Você está na página 1de 8

Geotextiles and Geomembranes xxx (2015) 1e8

Contents lists available at ScienceDirect

Geotextiles and Geomembranes


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

Evaluation of permanent deformation of geogrid reinforced asphalt


concrete using dynamic creep test
Sina Mirzapour Mounes a, *, Mohamed Rehan Karim a, Ali Khodaii b,
Mohamad Hadi Almasi a
a
b

Centre for Transportation Research, Faculty of Engineering, Civil Engineering Department, University of Malaya, 50603 Kuala Lumpur, Malaysia
Department of Civil and Environmental Engineering, Amirkabir University of Technology, 15914 Tehran, Iran

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 17 November 2014
Received in revised form
6 April 2015
Accepted 4 June 2015
Available online xxx

Permanent deformation (rutting) is one of the distresses that can adversely affect the bituminous surface
of pavement structures, particularly in hot climates. The geosynthetics reinforcement of hot mix asphalt
is one of the means to combat rutting. In this study, a dynamic creep test was performed on asphalt
concrete samples reinforced with four different types of berglass grid as well as on unreinforced
samples. The berglass grids used in this study contained two different sizes of grid openings and two
tensile strengths, allowing us to test for the mesh size and tensile strength effects of the grids on the
permanent deformation behavior of double layered asphalt concrete. In addition, we tested a recently
developed creep curve model has been veried and used this to study the creep behavior of the samples
in the primary and secondary regions of the creep curve, as well as determining the boundary point of
the regions. The results suggest that not only grid tensile strength, but also grid mesh size is of great
importance in combatting permanent deformation of berglass grid reinforced asphalt concrete within
the conditions and grids used in this study. In a nutshell, higher tensile strength and/or smaller mesh size
grids lead to overall better performance of grid reinforced samples. Moreover, great care must be taken
when the creep curves are not reached in the tertiary region, and the creep rate must be taken into
account to avoid any misinterpretation of the results.
2015 Elsevier Ltd. All rights reserved.

Keywords:
Geosynthetics
Permanent deformation
Asphalt
Dynamic creep test
Creep curve model

1. Introduction
1.1. Overview
A bituminous mixture applied to the surface or the base layer of
a pavement structure serves to distribute the trafc load and prevent water from penetrating into underlying unbound layers (Epps
et al., 2000). Due to applied trafc loading, there are many different
types of distresses that can affect bituminous surface layers,
including permanent deformation (rutting), and fatigue cracking.
In recent years, because of increases in the volume of trafc and
of heavy vehicles, rutting is one of the most frequent defects found
in exible pavements, particularly in hot climates. Rutting shows
up as depressions formed in the wheel path in a pavement. It
normally occurs when a permanent deformation of each layer in

* Corresponding author. Tel.: 60 3 7967 5339; fax: 60 3 7955 2182.


E-mail address: sinamm.1981@gmail.com (S. Mirzapour Mounes).

the pavement structure accumulates under a repetitive trafc load


(Tayfur et al., 2007). There are generally two modes of ruts that
occur on pavements, compactive and plastic (Gabra and Horvli,
2006; Lee et al., 2010).
Accumulation of residual strains in wearing course may cause
serious problems, particularly through aquaplaning on wet pavements (Fwa et al., 2004; Sivilevi
cius and Petkevicius, 2002;
Verhaeghe et al., 2007). Thus, not only does pavement rutting
lead to higher road maintenance costs, but it also increases the risk
to human life through accidents caused by water accumulating in
depressions (ruts) in pavements.
Various laboratory testing methods have been developed to
investigate the resistance to rutting of asphalt concrete. These
include the static/dynamic creep test, wheel track test, and indirect
tensile test. Monismith et al. (1975), quoted by Kalyoncuoglu and
Tigdemir (2011), developed the dynamic creep test which is
thought to be one the best methods to evaluate the resistance of
asphalt concrete to permanent deformation. Furthermore, a report
by the NCHRP (Cominsky et al., 1998), quoted by Kaloush and

http://dx.doi.org/10.1016/j.geotexmem.2015.06.003
0266-1144/ 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

Witczak (2002), identied the dynamic creep test as having the


potential to be utilized as a eld quality control.
One of the most important outputs of the dynamic creep test is
the creep curve, which illustrates permanent deformation versus
loading cycles. Since creep curves obtained by the dynamic creep
test are used to assess the resistance to permanent deformation of
asphalt concretes, not only is the behavior of each region of the
creep curve of great importance, but the identication of the
boundary points connecting the primary to the secondary region,
and the secondary to the tertiary region, is also important.
Analyzing the creep curves obtained from a dynamic creep test in
this way can provide a better understanding of the resistance to
permanent deformation of asphalt concrete.
1.2. Literature review
During the past decade or so, there have been a lot of studies by
various researchers into how to hinder rutting in pavements
through geosynthetic reinforcement (Austin and Gilchrist, 1996;
Collin et al., 1996; Laurinavi
cius and Oginskas, 2006; Leshchinsky
and Ling, 2012; Ling and Liu, 2001; Perkins, 1999; Thakur et al.,
2012; Yang et al., 2012). Virgili et al. (2009) reported that the
reinforcement of bound layers can be divided into three elements:
fatigue life extension, the reduction of reection cracking, and the
reduction of permanent deformation. Geogrids are high-strength
extruded sheets of polyethylene or polypropylene with holes
punched to produce a regular, grid-like pattern. Geogrids are quite
stiff compared to the bers of geotextiles, and have a higher
modulus (Appea, 1997). Various literature points to the fact that
stiffer geogrids lead to better performance in pavements (Collin
et al., 1996; Ling and Liu, 2001; Perkins, 1999). Past research
further suggests that some geosynthetics, and particularly certain
geogrids, have a positive inuence on permanent deformation in
asphaltic pavements. This inuence is stronger when the geogrid
reinforcement is laid at mid-depth of the asphalt concrete, rather
than embedded at the bottom (Sobhan et al., 2005). Similarly
Perkins (1999) reported that the closer the placement positions of
the geosynthetic to the applied load, the higher the reinforcement
effect.
The elasticity modulus of asphalt concrete can be improved by
incorporating materials with certain properties, such as grid reinforcement, within the asphalt mixture. There is evidence that the
rutting depth of asphalt concrete depends on its modulus of elasticity, while the modulus of elasticity is in turn dependent on the
type of geosynthetic material used (Laurinavi
cius and Oginskas,
2006). Moreover, lateral tensile strains can be restricted through
geosynthetics, which are stiff when under tension (Perkins and
Ismeik, 1997). FEM calculations have shown that the mechanically
restrained system generated by geogrids can hold back the movement of aggregates and increase the transverse binding force of an
asphalt layer, which can contribute to increased resistance to
rutting (Fei and Yang, 2009). Furthermore, the opportunity provided by grid-shaped geosynthetics for aggregates from the top and
bottom layers to interlock should logically lead to greater friction
between layers (Tutumluer et al., 2010).
Researchers generally agree that geogrid reinforced asphalt
concrete is more resistant to surface deformation than unreinforced concrete (Austin and Gilchrist, 1996; Bertuliene et al., 2011;
Komatsu et al., 1998; Siriwardane et al., 2010; Sobhan et al., 2005).
Yet very few researchers have provided a full-range comparison of
the effects of the mesh size and tensile strength of such grids on the
permanent deformation of bituminous systems. Tests of asphalt
concrete plastic ow resistance suggest that durability increases
when the size of grid openings (Komatsu et al., 1998). Similarly,
Jenkins et al. (2004) observed slightly less rutting in grids

reinforced with smaller grid mesh sizes than those with larger
mesh sizes.
As quoted by Zhou et al. (2004), different mathematical models
such as Barksdale's Semilog model (1972), Power-law models based
on the Monismith model (1975), and Tseng and Lytton's model
(1989), have been developed in order to t the creep curve and
estimate the ow number parameter in asphalt mixtures. For tting
and distinguishing between regions of the unmodied asphalt
mixture creep curve, Zhou et al. (2004) proposed a three-stage
model: a power model for the primary region, a linear model for
the secondary, and an exponential model for the tertiary region. In
other words, they modeled each region of the creep curve separately. At the same time, some other researchers believe that the
logarithmic model simulates more accurately the primary region of
the creep curves in SBS modied asphalt mixtures (Kalyoncuoglu
and Tigdemir, 2011; Ahari et al., 2013).
Ahari et al. (2013) developed a two-stage model for the primary
and secondary regions of creep curves in SBS modied asphalt
mixtures. They proposed two different approaches for modeling
the primary and secondary regions of the creep curve as follows:
Approach 1. Both the primary and secondary regions of the creep
curve can be modeled simultaneously using the following logarithmic function:

3P a  LnX b
where:
X: is loading cycle
3P: is accumulated permanent strain at loading cycle X
a, b: are constants
Then, in order to check if the developed logarithmic function ts
well with both regions, the deviation errors of all the points, re
calculated as below, must be less than or equal to 1%:

De





3PCalculated  3PMeasured
3PMeasured

 100

where:
De: is deviation error (%)
Approach 2. In order to identify the boundary points of the primary and secondary regions of the creep curve, the following steps
are taken:
1 Visual selection of loading cycles among the initial loading cycles of the secondary region. [It must be noted that this loading
cycle is not necessarily the boundary point]
2 Removal of the loading cycles before the selected loading cycles,
and plotting a new graph representing the approximate secondary linear region.
3 Fitting a linear model to the approximate secondary region and
determining the model coefcients.
4 Calculating the accumulated permanent strain for all the loading
cycles of the approximate secondary region, based on obtaining
model coefcients.
5 Determining the deviation error (De) for all of the calculated
accumulated permanent strains.
6 If simultaneously all the De(s)  1% / the criterion is met and
the linear model is assumed to be representative.

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

If at least one of the De(s) > 1% / go to the next loading cycle


and repeat steps 2e6 until the former criterion is met.
7 Fitting the logarithmic function resulting from approach 1 to the
primary region.
8 Solving the set of simultaneous equations, called logarithmic
and linear respectively, for the results of the primary and secondary regions, in order to identify the accumulated permanent
strain and its corresponding loading cycle, where the primary
region is connected to the secondary region.
In the present study, comparisons were carried out on the dynamic creep curves of asphalt concrete samples reinforced by four
types of berglass grids with two different tensile strengths and
two different grid opening sizes, as well as on unreinforced samples, in order to assess their respective resistance to pavement
deformation. As the maximum number of cycles applied in this
experiment was 10,000 due to time limitations, none of the samples reached the tertiary region. We rst tested out a mathematical
model by Ahari et al. (2013), recently developed to model the primary and secondary regions of SBS modied asphalt concrete creep
curves, to see if could model the creep curves of the materials in the
current study. Using this, we then investigated the effects of combined and separate variations in both grid tensile strength and
opening size, applied at the mid-depth of asphalt concrete, on the
samples' resistance to permanent deformation. Finally, the
behavior of the primary and secondary regions and their boundary
points in the creep curves obtained for, the various types of samples
were analyzed and compared.
2. Experimental program
2.1. Materials and sample preparation
Crushed granite supplied from the Kajang region of Selangor
state in Malaysia was used as aggregates in this study. Fig. 1 shows
the aggregate gradation for the dense graded mixture utilized in
this research, with a nominal maximum aggregate size of 9.5 mm in
accordance with ASTM D3515 (2000). Bearing in mind the opening
size of the berglass grids used in this study, selecting this aggregate gradation should allow the grids to provide better interlocking
with the asphalt concrete. The applied bitumen was 80/100 penetration grade, and the optimum asphalt content of the dense graded
asphalt mixture was determined to be 5% by mass of the total
mixture, using the Marshall Test. The asphalt concrete slabs were
compacted using a roller compactor in accordance with EN 1269733 (2003) in two lifts to the target air void of 8%, in order to

simulate compaction at the time of eld construction (Kandhal and


Chakraborty, 1996). The layer thickness of each lift of the slabs was
40-mm, resulting in 80 mm thick compacted slabs. Four types of
berglass grid manufactured by a European corporation, with two
different tensile strengths and two different opening sizes, were
employed as the reinforcing material applied at mid-depth in the
reinforced specimens. This study compares two levels (high and
low) of grid tensile strength and two levels (large and small) of grid
opening size. It should be noted that the dimensions of the girds
with large opening sizes differed slightly from those with small
opening sizes; however, since the difference was very small, this
study assumes that they were of identical size. The basic properties
of all these reinforcements are presented in Table 1.
The specimens to be cored and trimmed into cylindrical shapes
had dimensions of 150-mm diameter and 60-mm height as recommended by EN 12697-25 (2005) so that the applied berglass
grid was placed at mid depth of the sample. The average volumetric
properties of the testing samples are illustrated in Table 2. The code
for each sample in Table 2 includes whether the sample was reinforced or unreinforced, as well as the type of glass grid used for
reinforcement in that particular sample.
2.2. Dynamic creep test
The creep test was conducted using a uniaxial cyclic compression test with the connement method, as recommended by EN
12697-25 (2005). However, since only three cores were attainable
from each slab, three test repetitions were carried out so as to
minimize any variability among replicates of one type of specimen.
For that purpose, UTM-5P from IPC was used to apply a constant
dynamic load at a certain periodic rate onto the cylindrical asphalt
samples, and vertical deformation was measured using a Linear
Variable Displacement Transducer (LVDTs). The servo pneumatic
UTM-5P machine has integrated software that allows the operator
to select several input parameters such as loading function, stress,
frequency and seating stress Static pre-loading fora certain period
of time can also be applied to the samples before cyclic loading is
started. The loading jig is moreover located in an environmental
chamber so as to control the testing temperature.
In the present study, in accordance with EN 12697-25 (2005),
the test was performed for both reinforced and control samples at
40  C, at a cyclic stress level of 100 kPa and a frequency of 0.5 Hz,
with 1000 ms allocated for each cycle width and the corresponding
rest period. For all the samples, a constant stress of 100 kPa was
applied for up to 10,000 cycles due to time limitations. Moreover, a
static preloading stress of 10 kPa was applied to all the samples for a
period of 10 min prior to initiating the dynamic load, in order to

Fig. 1. Grading curve for crushed aggregate.

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

Table 1
Basic properties of ber glass grid applied.
Identication

Glass grid A

Glass grid B

Glass grid AA

Glass grid BB

Tensile strength (kN/m) (MD  XD)

115  115
/15
12.5  12.5

115  115
/15
25  25

115  215
/15
12.5  12.5

115  215
/15
25  19

2.5
4600  4600

2.5
4600  4600

2.5
4600  8600

2.5
4600  8600

Grid size (mm)


Center to center of strand
Tensile elongation (%)
Secant stiffness (N/mm)

Table 2
Physical properties of tested samples.
Sample code

Applied glass grid

Bulk specic gravity

Maximum specic gravity

Air void (%)

C
R1
R2
R3
R4

e
Glass
Glass
Glass
Glass

2.226
2.225
2.225
2.225
2.222

2.425
2.425
2.425
2.425
2.425

8.23
8.27
8.25
8.23
8.36

grid
grid
grid
grid

A
B
AA
BB

ensure proper contact between the core surface and loading platen.
Moreover, all the samples were conditioned at 40  C for about 4 h in
a temperature chamber to make sure that they had reached the
testing temperature.

3. Test results and discussion


3.1. Permanent strain comparison
The permanent deformation potentials of asphalt concrete
reinforced with four different types of geosynthetics were
compared with each other, as well as with unreinforced (control)
samples in order to identify which type had the highest resistance
to permanent deformation. Considering that there are three replicates for each type of sample, the diagrams are derived from the
average amount of parameters. Fig. 2 illustrates the creep curves of
the samples tested in this study.
Thereafter, Ahari's stepwise model was veried for the materials
used in this study and then used to t the creep curves and
determine the connecting points between the primary and secondary phases. This method consisted of eight steps as shown in
Section 1 (Ahari et al., 2013).

Fig. 2. Creep curves of tested mixtures, including reinforced and control samples.

The application of berglass grids at the mid-depth of the


samples of asphalt concrete, notably increased their resistance to
permanent deformation over that of the unreinforced samples.
Moreover, as can be seen from Fig. 2, the samples reinforced by
berglass grids with greater tensile strength and greater mesh size
(R4) showed the lowest permanent deformation throughout all the
cycles conducted in this study.
Fig. 2 also clearly shows that the control (unreinforced) samples
had substantially higher permanent deformation and accumulation
rates of permanent deformation than the reinforced ones e due
presumably to the tensile forces and lateral connement provided
by the grids.
A further important nding is that the samples of identical mesh
size reinforced by grids of lower tensile strength experienced more
permanent deformation than those reinforced by grids with a
higher tensile strength. Furthermore, the difference between the
permanent deformations in samples reinforced with large mesh
size was higher than for samples reinforced with smaller mesh size
grids. In other words, increasing the tensile strength of grids with a
large mesh size had a greater impact on their ability to resist permanent deformation than such increase in small mesh size grids
within the test conditions performed in this study. In addition, by
applying 10,000 loading cycles in this experiment, we found that
larger permanent deformation occurred in samples with small
rather than large mesh sizes, regardless of whether they had high
or low tensile strength grids.
Table 3 illustrates the results of a quantitative comparison between the measured permanent strain and grid tensile strength
and grid opening size respectively, during the last loading cycle of
the tests carried out in this study. In this table, reinforced samples
with the same size of opening (mesh) are compared in terms of
their grid tensile strength, and those with the same tensile strength
in terms of their grid opening size. The improvements in resistance
to permanent deformation shown in Table 3 were all determined
based on the permanent deformation of the control samples. The
clear conclusion from Table 3 is that, based on testing through
10,000 cycles, samples reinforced by grids with greater tensile
strength and with larger mesh size achieve the best performance.
In can also be seen from Table 3 that increasing the tensile
strength in samples with small grid openings from R1 to R3 leads to
a 4% improvement in permanent strain resistance by the last
loading cycle. However, doing the same thing with grids with a
large grid opening size from R2 to R4 leads to a 12% improvement,
ie three times as much.

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

Table 3
Comparison of permanent strains in the last cycle.
Sample code

Tensile strength
(kN/m) (MD  XD)

Grid mesh size (mm)


Center to center of strand

Measured last cycle


permanent strain (m3)

Improved permanent strain resistance


compared to control sample (%)

C
R1

e
115  115*L
/15
115  115*L
/15
115  215*H
/15
115  215*H
/15

e
12.5  12.5**S

11,438
8756

0
31

8564

34

8495

35

7819

46

R2
R3
R4

25  25**OL
12.5  12.5**

25  19**OL

*L: Low level for grid tensile strength; *H: High level for grid tensile strength; **S: Small level for grid opening size; **OL: Large level for grid opening size.

Conversely, comparing the samples with grids of the same


tensile strength, but different size of opening (mesh), it emerges
that in samples with low tensile strength grids, an increase in grid
opening size from small to large (R1 to R2) leads to a 3%
improvement in permanent strain resistance: while doing the same
(R3 to R4) with samples with grids of high tensile strength grids
leads to a much larger, 11% increase in such resistance.
In research studying the shear behavior of bi-layer asphalt
concrete specimens, geogrid reinforced samples showed less
interlayer shear resistance than unreinforced ones, even though
some of the geogrid surface coatings were found to be able to
maximize bonding between the interlayer and asphalt concrete
(Ferrotti et al., 2012). It may be, therefore, that the effects of the
smaller mesh size observed in the current testing condition of this
study were due to reduced bonding between the lower and upper
lift of the asphalt concrete, leading to the development of higher
shear deformation. Comparing the samples reinforced by small
mesh grids and the control ones, it should be noted that, although
the bonding of two lifts was important in the reinforced samples,
the reinforcing effect of the grid was much more signicant than its
effect on the bonding condition of the lifts; the upper and lower
lifts in fact remained in full contact in the control samples. It can
likewise be seen that the more the tensile strength increases, the
greater the effect of mesh size on strength and resistance.
In sum, if we look merely at the accumulated permanent strain
up to the last cycle of the creep test conducted in this work, this
leads to the conclusion can be drawn that not only increasing the
tensile strength, but also enlarging the mesh size of glass grid
reinforced asphalt concrete can increase its resistance to permanent deformation. However, it should be noted that the ow point
was not reached in the performed test conditions, and that closer
investigation of the creep curves after model tting resulted in
rather inferences from the ones drawn in this section.

3.2. Fitted models comparison


Unfortunately, none of tested samples reached the third phase
of the creep curve in the course of the 10,000 loading cycles conducted in this experiment. As a results, only the primary and secondary phases could be modeled; the two regions for which Ahari's
model was developed. Table 4 presents the results of mathematical
functions and estimated permanent strains at the boundary points
at the last cycle for each phase of testing samples. Based on
Figs. 3e5, and the coefcients of determination in Table 4 it can be
seen that the tted models, both for the logarithmic and linear
regions, t acceptably with the measured creep curves. Thus, it can
be concluded that Ahari's model is suitable for modeling the primary and secondary regions of the creep curve for both the ber
glass grid reinforced samples and unreinforced hot mix asphalt
samples. The slopes of both the primary and secondary regions are
important, particularly the secondary region generally known as
the creep rate. In Ahari's proposed model, a linear logarithmic
model is utilized to model the primary region of the creep curve as
shown bellow:

y a blnx
b is the ratio of absolute change in y to the relative change in x. In
other words, if x changes by 1%, then the absolute change in y is
0.01b unit (Thomas et al., 2001). However, the slope of the primary
region is not as important as that of the secondary region. In our
study, the slope of the tted curve in the primary and secondary
regions was determined for each individual sample type. The
extend of improvement for each (in terms of smaller permanent
strain accumulation rates) was then determined based on the
control sample. These results are shown in Table 5. It is worth
noting that the control samples in this table have their maximum
slopes in both primary and secondary regions of the creep curves.

Table 4
Creep curve models based on Ahari's model and estimated critical values.
Code First stage model

End of rst stage

Second stage model

C
R1
R2
R3
R4

3p 1523.945 Ln(N)  2653.456


R2 0.9842
3p 1114.477 Ln(N)  1226.826
R2 0.9927
3p 1075.835 Ln(N)  1099.156
R2 0.9893
3p 1062.883 Ln(N)  843.671
R2 0.9902
3p 954.627 Ln(N)  980.730
R2 0.9918

7072

10,863

3945

8001

36

3223

7592

43

3001

7666

42

4847

7120

53

Last cycle
Cycle (N) 3p(modeled) Improved 3p
compared to
control sample (%)

Cycle (N) 3p (modeled) Improved 3p


compared to
control sample (%)
3p 0.2155(N
R2 0.9913
3p 0.1336(N
R2 0.9864
3p 0.1515(N
R2 0.9957
3p 0.1291(N
R2 0.9824
3p 0.1436(N
R2 0.9868

 7072) 9346.539 10,000

11,502

 3945) 7474.168 10,000

8810

31

 3223) 7103.268 10,000

8618

33

 3001) 7278.914 10,000

8570

34

 4847) 6424.536 10,000

7860

46

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

Fig. 3. Measured and estimated permanent deformation for R1 sample.

estimated values were rather close to each other. We therefore,


used the tted curves from the measured values to nd the turning
point between the primary and secondary regions of the creep
curves.
When we only took into account the creep curves (as in Fig. 2),
this pointed to the conclusion that enlarging the grid mesh size at
the same level of tensile strength leads to better performance
(resistance) within the used grids in this study. However, it can be
seen from Table 5 that enlarging the mesh size has the effect of
increasing the secondary region slope e something which does not
emerge from just looking at the creep curves. In the secondary
region, in which the mixture has reached to an optimum density
level (Mehta et al., 2014), the presence of steeper slopes for grids
with a larger mesh size but with the same tensile strength may be
due to there being a lower number of grid junctions on the grid
applied area. In other words, the number of stripes or threads of
grid per unit area of the sample increases as the size of the opening
(mesh) is reduced, leading to greater structural and dimensional
stability through the higher number of grid junctions. This could
possibly explain the smaller slope in the secondary region of the
creep curve.
In sum, looking at the effects of the mesh size of grids on a
combination of permanent deformation and creep rate, the results
suggest that larger mesh size grids performed better only in the
initial stages of loading, whereas over the longer term, smaller
mesh size grids will outperform large ones with same the tensile
strength. These results showing small gird mesh sizes to perform
better than large ones with the same tensile strengths are similar to
those reported by some previous studies (Jenkins et al., 2004;
Komatsu et al., 1998).
4. Conclusions

Fig. 4. Measured and estimated permanent deformation for R2 sample.

Fig. 5. Measured and estimated permanent deformation for C, R3, & R4 samples.

Fig. 6 is a one-to-one graph of the measured versus estimated


values of permanent strain in the last cycle, including the intercept,
slope and correlation coefcients. Comparing the measured and
estimated permanent strain in the last cycle for each type of sample
(Fig. 6) and the improvements in the reinforced samples in Tables 3
and 4 as well as in Figs. 3e5, it can be seen that the measured and

The reinforcement of asphalt concrete with berglass grids is


one of the means to combat permanent deformation. Fiberglass
grids are manufactured with different tensile strengths and aperture (mesh) sizes. In this study, an attempt was made to study the
effects of ber glass grids with different tensile strengths and mesh
sizes, applied at the mid-depth of bi-layer asphalt concrete samples, on the resistance of these samples to permanent deformation.
Our results suggest that berglass grid reinforcement is
remarkably effective in lowering the permanent deformation of
asphalt concrete, probably due to the tensile forces and lateral
connement provided by such grids.
Secondly, our study conrms that Ahari's creep curve model can
be used with both berglass grid reinforced, and unreinforced hot
mix asphalt. In the secondary region of the creep curves, in which
the optimum density of the mixture is achieved, higher tensile
strengths and smaller mesh size result in gentler slope (meaning a
lower permanent strain accumulation rate).
Another conclusion from our results is that increasing the tensile strength of a berglass grid can lead to a reduction in permanent strain, depending on the type of grid used. Moreover, not only
is the tensile strength of a berglass gird effective in increasing the
resistance of asphalt concrete to permanent deformation, but the
mesh size of the grid is also of considerable importance. Our results
suggests that, of the grid mesh sizes used in this study, larger mesh
size berglass grids perform better than small ones only in the
initial stages of loading. This is because enlarging the mesh size
causes the creep rate to increase which eventually, in the longer
terms, outweighs the smaller deformation achieved in the rst
stages of the creep curve.
In conclusion, the range of experiments we carried out suggest
that smaller mesh sizes should provide more resistance to permanent deformation in the long run. This could be due to the

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

Table 5
Slope comparison of primary and secondary regions.
Code

C
R1
R2
R3
R4

Tensile strength
(kN/m) (MD  XD)

Grid mesh size (mm)


Center to center of strand

Primary region
Slope

Improved slope compared


to control sample (%)

Secondary region
Slope

Improved slope compared


to control sample (%)

e
115  115*L
/15
115  115*L
/15
115  215*H
/15
115  215*H
/15

e
12.5  12.5**S

1523.9
1114.5

0
37

0.2155
0.1336

0
61

25  25**OL

1075.8

42

0.1515

42

12.5  12.5**S

1062.9

43

0.1291

67

25  19**OL

954.6

60

0.1436

50

*L: Low level for grid tensile strength; *H: High level for grid tensile strength; **S: Small level for grid opening size; **OL: Large level for grid opening size.

Fig. 6. One-to one graph of measured vs. estimated permanent strain of last cycle.

greater number of bers per unit width in such smaller meshes


than in the larger mesh geogrids. This nding that the best resistance to permanent deformation can be achieved by asphalt concrete reinforced grids with greater tensile strength but also with
smaller mesh sizes is in line with what previous researchers have
reported.
Finally, our study shows that interpreting creep curves without
creep rate consideration can be misleading when the tertiary region of creep curves is not achieved in tests. Further research on
other types of asphalt mixtures, reinforced with other types of
grids, at various depths and under other testing conditions, is
recommended.
Acknowledgment
The authors would like to acknowledge the Ministry of Higher
Education of Malaysia for their nancial support under grant
number FP021/2011A.
References
Ahari, A.S., Forough, S.A., Khodaii, A., Moghadas Nejad, F., 2013. Modeling the Primary and the Secondary Regions of Creep Curves for SBS Modied Asphalt
Mixtures under Dry and Wet Conditions. J. Mater. Civil Eng. 26 (5), 904e911.
Appea, A.K., 1997. In-Situ Behavior of Geosynthetically Stabilized Flexible Pavement.
Polytechnic Institute and State University, Virginia.
ASTM, D3515-00, 2000. Standard Specication for Hot-Mixed, Hot-Laid Bituminous
Paving Mixtures. ASTM International.
Austin, R., Gilchrist, A., 1996. Enhanced performance of asphalt pavements using
geocomposites. Geotext. Geomembr. 14 (3), 175e186.

Barksdale, R.D., 1972. Laboratory Evaluation of Rutting in Base Course Materials. Paper
presented at the Third International Conference on the Structural Design of Asphalt
Pavements, Grosvenor House, Park Lane, London, England, Sept. 11e15, 1972.
Bertuliene, L., Oginskas, R., Bulevicius, M., 2011. Research of Rut Depth in Asphalt
Pavements Reinforced with Geosynthetic Materials. Paper presented at the 8th
International Conference Environmental Engineering, Lithuania.
Collin, J., Kinney, T., Fu, X., 1996. Full scale highway load test of exible pavement
systems with geogrid reinforced base courses. Geosynth. Int. 3 (4), 537e549.
Cominsky, R., Killingsworth, B., Anderson, R., Anderson, D., Crockford, W., 1998.
Quality Control and Acceptance of Superpave-Designed Hot Mix Asphalt.
NCHRP Report 409.
EN, C., 12697-33, 2003. Bituminous Mixtures. Test Methods for Hot Mix Asphalt.
Part 33: Specimen Prepared by Roller Compactor.
EN, C., 12697-25, 2005. Bituminous Mixtures e Test Methods for Hot Mix Asphalt
Part 25: Cyclic Compression Test.
Epps, A., Harvey, J.T., Kim, Y.R., Roque, R., 2000. Structural Requirements of Bituminous Paving Mixtures. Transportation in the New Millennium.
Fei, Y., Yang, Y., 2009. Fem analysis on geogrid reinforced asphalt concrete pavement. Geosynth. Civ. Environ. Eng. 677e682.
Ferrotti, G., Canestrari, F., Pasquini, E., Virgili, A., 2012. Experimental evaluation of
the inuence of surface coating on berglass geogrid performance in asphalt
pavements. Geotext. Geomembr. 34 (0), 11e18. http://dx.doi.org/10.1016/j.
geotexmem.2012.02.011.
Fwa, T., Tan, S., Zhu, L., 2004. Rutting prediction of asphalt pavement layer using
cef model. J. Transp. Eng. 130 (5), 675e683.
Gabra, R., Horvli, I., 2006. Simplied Testing Method for Evaluation of Asphalt
Mixtures for Their Susceptibility to Permanent Deformation.
Jenkins, K.J., Dennison, P., Ebels, L.J., Mullins, L.S., 2004. 3-D Polymer Grid Reinforcement of Asphalt for Rut Resistance. Paper presented at the Proceedings of
the 8th Conference on Asphalt Pavements for Southern Africa (CAPSA'04),
Southern Africa.
Kaloush, K., Witczak, M., 2002. Tertiary ow characteristics of asphalt mixtures
(with discussion and closure). J. Assoc. Asphalt Paving Technol. 71, 248e280.
Kalyoncuoglu, S., Tigdemir, M., 2011. A model for dynamic creep evaluation of SBS
modied HMA mixtures. Constr. Build. Mater. 25 (2), 859e866.
Kandhal, P.S., Chakraborty, S., 1996. Evaluation of Voids in the Mineral Aggregate for
HMA Paving Mixtures. Paper presented at the Proceedings of the Annual
Conference-Canadian Technical Asphalt Association.
Komatsu, T., Kikuta, H., Tuji, Y., Muramatsu, E., 1998. Durability assessment of
geogrid-reinforced asphalt concrete. Geotext. Geomembr. 16 (5), 257e271.
Laurinavi
cius, A., Oginskas, R., 2006. Experimental research on the development of
rutting in asphalt concrete pavements reinforced with geosynthetic materials.
J. Civ. Eng. Manag. 12 (4), 311e317.
Lee, K.W., Gundapuneni, S.K., Singh, A., 2010. Effects of Asphalt Binder Grade on the
Performance of Rhode Island Hot-Mix Asphalt.
Leshchinsky, B., Ling, H., 2012. Effects of geocell connement on strength and
deformation behavior of gravel. J. Geotech. Geoenviron. Eng. 139 (2), 340e352.
Ling, H.I., Liu, Z., 2001. Performance of geosynthetic-reinforced asphalt pavements.
J. Geotech. Geoenviron. Eng. 127 (2), 177e184.
Mehta, Y., Guercio, M.C., McCarthy, L., 2014. Determine Viscoelastic Mechanical
Properties of Warm Mix Asphalt (WMA)-Reclaimed Asphalt Pavement (RAP)
Mixes under High Stresses in Aireld Flexible Pavements and Its Impacts on
Design Life.
Monismith, C.L., Ogawa, N., Freeme, C.R., 1975. Permanent deformation characterization of subgrade soils due to repeated loading. Transp. Res. Rec. 537, 1e17.
Perkins, S., 1999. Mechanical response of geosynthetic-reinforced exible pavements. Geosynth. Int. 6 (5), 347e382.
Perkins, S., Ismeik, M., 1997. A synthesis and evaluation of geosynthetic-reinforced
base layers in exible pavements e part I. Geosynth. Int. 4 (6), 549e604.
Siriwardane, H., Gondle, R., Kutuk, B., 2010. Analysis of exible pavements reinforced with geogrids. Geotech. Geol. Eng. 28 (3), 287e297.
Sivilevi
cius, H., Petkevi
cius, K., 2002. Regularities of defect development in the
asphalt concrete road pavements. J. Civ. Eng. Manag. 8 (3), 206e213.

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003

S. Mirzapour Mounes et al. / Geotextiles and Geomembranes xxx (2015) 1e8

Sobhan, K., Genduso, M., Tandon, V., 2005. Effects of Geosynthetic Reinforcement on
the Propagation of Reection Cracking and Accumulation of Permanent
Deformation in Asphalt Overlays. Paper presented at the Third LACCEI International Latin American and Caribbean Conference for Engineering and Technology (LACCET2005) Advances in Engineering and Technology: a Global
Perspective, Cartegena, Columbia.
Tayfur, S., Ozen, H., Aksoy, A., 2007. Investigation of rutting performance of asphalt
mixtures containing polymer modiers. Constr. Build. Mater. 21 (2), 328e337.
Thakur, J.K., Han, J., Pokharel, S.K., Parsons, R.L., 2012. Performance of geocellreinforced recycled asphalt pavement (RAP) bases over weak subgrade under
cyclic plate loading. Geotext. Geomembr. 35, 14e24.
Thomas, G.B., Finney, R.L., Weir, M.D., Giordano, F.R., 2001. Thomas' Calculus.
Addison-Wesley.
Tseng, K.-H., Lytton, R.L., 1989. Prediction of permanent deformation in exible
pavement materials. In: Implication of Aggregates in the Design, Construction,
and Performance of Flexible Pavements, pp. 154e172. ASTM STP, 1016.

Tutumluer, E., Huang, H., Bian, X., 2010. Geogrideaggregate interlock mechanism
investigated through aggregate imaging-based discrete element modeling
approach. Int. J. Geomech. 12 (4), 391e398.
Verhaeghe, B., Myburgh, P., Denneman, E., 2007. Asphalt Rutting and Its Prevention.
Paper presented at the Proceedings of the 9th Conference on Asphalt Pavements for Southern Africa (CAPSA'07), Gaborone, Botswana.
Virgili, A., Canestrari, F., Grilli, A., Santagata, F., 2009. Repeated load test on bituminous systems reinforced by geosynthetics. Geotext. Geomembr. 27 (3),
187e195.
Yang, X., Han, J., Pokharel, S.K., Manandhar, C., Parsons, R.L., Leshchinsky, D.,
Halahmi, I., 2012. Accelerated pavement testing of unpaved roads with geocellreinforced sand bases. Geotext. Geomembr. 32, 95e103.
Zhou, F., Scullion, T., Sun, L., 2004. Verication and modeling of three-stage permanent deformation behavior of asphalt mixes. J. Transp. Eng. 130 (4),
486e494.

Please cite this article in press as: Mirzapour Mounes, S., et al., Evaluation of permanent deformation of geogrid reinforced asphalt concrete
using dynamic creep test, Geotextiles and Geomembranes (2015), http://dx.doi.org/10.1016/j.geotexmem.2015.06.003