Accelerated rutting tests were performed on the brous and
plain mixtures using a Hamburg wheel test setup. The average
rut depth of the brous mixture was 3.8 mm upon completing
20,000 loading repetitions. The rut depth for the plain mixtures
without bres exhibited increased by at least tenfold compared
to the initial value before completing 5000 loading passes.
420 P.J. Yoo, T.W. Kim / Construction and Building Materials 75 (2015) 415420
This signicant increase indicates that the bres bridging effect
could resist the tensile or compressive movement of the brous
mixture even on the rigid plastic foundation of the Hamburg wheel
test setup. The rut depths of the plain HMA were at least three
times higher than those of the bre-reinforced HMA. The brous
HMA can prolong the performance life of the pavement system
due to the bridging effect of the bres.
This result indicates that the tensile or compressive strength of
the bre-reinforced mixture was increased by the bridging effect of
the bres along the failure plane.
Acknowledgements
This study was supported by the Internal Research Program
(Pothole-Free Pavement System) of the Korea Institute of Civil
Engineering and Building Technology, which is funded by the Min-
istry of Knowledge and Economy.
References
[1] Kaloush KE, Biligiri KP, Zeiada WA. Evaluation of FORTA ber reinforced
asphalt mixtures using advanced material characterization tests-Evergreen
drive. Tempe. Arizona, Research report. Grove city (PA, USA): FORTA
Corporation; 2008.
[2] Lee SJ, Rust JP, Hamouda H, Kim YR, Bordon RH. Fatigue cracking resistance of
ber reinforced asphalt concrete. Text Res J 2005;75(2):1238.
[3] Mahrez A, Karim MR. Rutting characteristics of bituminous mixes reinforced
with glass ber. Eastern Asia Soc Transport Stud 2007;6:216872.
[4] Fitzgerald RL. Novel applications of carbon ber for hot mix asphalt
reinforcement and carboncarbon preforms. M.S. Thesis. Huoghton (MI):
Department Chemical Engineering, Michigan Technological University; 2000.
[5] Yoo PJ, Kim K. Thermo-plastic bers reinforcing effect on hot-mix asphalt
concrete mixture. Constr Build Mater 2014;59:13643.
[6] Li VC, Wang Y, Backer S. A micromechanical model of tension-softening and
bridging toughening of short random ber reinforced brittle matrix
composites. J Mech Phys Solid 1991;39(5):60725.
[7] Wallenberger FT, Watson JC, Li H. Glass bers. ASM hand book, vol.
21. Materials Park (OH): ASM International; 2001.
[8] AASHTO T 283-07. Standard method of test for resistance of compacted asphalt
mixtures to moisture-induced damage. Washington D.C.: American
Association of State Highway and Transportation Ofcials; 2011.
[9] Lu Q, Harvey JT. Evaluation of Hmaburg wheel tracking device test with
laboratory and eld performance data. Transportation research record 1970.
Washington, D.C.: TRB, National Research Council; 2006. p. 1424.
[10] Rahman F, Hossain M. Review and analysis of Hamburg wheel tracking device
test data. Report No. KS-14-1. Mahattan (Kansas): Kansas State University
Transportation Center; 2014.