Académique Documents
Professionnel Documents
Culture Documents
on
Performance Evaluation of Geosynthetic Reinforced Unpaved Roads
(CIST Phase III/CIST0026/2012-2013)
Submitted to:
CiSTUP
Indian Institute of Science
Bangalore 560 012
Investigator:
Prof. Gali Madhavi Latha
APRIL 2013
Keywords:
Unpaved roads, geosynthetics, cyclic loading, CBR tests, cyclic triaxial tests, field
tests, numerical analysis
Deliverables
-Bearing resistance of reinforced soil-aggregate systems in terms of load-penetration
response and CBR tables
-Cyclic deformation characteristics of subgrade materials in terms of stress-strain
graphs and modulus tables
-Results from field tests in terms of number of vehicle passes vs. deformation of road
sections and Traffic benefit ratio comparisons
-Results from numerical simulations
-Design guidelines for geosynthetic reinforced unpaved roads
Results, discussion and outcome of the project are presented in following sections in
the order of the deliverables listed above.
INTRODUCTION
A systematic series of unsoaked CBR tests were carried out on unreinforced
and reinforced soil-aggregate systems in the conventional CBR mould of 150 mm
internal diameter and 175 mm total height. These tests were carried out as per ASTM
D1883-07. The total height of the prepared soil or soil-aggregate systems was 125
mm in all the cases. A surcharge weight of 5 kg was applied through a steel plate of
50 mm thickness in all the tests. A plunger of 50 mm diameter was used for applying
the load. The resistance offered by the sample to the penetration of the plunger was
measured using a load cell.
MATERIALS USED IN EXPERIMENTS
In these experiments, two types of subgrade soils 1 and 2 (SS1 and SS2) are
used for preparing the soil layer and aggregate (A1) is used as sub-base course. For
filling the soil and aggregates in the mould, modified Proctor compaction effort was
used. Various reinforcing materials used in the experiments are geotextile, strong
geogrid and geonet. Physical and mechanical properties of these materials are
discussed below.
Subgrade Soil SS1
The grain size distribution curve of SS1 soil is shown in Fig. 1. SS1 soil is
classified as clay of low plasticity (CL) according to the Unified Soil Classification
System. Properties of SS1 are given in Table 1. Compaction curves for SS1 soil from
both standard and modified Proctor tests are shown in Fig. 2. The California Bearing
Ratio of the SS1 soil is computed as per ASTM D 1557 07. SS1 soil has unsoaked
and soaked CBR values of 30 % and 19 % respectively corresponding to modified
Proctor effort (optimum moisture content of 12.5 % and maximum dry unit weight of
18.3 kN/m3).The load-penetration curve of the SS1 soil from unsoaked and soaked
CBR tests is shown in Fig. 3. The shear strength properties of the soil were
determined from consolidated undrained (CU) triaxial compression tests on the soil
samples compacted to modified Proctor effort. The stress-strain response of SS1 soil
in CU test at three different confining pressures of 50, 100 and 150 kPa is shown in
Fig. 4. The soil showed effective cohesion of 45 kPa and effective friction angle of
25.5 as determined from CU test.
Fig. 1 Grain size distribution of the subgrade soils used in the experiments
Fig. 2 Compaction curves and zero air void line for subgrade soil 1
Fig. 3 Load-penetration curves from unsoaked and soaked CBR tests on SS1
soil
Fig. 5 Compaction curve and zero air void line for subgrade soil 2
Table 1 Properties of subgrade soils used in the experiments
Type of Soil
Colour
Specific gravity
Soil classification
Liquid limit, %
Plastic limit, %
Maximum dry unit weight (kN/m3) from standard
Proctor test
Optimum moisture content from standard Proctor
test (%)
SS1
Reddish
Brown
2.7
CL
36
22
SS2
Reddish
Brown
2.71
CL
36
24
17.2
18.24
15.5
15.5
AGGREGATE
Aggregate layer was used to simulate the base/sub-base course layer in the
unpaved road and was placed on top of the subgrade soil in experiments. Three
different types of aggregates (A1, A2 and A3) were used in the experiments and these
were obtained from a nearby quarry. All the aggregates were grey in colour. Major
difference between different types of aggregates is the variation in their grain size
distribution. The following section describes the aggregates used in experiments.
Aggregate 1 (A1)
A1 comprises of aggregates passing through 6.3 mm sieve and retained on
4.75 mm sieve. The grain size distribution curve of this aggregate is shown in Fig. 6.
The specific gravity of A1 type aggregate determined using pycnometer according to
ASTM C 128 -2012 was 2.65. CBR test was performed on this aggregate by filling it
in 5 layers in the CBR mould and it was observed that the aggregate has a CBR value
of 23 %. The maximum unit weight achieved by A1 aggregate during CBR test (at
modified compaction effort) was 16 kN/m3. The load-penetration response from CBR
tests on A1 aggregate is shown in Fig. 7. The irregularities in the load versus
penetration response indicate crushing of the aggregate during the test.
Aggregate 2 (A2)
A2 type of aggregate comprises of aggregates passing through 12.5 mm sieve
and retained on 6.3 mm sieve. The average size of A2 type aggregate is 10 mm. The
grain size distribution curve of this aggregate is shown in Fig. 7. This aggregate has a
specific gravity of 2.67. Photograph of A1 and A2 type aggregates is shown in Fig. 8.
The maximum dry unit weight achieved by A3 aggregate using wet method
was 20.6 kN/m3 (bulk unit weight of 21.96 kN/m3 ) at a water content of 6.7 %. The
maximum dry unit weight so achieved was 20.5 kN/m3 (bulk unit weight of 21.4
kN/m3) at a water content of 4.5 %.
Geotextile
The geotextile used in the experiments is a polypropylene multifilament
woven fabric. The individual multifilaments are woven together in such a manner so
as to provide dimensional stability relative to each other. The properties of geotextile
are given in Table 2. Ultimate tensile strength of the geotextile was determined by the
wide-width strip method as per ASTM D 4595 01. The geotextile has an ultimate
tensile strength of 55.16 kN/m in the warp direction. The mobilized tensile strength of
the geotextile material corresponding to 2% strain is 3.02 kN/m, and the
corresponding secant modulus is calculated as 151 kN/m. The load-elongation
response of the geotextile in the warp direction is shown in Fig.11.
Table 2 Properties of the geotextile
Breaking strength:
Elongation at break
warp
55.16 kN/m
weft
46.0 kN/m
warp
38%
weft
21.3%
Thickness
1 mm
230 gm/m2
Fig. 11 Load elongation response of the geotextile from wide-width tension test
11
Geogrids
Two varieties of biaxial geogrids (biaxially oriented integrally extruded
geogrids with rigid junctions and stiff ribs), made of polypropylene (PP) are used in
the present study. They are designated as strong biaxial geogrid (SG) and weak
biaxial geogrid (WG) in this thesis based on their ultimate tensile strength. Fig. 12
shows the nomenclature used to describe the dimensional details of a typical biaxial
geogrid. Dimensions of both the biaxial geogrids as per the nomenclature used in Fig.
12 are presented in Table 3. The tensile properties of the geogrids were obtained from
standard multi-rib tension test (as per ASTM: D 6637-01). The tensile strength of both
these biaxial geogrids with respect to strain as obtained from standard multi-rib
tension test is presented in Fig. 13.
AT
WLR
WTR
tJ
tLR
tTR
Unit weight
(Kg/m2)
SG
30
30
2.6
3.0
5.8
2.2
1.4
0.53
WG
35
35
2.3
3.0
4.1
1.4
1.1
0.22
Geonet
The geonet (GN) used in the tests is an extruded polymeric flexible mesh with
square openings of size 1.5 mm 1.5 mm, typically used for insect screens and is gray
in colour. The load elongation response of the geonet obtained from wide width
tension test is shown in Fig. 14. The geonet has sustained a peak tensile load of 7.6
kN/m at 2.4 % strain at which it failed. Table 4 summarizes the properties of various
geosynthetic materials used in experiments.
Geotextile
WG
SG
Geonet
3535
3030
1.51.5
55.16
26.4
38.1
7.6
38
16.50
16.7
2.40
151
219
588
319
230
220
530
125
13
EXPERIMENTAL SET-UP
Schematic sketch of the reinforced soil-aggregate system is shown in Fig. 15.
In these experiments, subgrade soil 1 (SS1) is used for preparing the soil layer and the
aggregate 1 (A1) is used as sub-base course. For filling the soil and aggregates in the
mould, modified Proctor compaction effort was used. A total of 52 CBR tests were
conducted under 13 different series and many of them were repeated to check the
repeatability of the test results. Details of different test series are presented in Table 5.
Tests in each series were conducted at four different water contents of the soil layer
i.e., 12.5 % (corresponding to Optimum moisture content), 14.5 %, 16.5 % and
18.5 %.
14
Series
No
Notation
Soil alone
SA
SAGT
SABG
10
11
Schematic
sketch
12
SABGGCBG
13
SAGNGCGN
value of the unreinforced and reinforced systems were estimated. In case of soil alone
systems the CBR value at 2.5 mm penetration was higher than that at 5 mm
penetration whereas for unreinforced and reinforced soil-aggregate systems, the CBR
at 5 mm penetration was higher than that at 2.5 mm penetration CBR values of
various systems computed at four different water contents are summarized in Table 6.
Table 6 CBR values for various soil-aggregate systems
CBR values (%) at
water content %
OMC
14.50
16.50
18.50
29.80
10.40
2.70
1.30
SA
36.70
13.20
5.60
2.80
SAGT
40.20
14.60
6.50
3.90
SAGTBG
47.60
16.90
8.40
4.60
SAGTGN
42.70
15.40
7.40
3.90
SAGN
37.70
13.40
6.50
3.70
SAGN2
40.20
14.90
7.40
3.90
SAGNGCGN
32.70
11.90
6.90
3.70
SAGTGCBG
50.60
17.90
8.70
4.70
SABG
37.70
13.50
8.30
4.71
SABG2
54.10
20.70
9.20
5.40
SABGGCBG
60.50
23.30
10.20
5.60
SAGTGCGN
33.20
12.60
6.50
3.60
Test Series
(1)
Where CBRr is the CBR value of the reinforced soil-aggregate system and CBRu is
the CBR value of the unreinforced soil-aggregate system at the same water content
and density.
CBR improvement factor for various reinforced systems was estimated and
the values are summarized in Table 7. An improvement factor of 1 implies that there
is no extra benefit in using geosynthetic reinforcement, whereas a value less than 1
indicates that the performance of reinforced system is inferior compared to the
unreinforced system. From Table 7 it is observed that the improvement factors
increase with water content for all the cases. Improvement factors below 1 for soilaggregate systems reinforced by geonets in geocell form at lower water contents
16
showed that these systems had an inferior performance compared to the unreinforced
systems. The reason for this low improvement factor could be attributed to the
rupture of geocells in these cases.
Table 7 Improvement factors for various reinforced soil-aggregate systems
Test Series
SAGT
1.10
1.11
1.16
1.39
SAGTBG
1.30
1.28
1.50
1.64
SAGTGN
1.16
1.17
1.32
1.39
SAGN
1.03
1.02
1.16
1.32
SAGN2
1.10
1.13
1.32
1.39
SAGNGCGN
0.89
0.90
1.23
1.32
SAGTGCBG
1.38
1.36
1.55
1.68
SABG
1.03
1.02
1.48
1.68
SABG2
1.47
1.57
1.64
1.93
SABGGCBG
1.65
1.77
1.82
2.00
SAGTGCGN
0.90
0.95
1.16
1.29
Fig. 18 Load-penetration response for unreinforced and planar reinforced soilaggregate systems at OMC and at a water content of 18.5%
The load-penetration response of the SAGN series was found to be on par with
SAGT series at low levels of penetration. For the test series carried out at OMC (Fig.
18a), after a penetration of 5 mm, SAGN exhibited inferior performance even
compared to unreinforced soil-aggregate system because the geonet got punctured
during the test due to its low tensile strength and in this case it was sandwiched
19
between the aggregate layer and stiff soil layer compacted at OMC. At higher water
contents, the soil layer became softer and no puncture was observed in the geonet
layer. Hence, the performance of SAGN series is better when compared to
unreinforced soil-aggregate system at higher water contents as seen in Fig. 18b.
However, at higher water contents, the geonet was getting clogged. The photographs
of the exhumed geonets after the tests are shown in Fig. 19. It is clearly seen in the
figure that the geonet is punctured in the test with OMC and remained intact in tests
with higher water content.
5.0
7.5
10.0
12.5
14.0
SAGT
1.17
1.08
1.05
1.13
1.12
1.18
SAGN
1.02
1.03
0.89
0.93
0.95
1.02
SABG
1.04
1.03
1.03
1.09
1.15
1.22
SAGTGN
1.22
1.18
1.08
1.00
1.00
1.11
SAGTBG
1.39
1.26
1.24
1.22
1.22
1.37
5.0
7.5
10.0
12.5
15.0
SAGT
1.25
1.39
1.44
1.49
1.58
1.62
SAGN
1.31
1.26
1.41
1.48
1.56
1.62
SABG
1.41
1.70
1.80
1.89
1.85
1.81
SAGTGN
1.28
1.39
1.47
1.67
1.71
1.79
SAGTBG
1.53
1.58
1.84
1.91
2.04
2.10
21
Fig. 23b summarizes the results obtained from the test series with geonet in
planar and geocell forms at OMC. The bearing resistance of the soil-aggregate system
was the maximum with geocell reinforcement. However, the difference in the
performance of the planar layers and geocells is not as significant as in the case of
biaxial geogrid reinforcement because the reinforcement is very weak and was
ruptured during the tests. Since there is no covering material, the load was directly
applied to the geocell joint and the geonet being weak it, got ruptured at very low
levels of penetration. The geocells were subjected to tearing along the joints, and
rupture initiated at the central portion and propagated outwards, showing the failure of
joints at several places. Exhumed geocells showed damage at many places. The base
geonet got punctured at lower water contents, as explained in earlier sections. The
performance of the geocells was quite similar at all water contents and at higher water
contents the geocells made of geonet were clogged. Photographs of geocells made of
geonet exhumed after the tests are shown in Fig. 24. These photographs clearly show
the rupture of the geocells at low water contents and the clogging of geocells at higher
water contents. Geonet failed to act as a separator and there was some mixing of the
aggregate and soil layers at higher water contents. Tests with geotextile basal layer
showed better performance because it was intact during loading and also acted as a
separator. For reinforcing unpaved roads, stronger grids which do not rupture within
the service load limits should be used.
Fig. 24 Photographs of geocells made of geonet exhumed after the tests at water
contents (a) OMC (b) 14.5 % (c) 16.5 % (d) 18.5 %
25
Effect of Soaking
Soaked CBR tests were carried out on unreinforced and reinforced soilaggregate systems at optimum moisture content to understand the effect of type of
reinforcement on the soaked CBR value. For this purpose, a total of 5 tests were
carried out viz., S, SA, SAGT, SABG and SAGN under soaked condition. The results
of soaked CBR tests carried out on unreinforced and reinforced soil-aggregate
systems are presented in Fig. 25. It is clear that the bearing resistance of soaked test
specimens is less compared to that of unsoaked test specimens. The exhumed geonet
sample after the soaked test on soil-aggregate reinforced with geonet (SAGN) showed
a rupture similar to the unsoaked test on an identical system. The increasing order of
performance of the various soaked tests were SAGN, SAGT & SABG. The
performance of SABG in soaked CBR test is almost similar to that of SAGT. The
order of performance improvement has not differed much in soaked and unsoaked
CBR tests.
26
unsaturated large diameter granular sub-base samples out of which 6 were static tests
and 5 were cyclic tests. All the tests were carried out at a confining pressure of 50
kPa. Details of the experiments carried out are summarised in Table 9.
Table 9 Details of large diameter triaxial tests carried out
Type of the test
Unreinforced (static & cyclic)
2 layers of strong geogrid (static & cyclic)
3 layers of strong geogrid (static & cyclic)
4 layers of strong geogrid (static & cyclic)
5 layers of strong geogrid (static)
Geocell enclosed sample (static & cyclic)
Notation
UR
SG-2
SG-3
SG-4
SG-5
GC
29
Edyn = d
(2)
31
the system showed higher modulus in the initial cycles (up to 2000) after which the
modulus dropped drastically. From the figures it is seen that the reinforced systems
had higher modulus compared to unreinforced systems. Among the reinforced
systems, geocell reinforced system has the least modulus. The 2 and 3 layer geogrid
reinforced systems had high modulus in comparison to 4 layer geogrid reinforced
system.
Aggregate
Grey coloured stone chips obtained from a near-by quarry were used as the base
course aggregate for the tests. The average size of the aggregate was 12 mm. An
aggregate layer of 10 cm thickness was placed directly over the subgrade soil in case
of unreinforced test sections or over the geosynthetic layer in case of reinforced test
sections. The unit weight of aggregate layer was maintained at 13 kN/m3 for all tests.
Surface Course
A leveled surface course layer of 5 cm thickness was placed above the base course
aggregate for better rideability of vehicle in experiments. The surface course layer
was prepared by placing in-situ dry soil and then rolling with sufficient quantity of
water to avoid dust and ravelling when a vehicle passes over it.
Reinforcing Materials
Three different types of geosynthetics, namely, a woven geotextile, a biaxial geogrid
and a uniaxial geogrid were used in different reinforced test sections. Two tests were
conducted with geocell layer reinforcement. In these tests, geocell layers were formed
at site using biaxial geogrid. In one test, tyre shreds were used as reinforcing layer.
Geotextile
The woven geotextile used in experiments is white in colour with negligible pore size
of <0.075 mm. The geotextile is made of polypropylene and the ultimate tensile
strength is 55 kN/m at an axial strain of 38%.
Biaxial Geogrid
The biaxial geogrid used in experiments is made from High Density Polyethylene
(HDPE). It is a stiff grid with square openings of size 30 mm 30 mm and the
ultimate tensile strength is 40 kN/m in both longitudinal and transverse directions at
an axial strain of 10%.
Uniaxial Geogrid
The uniaxial geogrid used in experiments is made from high strength polyester yarns
with black PVC coating. It is a stiff grid with rectangular openings of size 220 mm
17 mm and ultimate tensile strength of 60 kN/m in longitudinal direction and 40 kN/m
in transverse direction at an axial strain of 10%.
Loading Vehicle
A 4-stroke, 102 cm3, single cylinder scooter of dimensions 1765 715 1130 mm
with 1235 mm wheel base and ground clearance of 145 mm was used in experiments.
The weight of the vehicle was 106 kg.
CONSTRUCTION OF UNPAVED ROAD SECTIONS
In all these experiments, initially the unpaved road was constructed in stages.
First stage involved the preparation of soil subgrade to the required density and water
content. In the second stage, aggregate layer of 10 cm thickness was prepared to the
required density above the subgrade for unreinforced cases. For reinforced cases,
prior to the placement of aggregate layer, the reinforcing layer was placed above the
35
subgrade. After the base course was placed and leveled, a surface course of 5 cm
thickness was constructed using in-situ dry soil and rolled with sufficient quantity of
water and leveled. These different stages are illustrated in following subsections. The
original soil at the location was mixed with excess amount of water and made slushy
for a depth of 10 cm and leveled. This bed was left as such for at least 24 hours so that
the soil attained uniform consistency. Once the bed was ready, undisturbed tube
samples were taken at three different sections of the soft bed to determine the
placement water content and unit weight. For all the experiments, the water content
and unit weight were maintained as 30% and 17 kN/m3 respectively. Most of the
times, it required 2-3 trials of mixing to achieve these uniform values for all the tests.
For tests involving reinforced road sections, reinforcing layers were placed above the
soft soil subgrade before placing the aggregate base course. In case of geotextile and
geogrids, a geosynthetic layer was cut from the rolls and placed over the test section,
covering the entire test section. The longitudinal direction of geosynthetic layer was
coinciding with the length direction of the road for all the tests to achieve maximum
benefit.
In case of geocell reinforcement, initially a geotextile layer was placed over
the subgrade. A layer of geocells was constructed in diamond pattern at the site to a
size of 2 m 1 m using biaxial geogrid and anchor pins of 6 mm diameter and 10 cm
effective height and placed above the geotextile as shown in Fig. 33. Geotextile layer
was needed for this case to separate the subgrade and base course and to avoid mixing
of layers during vehicle passage. Tests were done with geocell layers of two different
geometries. The area of biaxial geogrid used to prepare the layer of geocell was 5.85
m2 in one case and 2 m2 in the other case.
density. Care was taken not to fill any geocell to the total height until the adjacent cell
was at least filled to half of the height, to ensure the proper shape of the geocell layer
with aggregate infill. The in-situ dry soil was mixed with 10% water and placed over
the aggregate layer to prepare a comfortable riding surface. The thickness of this layer
was maintained as 5 cm and it was levelled using a drop hammer of 5 kg mass falling
from a height of 450 mm on a square base plate of 150 mm 150 mm in size.
RUT DEPTH
ASTM: E 1703/E 1703M 95, defined rut depth as the maximum measured
perpendicular distance between the bottom surface of the straightedge and the contact
area of the gage with the pavement surface at a specific location. This is shown in Fig.
38. When vehicle is passed over the prepared road surface, the surface gets deformed,
forming ruts. Arrangements were made to measure the rut depth at 11 equally spaced
grid points across the width of the road at three sections spaced uniformly along the
length of the road. The schematic diagram showing the layout of the grid points
marked on the plan of the road section is shown in Fig. 39.
Testing Procedure
A scooter of mass 106 kg was driven by a person weighing 55 kg along the
centerline of the finished roadbed. The speed of the vehicle was maintained as 18 to
20 km/hr and the vehicle was passed in one direction only. The rut depths were
measured at all grid points after every 20 passes until 200 passes were completed.
Then it was passed continuously for 50 times more and the final rut depths were noted.
If the vehicle started skidding in any point of time, the test was stopped at that
particular stage and the corresponding number of passes and rut depths were noted.
The testing arrangement is shown in Fig. 40.
38
The next test was with biaxial geogrid placed along with the geotextile at the
subgrade and base course interface. This section was stable till 250 passes of the
vehicle, where the experiment was stopped because the rut was stabilized and there
was no visible displacement for increasing number of passes. The cross section profile
of the road for this test is shown in Fig. 43. The maximum depression as observed in
comparison to the initial ground surface after 250 passes were 92, 56 and 90 mm for
Sections 1, 2 and 3 respectively. As seen from Fig. 43, the heave and subsidence
were increasing with the number of passes but they were stabilized at 250 cycles in
this test.
Fig. 43 Cross section profile at various sections for Biaxial Geogrid reinforced road
In the next test, uniaxial geogrid was used along with geotextile for
reinforcing the road section. In this test also, the rut depths were increasing till 250
passes and beyond 250 passes, the increase in deformations were very minimum
hence the test was stopped after 250 passes. The cross section profile for this test is
shown in Fig. 44. This test section showed a depression of 76, 66 and 66 mm when
compared to the initial ground surface in Sections 1, 2 and 3 respectively. The next
two tests consisted of road section reinforced with geocell layer placed at the interface
of soft subgrade and aggregate base along with the geotextile layer. Geotextile layer is
used to perform the function of separator in these tests. In one test, the geocell layer
was formed using 5.85 m2 of biaxial geogrid and 150 connecting pins. The aspect
ratio of geocells for this case was 1. Results from this test are presented in Fig. 45.
Unlike in case of geotextile and biaxial geogrid, in case of road section reinforced
with geocell layer, there was no progressive change in the cross section profile with
the number of passes. The maximum heave and subsidence were observed within 100
passes and afterwards remained constant from 100 passes to 200 passes. This
behaviour is because the geocell layer acts as stiff reinforcing mat for the road and
40
supports the loads. Even the heave and subsidence observed for this case was
relatively small compared to other reinforced sections because geocell layer allows
uniform distribution of loads and reduces differential settlements. The maximum
depression as observed in comparison to the initial ground surface after 200 passes
were 73, 47 and 76 mm for Sections 1, 2 and 3 respectively.
Fig. 44 Cross section profile at various sections for uniaxial geogrid reinforced road
Fig. 45 Cross section profile at various sections for geocell (GC 5.85) reinforced road
41
In the next test, geocell layer was formed using 2 m2 of biaxial geogrid, which
is equivalent to the area of geogrid used in test with planar biaxial geogrid. Fig. 46
shows the cross section profile observed from this test. Even in this case, the
settlement was almost immediate and after that, remained constant with the increase
in number of passes. However, the maximum subsidence observed in this test at the
end of 250 passes was 100, 68 and 82 mm, which is relatively higher than the
previous test with geocell layer made up of 5.85 m2 of geogrid. This is because the
area of biaxial geogrid used for preparing this geocell is very less compared to the
previous test and the aspect ratio is 0.25 for geocells, making it less stiffer compared
to the geocell layer with cells having an aspect ratio of 1.
Fig. 46 Cross section profile at various sections for geocell (GC 2) reinforced section
Comparison of Rut Depth with Number of Passes
When no reinforcing material was used, in the control section, the vehicle was
able to pass only 17 times and thereafter it started skidding. The road is considered as
totally failed at that point. All reinforced test sections performed better than the
control section in terms of sustaining more vehicle passes for the same rut depth. The
test section reinforced with geotextile layer failed at 100 passes of vehicle, whereas all
other sections were in operating condition even after 250 vehicle passes. Comparison
of the performance of unreinforced and geotextile reinforced test sections is shown in
Fig. 47. As observed in the figure, the geotextile was efficient in increasing the
number of vehicle passes at failure to 100 against 17 for control section. As observed
from the figure, in the first and third sections, the rut depth observed for geotextile
reinforced section was significantly less compared to that observed in the control
section. However, in the central section, not much different was observed in the rut
depths.
42
Fig. 47 Comparison of rut depth for geotextile-reinforced road section with the
control section
When the road sections were excavated and seen after the test, there was
mixing of layers and intrusion of aggregate into subgrade observed for the control
section. In case of geotextile-reinforced section, the layers were separate even after
the test, demonstrating the role of geotextile as separator apart from providing
membrane support to the road as shown in Fig. 48. Hence the beneficial effect of the
geotextile layer is seen clearly in terms of arresting the mixing of layers as well as in
increasing the vehicle passes.
(a)
(b)
Fig. 49 Comparison of rut depth for geotextile-reinforced road section and Biaxial
Geogrid reinforced section with the control section
Fig. 50 compares the results from tests done with biaxial geogrid and geocell layer
made of biaxial geogrid of area 2m2. It should be noted that the quantity of grid used
in both these tests is same, i.e. 2 m2. From Fig. 50, it is observed that the geocell
layer in this case was not much effective in reducing the rut depth, except in Section 1.
The reason for this is the low aspect ratio (0.25) for geocells in the layer. The pocket
size being more, the cells are not effective in holding the aggregate. Whereas the
geogrid layer, being continuous throughout the road section, provided better support
and effective friction development at the interface.
44
For all the sections, the TBR value of geocell layer (5.85 m2) was the highest
among all the reinforcements that have been used. Hence it is the most
efficient reinforcement. The TBR of geocell layer (5.85 m2) at a rut depth of 71
mm is 16.5, whereas that of biaxial geogrid is 6.5. Hence the geocell layer
(GC 5.85) is around 61% more efficient than biaxial geogrid in reducing rut
depth.
The TBR value of shredded tyre layer is greater than that of the geotextile layer.
Hence as far as traffic benefit ratio is compared tyre shreds are a better option
than geotextile. However, in case of tyre shreds, initially the TBR value was
low and started increasing with the passes. The reason for this could be the
initial compression of the shredded tyre. Hence shredded tyre forms more rut in
the Section 1 than unreinforced section.
45
46
FLAC uses an explicit finite difference formulation to find solutions to the dynamic
equations of motion for the specific problem to be analyzed. This process cycle of
FLAC in arriving at a solution to the problem is repeated until force equilibrium is
reached (Fig. 52).
using large strain mode. The concept of using large strain analysis is justified because
the large strain represents the deep ruts that are allowed in unpaved roads. In case of
reinforced road sections, soil-geogrid and geogrid-base contacts are governed by an
interface that has a behavior, elastic perfectly plastic of Mohr-Coulomb. Boundary
conditions of the problem are given in Fig. 53.
15.6
Poissons ratio
0.33
10
30
Base Properties
Unit weight, kN/m3
22
Poissons ratio
0.25
50
48
Cohesion, kPa
40
20
Reinforcement Properties
Reinforcement material
Weak geogrid
Geonet
183
1151
Thickness, mm
1.1
0.275
Poissons ratio
0.3
0.3
Weak geogrid
Geonet
3.58E+04
3.65E+04
Cohesion, kPa
0.1
21.38
28
33
Interface Properties
Reinforcement material
Stiffness per unit area k, (kN/m3)
100
200
300
400
500
600
700
0
-2
-4
)
%
( -6
B
/
,
h
t -8
p
e
D
t -10
u
R
d -12
e
zi
l
a -14
m
r
o
N-16
Unreinforced
Weak Geogrid
Geonet
-18
-20
Fig. 55 Load bearing capacity of unreinforced and reinforced unpaved road sections
Parametric Studies
Parametric numerical studies are carried out on the unpaved road section, varying the
stiffness of the reinforcement, number of reinforcing layers and undrained cohesion of
the soil layer below. Results are presented in Figs. 56-58. From these results, it is
clear that the stiffness of the reinforcement has significant effect on the load carrying
capacity of pavement sections. It is the most influencing parameter in the analysis.
50
Fig. 57 Effect of undrained cohesion of soil layer on the load bearing ratio
51
200
400
600
800
1000
1200
0
-2.5
)
%
( -5
/B
,
-7.5
o
ti
a
R-10
h
t
p
e
D
-12.5
t
u
R
-15
Unreinforced
N=1
N=2
N=3
-17.5
-20
Fig. 58 Effect of number of geogrid layers on the load bearing pressure of the road
section
52
REFERENCES
1.
2.
53
Journal Publications
Asha M Nair and Madhavi Latha, G. (2012) "Taming of large diameter triaxial
setup", Geomechanics and Engineering, Techno Press, Vol. 4, No.4. pp. 251262.
International Conference
National Conferences
54