Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
ceymchen@polyu.edu.hk
environment, there are various problems associated with the use of conventional steel
reinforcement bars as soil nails. For loose fill slopes or clay slopes, the development
of adequate nail bond strength is another practical issue which should be considered.
Carbon fiber reinforced polymer (CFRP) and glass fiber reinforced polymer (GFRP)
in several forms and installation methods have been studied as the alternatives to the
classical steel bar. Extensive laboratory tests on the materials and field tests on
different types of soil nails with different methods of installation have been carried
out in Hong Kong, Korea and Australia. The test results have supported the use of
these materials with innovative installation method as soil nails under different
geological conditions, and the suitability and performance of these materials under
Keywords: Soil Nail; Glass Fiber; Stress Transfer; Pullout Test; Numerical Modeling
1. Introduction
Soil nailing is first used in Hong Kong to provide support for highly weathered
zones until the 1980s (GEO, 2008). Currently, high yield steel bars are commonly
used as soil nails to stabilize slopes in Hong Kong and many other countries. However,
1
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 2 of 38
there are several drawbacks for the steel bar nails which are particularly critical in
Hong Kong and some other Asian countries. Soil nailing in very low shear strength
soil may require a very high soil nail density which is uneconomical. The use of steel
measures (double corrosion system), which add additional cost to the soil nail are
structures. The use of sacrificial thickness for conventional steel bar is now a general
practice in Hong Kong (GEO and HKIE 2011), and expensive corrosion protection
measures similarly to that for the anchor system are also adopted for corrosive
ground.
In Hong Kong, large diameter steel bars (32 mm to 40 mm) with a length
highways. Due to the space constraints, it is difficult to maneuver these heavy steel
bars on site. The steel bars are hence divided into several shorter segments (3-5 m) for
easy transportation, and couplers are used to connect the bars to the required total
length. This arrangement is not common in other countries but is quite popular in
Hong Kong. The time and cost of soil nail stabilization are thus extremely high for
slopes with difficult access in Hong Kong. Many engineers also have reservation on
the nail bond stress transfer within loose fill slopes or soft clays. In particular, it is
found that even though good compaction has been carried out in loose fill, the
compacted dry density of the fill can decrease with time, possibly due to washout of
unpleasing concrete grillage is commonly used in loose fill/clay slopes in Hong Kong,
Taiwan, China, Japan and other places, and this option will further increase the total
cost and time for soil stabilization . Currently, there are many research works in the
use of alternate soil nail materials with different installation methods in different
countries. In particular, for soft ground tunneling in clays where there is only limited
working space, tunnel face stability is difficult to be maintained, and the uses of
2
Page 3 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
The desired features of a good soil nail for a site with difficult access include: (a)
light weight and high strength; (b) a good bond strength can always be maintained in
all types of soil; (c) minimum corrosion problems (for permanent nail); (d) acceptable
cost; and (e) ease of construction – handling, joining and cutting. However, there is
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
not a single material which can fulfill all of the above requirements. To investigate the
various advantages and limitations of different soil nail materials and systems, the
authors have carried out several research studies in different types of soil in Hong
Many works have been carried out in investigating and improving the
performance of soil nails. The direct shear test is used for characterization of the
failure mode of nailed soil mass and the bond between the reinforcement and the
grout (Miller et al. 1999, Chajes et al. 1996, De Lorenzis et al. 2001). In general, they
have found that the controlling factor is the failures at the interface between the
reinforcement and the grout, while the strength and stiffness of the reinforcement are
not usually critical in the tests. De Lorenzis et al. (2001) have reported that the bond
strength is not influenced by the bonded length or by the width of the FRP composite
sheet. Jeffries (2004) has carried out further research and found the bond strength to
(1996, 2002) have carried out some pilot works for the bond strengths of FRP and
CFRP bars as rebars and grouted anchor. There are also research works about the
time-dependent behavior of GFRP nail by Li et al. (2013) and Zhang et al. (2015).
Cheng and Wei (2007) and Cheng et al. (2009) have carried out large scale pilot
studies (laboratory and field tests) on the use of glass fiber reinforced polymer pipe
(GFRP) with the Tube a’ Manchette grouting method as a soil nail material in Hong
Kong and Korea. Another large scale study involving conventional steel bar, glass
fiber reinforced plastic rod, carbon fiber reinforced polymer rod (CFRP) and GFRP
pipe has been carried out in Hong Kong, while the use of fracture grouting combined
3
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 4 of 38
with the use of GFRP rod has been applied in the Brisbane Airport Link project in
Australia (Cheng et al. 2013), with Fleming Award in 2011 and Ground Engineering
Tremendous field and laboratory tests on different soil nail materials and the
performance of the soil nails under different installation methods are summarized and
Currently, proprietary GFRP soil nail is available in the market (Aslan 2011 and
other), and the KGC slope in Hong Kong is stabilized by GFRP soil nail using
conventional grouting method. There are however only limited trial study in using
CFRP as soil nail. Based on only laboratory tests, Toufigh et al. (2014) have
suggested that CFRP can be a suitable soil nail material. The preliminary tests by
Cheung and Lo (2005) on CFRP soil nail have indicated that the material is brittle and
care should be exercised in the actual application. Unwin (2001) has commented that
the properties of the CFRP nail may be affected by the potassium ion in the soil, but it
appears that the engineering life of CFRP nail can be as long as 120 years for normal
condition. So far, there are only limited real applications of CFRP soil nail. In general,
these FRP based soil nails are installed using the conventional method, and it is
possible to extend and improve the use of FRP soil nail using refined methods which
The objectives of this paper are to investigate the properties of different soil nail
materials, the problems in installation and testing, the effect of different nail
installation/grouting method and bond stress distribution along the nail during pullout
process through vast amount of laboratory and field tests. The soil nails adopted in
this study are made of different materials and configuration which include steel
reinforcement bars, GFRP bars and pipes, and CFRP bars. The soil nails in the present
project can be classified into two major groups: (1) soil nails grouted without applied
4
Page 5 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
pressure and function simply as a classical soil nail and (2) soil nails grouted with
pressure (relatively high to very high pressure) with an additional function in ground
improvement besides enhancing the nail bond stress. For soil nailed grouted under
pressure, the field tests are further divided into two groups : (1) permeable loose
granular fill and (2) impermeable soft clayey soil. Such division is necessary because
mechanism. In loose granular fill, permeation grout will be dominant while fracture
grout will be dominant for soft clayey soil. The field evaluations include the
2. Laboratory tests
Basic laboratory pullout tests carried out in Australia, Hong Kong and Korea are
introduced and discussed in this section. In Hong Kong, the use of expansive grout
has also been considered which is absent in the laboratory tests in Australia and Korea.
Steel bar, GFRP bar and pipe and CFRP bar have been tested in the present study.
Group A and C nails are GFRP nails with normal grout and expansive grout while
group B and D nails are CFRP nails with normal grout and expansive grout. Group E
nails are conventional steel bar nails with normal grout. The test results are shown in
A series of laboratory tensile and pullout tests are carried out for the
conventional steel bar, CFRP bar, GFRP bar and GFRP pipe, and the results of the
laboratory tests are shown in Table 1. For material tests, tensile tests to the vinyl ester
GFRP pipe are performed according to ASTM D638 “Standard Test Method for
Tensile Properties of Plastics”. For tensile test of steel bar, it is tested according to
5
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 6 of 38
Hong Kong standard CS2:2012. It should be mentioned that for the GFRP pipe, it is
described by Cheng et al. (2009). Furthermore, for the FRP soil nail, it is found from
the laboratory tests that there is not an obvious yield stress and yield strain which is
During the tensile testing for CFRP and GFRP nails, it is found that slippage will
occur between the relatively smooth CFRP/GFRP surface and the hydraulic jack when
the classical gripping method is used. This problem has been reported by Cheng et al.
(2009) so that slippage will occur before the ultimate yield stress/strain is reached,
and the ultimate yield stress/strain will not be obtained then. To overcome this
problem in laboratory and in field, the FRP are bonded to steel pipes by using epoxy.
The hydraulic jack then applies the loads on the steel pipe instead of adding the load
to the FRP rods directly. Since the steel pipes are stronger than the FRP nail and will
not fail during the test, the ultimate yield stress/strain of the FRP material can then be
obtained. Through such arrangement, the ultimate load on CFRP/GFRP nail is then
determined.
A test box (2m in length, 2m in width and 1.6m in height) completely filled with
decomposed granite (CGD) as shown in Fig.1 is used for the pullout test. According
to the Proctor Test, in accordance with BS 1377-Part 4-1990, the maximum dry
density of the soil varies from 1801 to 1867 kg/m3. The optimum moisture content
varies between 12 to 15%. Vertical pressure is applied to the box to simulate the
hydraulic jacks mounted on a top rigid plate placed on top of the fill in the test box.
Since CDG has a relatively high permeability (in the order of 3-5x10-6 m/s), the
loading is applied only for half an hour which is sufficient for the soil to stabilize. The
pullout machine is inclined at an angle of 15° (same as the inclination of the soil nail)
during the pullout tests. The embedded length of the soil nail is 1650 mm, and the
6
Page 7 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
length of the soil nail protruded outside the test box is 450 mm. Five Linear Variable
Differential Transformers (LVDT) and seven pressure sensors are installed at the front
plate, as shown in Fig.1. Readings are taken at 15 seconds interval. The pullout rate is
maintained at 0.5 mm per minute. The locations of the strain gauges along the soil
nails are shown in Fig.2. The steel nails are grouted under gravity with cement grout
The two important factors controlling the design of GFRP pipe soil nail are the
tensile strength of GFRP and bond strength between the nail and the soil. Lazarte et al.
(2003) state that the pullout capacity of any nail is a function of the ultimate bond
strength, drilled hole diameter and effective pullout length. The pullout capacity (Rp),
is mobilized when the ultimate bond strength is achieved and is expressed as eq.(1):
[1] Rp=Tmax=QuLp
In eq.(1), Lp stands for the effective pullout length beyond the failure surface,
[2] Qu=πquDDH
where qu denotes the ultimate bond strength; DDH is the effective diameter of the drill
hole.
For the material tests, tensile tests for the vinyl ester GFRP pipe are performed
according to ASTM D638 “Standard Test Method for Tensile Properties of Plastics”.
Vinyl ester bonded GFRP pipe is found to fail in a brittle mode. Beyond an axial
strain of 1.8%, the GFRP pipe fails suddenly with a very small increase of the strain.
The stress-strain characteristics of the GFRP pipe are, however, dependent on the
resin used to embed the glass fibers. The failure mode of urethane bonded GFRP pipe
(more expensive) which has higher tensile strength is much more ductile than that of
the vinyl ester bonded GFRP pipe from the laboratory tests in Korea, but it is not
7
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 8 of 38
considered in the present study due to the cost consideration (unlikely to be adopted
For the laboratory pullout tests of the pressure grouted residual soil Seoul granite
in Korea which is shown in Fig.3, the length of the sample adopted is 250 mm while
around 10 bars during grouting. Different confining pressures are applied to the
Some of the non-destructive laboratory pullout tests in Hong Kong are given in
Fig. 4. Compared with Fig.4a and 4b, it is noticed that the use of expansive grout has
greatly improved the ultimate bond strength as well as the elastic range with little
plastic deformation under pull-out test. The gradients of the pullout force against
pullout displacement however appear to be not strongly related to the use of expansive
grout (see Fig.4a, 4b and 4c), and this relation is probably mainly controlled by the
elastic property of the surrounding soil. On the other hand, this gradient appears to be
strongly influenced by the use of expansive grout if CFRP nail is used. This is
possibly due to the fact that CFRP bar is much smaller in size, and the percentage
effect of the grout expansion on the compaction of the surrounding soil will be more
appealing if the initial volume under consideration is smaller in size. This new finding
In this group of tests, nails A and B are CFRP bars while nails C and D are steel
bars. The surface of nails B is roughened in order to increase the bond stress. For
material testing, the CFRP is also tested in accordance with ASTM D638. The CFRP
bar in test has a diameter of 12 mm with sand coated surface, and each bar is cut into
2000 mm in length. Two types of grout mix were used in the test: ordinary cement
grout with a water-cement ratio of 0.45 and expansive cement grout with a
water-cement ratio of 0.4 (on nail C as shown in Fig.5 and 6). The pullout tests are
8
Page 9 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
carried out in typical Hong Kong sandy soil CDG using a destructive pullout test and
non-destructive pullout test. In the destructive pullout test, the soil nail is allowed to
move for 25 mm to determine the maximum pullout forces. The non-destructive tests
are divided into 3 stages according to the typical soil nail testing procedure in Hong
Kong. The first cycle is 1/3 of the estimated maximum pullout force and the test load
is held with an hour. The second cycle is 2/3 of the maximum pullout force and the
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
load is also maintained for an hour. Finally, the designed maximum pullout force is
applied and the load is held for another one hour before it is released. The results of
The results for the destructive pullout tests for Soil nails A and C given in Fig.5.
From observation, it is clear that there is no slippage between the CFRP rod and the
cement grout. Failure is observed at the connection between the anchorage tube and
the weakened CFRP connection. As such a connection is used only for laboratory test
and will not be used for actual site implementation, the pullout strength of the CFRP
rod can be much greater than the test results of 40 to 50kN (nail C) as obtained in the
laboratory test. The bond strength between the CFRP rod and cement grout is strong
and the failure mode is the same as the steel soil nail where failure is initiated at the
interface between the soil and cement grout. There are, however, differences in the
development of skin friction between the two groups of nails, which will be discussed
in a later section.
For the non-destructive testing of nails as shown in Fig.6, it is noticed that soil
nail D which is subjected to the cycled pullout test in a saturated situation gives lower
(saturated) and nail E (unsaturated) are however basically the same, which indicate
that saturation alone has only limited effect on the stiffness of the soil nail system.
Nail D exhibits higher pullout strength as compared with nail E, and this is possibly
due to the fact that unsaturated soil exhibits higher apparent soil parameters. This
result also indicates that the stabilization effect of soil nail will depreciate with soil
9
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 10 of 38
In order to examine the axial load distribution, the corresponding data are taken
and summarized in Fig. 7. For both steel and CFRP nails, the axial stress is mobilized
from the nail head. However, the load distribution along the first 20 percent length
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
varies a lot among different nails: stress keeps at a “nearly” constant value for nails C
and E, while the stress drops rapidly for CFRP nails. This special phenomenon is
possibly due to the high elastic modulus for steel soil nail. It is also noticed from Fig.5
that the pullout force experienced by Nail A (steel) is greater than Nail C (CFRP)
under the same pull-out displacement, which is due to the higher elastic modulus of
steel as compared with CFRP as shown in Table 1. It is interesting to note that even
though the elastic modulus of steel is about twice that for CFRP, the results between
the steel nail and CFRP nail do not differ by twice in Fig.5. The elongation actually
depends on the combined modulus of the nail and the effective grouted zone, and the
smaller CFRP nail has a higher percentage effective zone which is possibly due to the
expansion action of the grout. The peak pullout force of Nail C can reach over 50kN
which is about 25% more than that in Nail A (Fig. 5), which is also possibly due to
the expansive action of the grout compacting the surrounding soil so that a larger
mass is required to be pulled out at failure. The importance of expansive action of the
grout compared with the contractive action of normal grout is clearly illustrated in the
present results.
If the elastic and plastic elongation is considered, it is found that the CFRP nail
actually gives lesser permanent elongation even though the elastic elongation is
greater than the steel nail, as indicated in Table 2. In this respect, CFRP nail appears to
perform even better than steel nail from the results in Table 2 and Fig.5, even though
After the laboratory tests, a series of field pullout tests are carried out in Hong
Kong, Korea and Australia in sandy soil and clay using different types of nails
(Steel/CFRP/GFRP) and grouting methods. The objective of the field pullout and soil
strength tests are to verify the design assumptions about the bond strength at the
interface between the ground and the cement grout and the improvement in the soil
properties, so that a better and more economic design of soil nails can be obtained.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Soil nails for field pullout tests require partial grout to form the bonded length.
A field pullout test programme is carried out at four existing slopes in Hong
Kong which is approximately 65m in length and 36m in height. The cut slope is
dipping towards the northeast: two lower faces are inclined at approximately 55° to
the horizontal; two upper faces are inclined at approximately 45° to the horizontal.
Site investigation results reveal that the slope is composed of completely decomposed
granite (CDG) and localized colluvial deposits at the slope crest, and their properties
The average value of the Standard Penetration Test (SPT) for colluvium is
approximately 10, while for completely decomposed granite, the value varies between
8 to 65. Two types of cement and two kinds of FRP rod are tested separately in this
test programme. Details of the cement mixes are summarized in Table 4 while the
basic properties of the FRP nails are summarized in Table 5. The field test results of
the soil nails are summarized in Table 6. The strain gauges are installed at 0.2, 0.6,
1.0, 1.4, 1.8 and 2.6 m from the lower end of each nail as shown in Fig. 8. It should be
noted that some strain gauges are not working properly during the measurement as
shown in Table 6, which are possibly due to the relatively poor workmanship in strain
gauges fixing and the disturbance effect of grouting. In fact, this problem is
The multi-stage pullout test is carried out in accordance with the typical
11
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 12 of 38
procedures as shown in Fig.9. Five percents of the maximum proposed test load (Ta)
is originally applied to the nail. Then, the soil nails are loaded in three equal stages to
the maximum test load. In each stage, the load is maintained constant for 60 minutes.
The outwards displacement near the nail head and the strain gauge readings are
recorded at a 5-minute interval. The maximum testing loads for the soil nails are
tabulated in Table 6. Besides that, another group of field tests on GFRP pipe nails has
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
also been carried out by Cheng et al. (2009), and the advantages and limitations of
these GFRP pipe nails have been discussed. Summing up the experience gained in
Hong Kong, the use of CFRP as soil nail is not considered to be a good solution
because:
2. For CFRP, as the bar diameter is small and the surface is relatively smooth, it is
not easy to maintain a good bonding between cement grout and the CFRP.
Furthermore, to achieve the same tensile load as a steel nail, up to 4 CFRP bars
may have to be used. It is also found that different CFRP bar has taken up
different portions of the nail load during the pullout testing. It can be concluded
that it is not easy to ensure a uniform distribution of loading among these four
3. The shear strength of the CFRP bar is low, and there is shattering of the bar with
the fibers separated from each other in one of the unsuccessful pullout test. It
appears that the shear failure of CFRP has not been reported previously, but it has
been found from the present study. For this case, it is believed that the pullout
the setup of the equipment which has generated a shear force during stressing.
The axial force distributions along some of the soil nails (with a constant 2.6m
grouted length) in field pullout tests are shown in Fig. 10. The soil nail forces remain
constant initially, as there is no grout for the initial portion of the soil nails. It should
12
Page 13 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
be noted that the grout is applied by simple gravity flow without the use of any
pressure in this test programme. It is interesting to note that X1 and X2 with a smaller
inclination of 20° have the highest pullout strength while X3 and X4 with higher
inclination of 30° have the lowest pullout strength. Nail X8 with an inclination of 30°
also has a high pullout strength, which is possibly due to the use of expansive grout in
the installation.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Field pullout tests have also been conducted on silty sandy soils in Korea and
Hong Kong using GFRP pipes and two-stage grouting. Specifically, conventional
gravity grouting is used in stage one, and pressure grouting in stage two. The system
an average increase of the cohesive strength to the soil by 18 kPa, which is a major
improvement to the soil properties. Besides that, Cheng et al. (2009) have also noted
that there is a major increase in the elastic modulus “E” and one-dimensional
reflect that the use of Tube a’ Manchette grouting has a very major improvement in
Fracture grouting combined with the use of GFRP soil nails (Geonail system) for
maintaining the tunneling face stability is first proposed and adopted by the authors in
the Airport Link tunnel project in Brisbane, Australia (Cheng et al. 2013). Using this
innovative fracture grouted GFRP soil nail, the tunnel has been constructed through
soft clay while maintaining the train passage above the clay embankment with
controlled settlement during construction. This project has received two technical
For the GFRP bar nails in Australia (Cheng et al. 2013), the bar has a nominal
13
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 14 of 38
dimensions of 12.7 mm x 6.4 mm and a special “D” shape (instead of being circular
as used in Hong Kong) to enhance the bond strength. The basic properties of the
compressive strength 300MPa; tensile strength 300~550MPa. The nails are casted in
concrete and tested according to the guidelines of American Concrete Institute, ACI
Besides various laboratory tests, six GFRP test nails are also installed and tested
on site in either firm or soft soils. The pullout resistances of the 4 pullout tests with
fracture grouting in soft clay (average pullout resistance of the Geonails is 85kN/m)
are found to be higher than the design requirement. On the other hand, the pullout
resistance of the geonail without fracture grouting is 12% less than that of the design
pullout resistance of the grouted Geonail. The result has clearly demonstrated the
importance of fracture grout in the Geonail system. There are dual effects of the
Geonail system which are improvement to the soil properties and stabilization of
tunnel face by the nail and grout finger. Compared with the field installation method
in Hong Kong where only the effect of stabilization can be achieved, the use of high
pressure fracture grouting at increased expenses may require less nail for stabilization.
A direct comparison of the cost benefit cannot be carried out at present because there
expected that the overall cost of construction will be similar for the two systems, even
though the Geonail in Australia will be more expensive than a traditional nail using
simple grouting method if only a single nail is considered. The use of fracture
grouting however has the advantage that in very loose/soft material, it is possible to
eliminate the use of concrete grillages as the soil properties are improved. For the
Tube a’ Manchette grouting method as used in Korea and the fracture grouting
method as used in Australia, the former grouting method is more suitable for sandy
soil while the later method will be better for clayey soil. The simple grouting method
as used in Hong Kong is more suitable for good and stable sandy soil. The suitability
14
Page 15 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
of the grouting method depends more on the permeability and strength of soil, and
based on this issue, engineers can choose a suitable grouting method and type of
The actual measured bond resistances of 73 kN/m and 55 kN/m in firm clay and
soft clay are higher than the prediction, despite the fact that the improved soil
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
strengths are generally slightly less than that as predicted. The greater improvement in
the nail bond resistance is attributed to the contribution of the grout network. Nail
pullout test results indicate that most of the nails satisfy the acceptance criteria under
the designed ultimate load condition. From the results using Tube a’ Manchette
grouting in GFRP pipe nail in Korea as given by Cheng et al. (2009) and the present
results using fracture grouting in Australia, it is clear that the use of pressure in
grouting can improve the soil properties which can reduce the number of nails to be
Besides the laboratory and field tests, vast amount of numerical modeling for the
pullout tests have also been carried out, and only limited results will be discussed in
this section. The method of numerical modeling for some of the selected nails from
the present study is basically similar to that by Cheng et al. (2009) and Wei and
Cheng (2010) using the program FLAC3D, and the details of the numerical modeling
will not be repeated here. The slope and the soil are modeled by brick elements
according to the field geometry, and a typical example is shown in Fig.11. Since the
soil nail is small in the actual size, a domain width of 1m as shown in Fig.11 will be
sufficient to model the side effect (as adopted by Cheng et al. 2009 and Smith and Su
1997 and others). The actual nail spacing is usually around 1.5m which is much
greater than the size of nail by more than 10 times, nail-nail interaction is hence
usually not required to be considered. The input parameters include the nail material
15
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 16 of 38
well as the soil parameters for individual nail location. It should be noted that the
grout properties are only assigned to the grouted portion. The grout stiffness and grout
cohesive strength are the properties that relate to the shear interaction between the
grout and the soil. Therefore, for the ungrouted portion, only air exists instead of grout
between the bar and the soil. The grout stiffness and grout cohesive strength in these
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
For the modeling of the pullout tests, a uniaxial velocity is applied to the nail
head. By increasing the time steps (cycles), the pullout rate of the nail increases
steadily at a magnitude of 1x10-6 m per step. With the development of nail head
displacement, axial force in the nail is increased and bond stress is transferred to the
surrounding soil. Results obtained at the intermediate load stages are recorded and
compared with those from field pullout tests. Since many comparative results are
available, only limited results for steel bar, CFRP and GFRP nails are shown in Fig.
12 and 13 for illustration. From these two figures, it can be observed that if reasonable
material parameters are used, the results from numerical simulation can reasonably
match well with the experimental outputs. It should be noted that a perfect match
between the numerical and experimental result is rare in reality, due to the variations
with only limited regions of constant parameters and other factors. It is however noted
that there are some minor fluctuations of measured results near to the end of the nail.
Such minor fluctuations are possibly due to: (1) measurement error, in particular, the
strain level is usually small near to the end of the nail and percentage of measurement
error is usually great and (2) adhesion problem of the strain gauges, and some
unsuccessful tests not shown in this paper are also due to this adhesion problem of
strain gauges. Furthermore, since an axial force is applied at the nail head, all
movements should be in the same direction, hence the nail load cannot increase within
a short region near to the end of nail. This kind of measurement error can be
16
Page 17 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
From numerical modeling and field test, the skin friction at the grout/soil
Fi − Fi +1
τs =
[3] l × ∆x
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
where F is the axial force within the nail, ∆x is a small distance along the nail and l is
the circumference of the grouted zone. It can be observed that the maximum skin
friction occurs at a small distance after the start of the grouted zone which is shown in
Figs. 15 to 17. When a load is applied to the nail bar, a small distance is needed for
the mobilization of the skin friction. The results are in agreement with the previous
Looking at the skin friction development within steel, GFRP and CFRP nails, an
interesting phenomenon is found as follows. For steel nails, there is a rapid increase of
the skin friction with distance within the grouted zone. After the peak skin friction,
the skin friction decreases rapidly with distance for the steel nail. On the other hand,
for CFRP and GFRP where the Young’s modulus of the material is lower than steel,
this phenomenon appears only when the axial load is small (46.47 kN in Fig.15). For
higher nail axial load, the peak skin friction appears to maintain for some distance
before the skin friction drops for CFRP and GFRP. The clear drop in the skin friction
after the peak stress for the steel nail as shown in Fig.16 is not obvious for the CFRP
and GFRP nails. This phenomenon which is not mentioned in other previous studies
GFRP/CFRP so that the skin friction can be fully mobilized over some distance after
the peak. The present analysis has also demonstrated the applicability in using
numerical method to carry out the initial assessment and design of FRP nails.
Although the basic stress distribution from field tests can match with that from
numerical analysis, there are also noticeable differences between the two results. Such
17
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 18 of 38
discrepancies are actually normal due to various reasons: nonhomogenity of the soil
medium, the precise values of the input parameters, the constitutive model of the soil
and the contacts and the negligence of the grout finger in field. Nevertheless, based on
the numerical analysis on pullout stress analysis, reasonable input parameters can be
analysis of a tunnel project has been carried out with satisfaction (Cheng et al. 2013).
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
The numerical pullout analysis can be used to confirm the suitability of the soil
parameters/constitutive model which are required for complicated projects, while the
field pullout test can be used to confirm the nail capacity and the quality of the nail
installation.
6. Discussions
In the past, the conventional steel bar has been a good solution to stabilize the
sandy soil. There are however also many cases where there is difficulty in access and
transportation, or the soil is too soft and impermeable to grouting so that the use of
alternative soil nail is required. The purpose of the present study is to investigate the
materials. The geological materials include loose granular soil as well as clayey soil,
and it is proposed that further works will be carried out for fractured rock in the future.
To achieve the aims of the present work, both laboratory and field tests on steel and
FRP soil nails are carried out together with the corresponding in-situ soil
improvement determination. Vast amount of laboratory and field tests for FRP and
steel nails have been carried out in Hong Kong, Korea and Australia, and only small
amount of important data can be given in the present paper. Based on vast amount of
laboratory and field test results, some results can be drawn as follows:
(a) Generally, the GFRP and CFRP nails exhibit similar behaviors as ordinary steel
bars, except for the brittleness beyond peak load from the present study. The strength
18
Page 19 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
of FRP nails is less than the steel nails, but the modulus is significantly less. While
flexible nail may be beneficial for the mobilization of the shear strength for soil
stabilization in some cases, the acceptability of the soil movement must also be
considered. With regards to the control of soil movement, the authors suggest that a
slightly higher factor of safety may be applied to FRP nail as compared with the
(b) From the present work, it is clear that the use of expansive grout can help to
increase the bond stress and the elastic range, and this is particularly useful for loose
sandy soil. From the present works, the uses of Tube a’ Manchette for sandy soil in
Hong Kong and fracture grouting for clayey soil in Australia have been demonstrated
to provide great benefit in the overall stabilization, and the combined use of the
pressure grouting and FRP can provide an alternative soil nail installation method
which can overcome the limitations of classical steel nail and installation method.
(c) It is found from the present study that the development of the nail axial load and
skin friction is different between steel and FRP nail, and it appears that this result has
not been reported in previous works. Through the field tests and actual applications in
Hong Kong, Australia and Korea, it appears that this factor alone this is not critical
(d) Contrary to the support of using CFRP as soil nail material based on laboratory
tests by Toufigh et al. (2014), the use of CFRP nail (which is usually small in
diameter size and has to be used in group form) is not encouraged for normal
application. Besides the material is expensive in cost, there are several technical field
issues which have been noted from the present study. The authors adopt 4 CFRP bars
in the trial tests in order to achieve the required nail strength comparable to that of the
steel nail used in Hong Kong, as each CFRP bar in the market is small in diameter.
Based on field test, it is found that it is difficult to maintain even load sharing among
different CFRP bars if there is more than one CFRP bars for each nail, and it is not
19
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 20 of 38
easy to transfer the load from the bearing plate to the CFRP nails due to the surface
condition and the size of CFRP nails. The shear failure of one CFRP bar in one of the
pullout test also warns that eccentricity in load sharing among different CFRP bars
(e) The use of numerical modeling by FLAC3D or similar software can provide good
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
modelling of the field pullout tests, and this has also been reported in other previous
works. The load-displacement relation, axial load and friction distribution from
numerical modelling are similar to those obtained from the test if suitable parameters
are used for the analysis. Since numerical modeling is much cheaper than field test,
the use of numerical modeling of FRP nail in routine design may be sufficient for
normal problems, while trial tests on site can be limited to a few numbers as a control
(f) Although there are some previous long term monitoring of FRP nails, there is not
enough database for general soil condition. The long term behavior of FRP soil nails
under stress should be studied, which is part of the works that the authors are working
on. On one hand, creep may lower the capacity of FRP nail. On the other hand, the
soil arching effect may disappear when there is enough time for the system to stabilize
so that the actual vertical stress acting on the soil nail will become the effective
7. Conclusions
In conclusion, the laboratory and field tests have demonstrated the applicability
of using FRP as a competitive alternative to the steel soil nail for sites with difficult
access. FRP is much lighter than steel and is easier to manipulate on site with limited
working space and difficult access which is particularly important in many congested
cities in Asian, and less labors are required for FRP nail installation as compared with
grouting which can give higher bond stress has been demonstrated to be useful for
normal GFRP/steel soil nails in loose sand. The authors would recommend the use of
classical steel nail for normal application. On the other hand, CFRP which has a very
high tensile strength but low shear strength has demonstrated some practical problems
in both installation and pullout test. In views of the high cost for CFRP and the
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
difficulty to construct a good nail head for load transfer, CFRP is not recommended to
Acknowledgements
Nails for Slopes with Poor Access & Mechanism” (account YBBY) funded by the
Reference
Benmokrane, B., Tighiouart, B., and Chaallal, O. 1996. Bond strength and load
pp. 246-253.
Benmokrane, B., Zhang, B., and Chennouf, A. C. 2000. Tensile properties and pullout
Cheng, Y.M., and Wei, W.B. 2007. Application of Innovative GFRP Pipe Soil Nail
3006-3009.
Cheng, Y.M., Choi, Y.K., Yeung, T., Tham, L.G., Au, S.K., Wei, W.B., and Chen, J.
21
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 22 of 38
2009. New soil nail material-pilot study of grouted GFRP pipe nails in Korea and
Cheng, Y.M., Au, S.K., Pearson, A.M. and Li, N. 2013. An innovative Geonail
Cheung, W.M., and Lo, D.O.K. 2005. Use of carbon fibre reinforced polymer
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Hong Kong.
De Lorenzis, L., Miller, B., and Nanni, A. 2001. Bond of FRP laminates to concrete,
GEO 2008. Guide to Soil Nail Design and Construction. GEO, HK SAR Government,
Hong Kong.
GEO and HKIE 2011. Design of soil nails for upgrading loose fill slopes,
Kong.
Hyett, A.J., Moosavi, M., and Bawden, W.F. 1996. Load distribution along fully
Lazarte, C.A., Elias, V., Espinoza, R.D., and Sabatini, P.J. 2003. Soil Nail Walls,
Washington, DC
Li, G.W., Ni, C., Pei, H.F., Ge, W.M., and Ng, C.W.W. 2013. Stress relaxation of
22
Page 23 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
grouted entirely large diameter B-GFRP soil nail. China Ocean Engineering,
27(4), 495-508.
Smith, I.M., and Su, N. 1997. Three-dimensional FE analysis of a nailed soil wall
Geomechanics, 21(9):583-597.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Toufigh, V., Desai, C.S., Saadatmanesh, H., Toufigh, V., Ahmari, S. and Kabiri, E.
2014. Constitutive Modeling and Testing of Interface between Backfill Soil and
1-8.
Unwin, H. 2001. Carbon fibre soil nailing for railway embankments, Proceedings of
Wei, W.B., and Cheng, Y.M. 2010. Soil nailed slope by strength reduction and limit
Zhang, C.C., Zhu, H.H., Xu, Q., Shi, B., and Mei, G.X. 2015. Time-dependent pullout
670–681.
Zhang, C.C., Zhu, H.H., Shi, B, Wu, F.D., and Yin, J.H. 2015. Experimental
23
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 24 of 38
Figure Captions
Figure 1 Location of the earth pressure sensors at 100mm above the centre of the
holes (800mm above bottom), with 4 LDTV at the corners and 1 LDTV at the centre
Figure 2 Location of the strain gauges along the soil nails (units in mm)
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Figure 5 Destructive pullout test for CFRP nail A and five steel nails type C at
different vertical pressure (from 1 to 5). The grout pressures for the five groups of
Figure 6 Non-destructive pullout test for CFRP nail A, steel nail D and E
Figure 7 Axial force distributions along soil nail at vertical stress 50kPa in laboratory
Figure 8 Strain gauges locations on soil nails (some strain gauges are not functioning
Figure 10 Axial force and friction along soil nails from field test (with same grout
24
Page 25 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
Figure 12 Comparison of field test and simulation results (X3 and X4 are steel nail
Figure 13 Comparison of field test and simulation results for X6 (GFRP nail)
25
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 26 of 38
Position of Nail A
0.3 EPS 1 EPS 2 EPS 3
Position of Nail B
0.3
Position of Nail C EPS 7 1.6 m
0.3
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Position of Nail D
0.3 EPS 4 EPS 5 EPS 6
Position of Nail E
2.0 m
Figure 1 Location of the earth pressure sensors at 100mm above the centre of the
holes (800mm above bottom), with 4 LDTV at the corners and 1 LDTV at the centre
Figure 2 Location of the strain gauges along the soil nails (units in mm)
1
Page 27 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
43kPa 43kPa
(a) Pullout force – displacement curve for 25mm (b) Pullout force – displacement curve of
GFRP nail using ordinary cement grout 25mm GFRP nail using expansive grout
2
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 28 of 38
0 10 20 0 10 20 0 10 20 0 10 20 0 10 20
Figure 5 Destructive pullout test for CFRP nail A and five steel nails type C at
different vertical pressure (from 1 to 5). The grout pressures for the five groups of
Figure 6 Non-destructive pullout test for CFRP nail A, steel nail D and E
3
Page 29 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
Figure 7 Axial force distributions along soil nail at vertical stress 50kPa in laboratory
2.6 m
1.8 m
1.4 m
1.0 m
0.6 m
0.2 m
Figure 8 Strain gauges locations on soil nails (some strain gauges are not functioning
4
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 30 of 38
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
X1 X2 X3 X4 X8 X11
250
200
Force(kN)
150
100
50
0
0 2 4 6 8 10
Distance from upper nail end (m)
Figure 10 Axial force and friction along soil nails from field test (with same grout
5
Page 31 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
150
Force(kN)
100
50
-50
0 2 4 6 8 10
Dista nce from upper nail end (m)
Figure 12 Comparison of field test and simulation results (X3 and X4 are steel nail
6
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 32 of 38
EXP force
Axial Load Comparsion at Different Loads for X6
model force
300 EXP force
model force
250 EXP force
Axial Load(kN)
model force
200
150
100
50
Can. Geotech. J. Downloaded from www.nrcresearchpress.com by University of Otago on 10/19/15
0
0 1 2 3 4 5 6
Location(m)
Figure 13 Comparison of field test and simulation results for X6 (GFRP nail)
300
250
Skin friction (kN/m2)
200
150
100
50
0
0 1 2 3 4 5 6
Distance from nail head
7
Page 33 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
120
100
80
60
40
20
0
0 2 4 6 8 10 12 14
Distance from nail head
8
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record. Page 34 of 38
Bar diameter Ultimate stress Elastic modulus Ultimate Yield stress Yield
Bar type
(mm) (MPa) (GPa) strain (%) (MPa) strain (%)
D shape GFRP
10 425 30 1150
Bar**
* GFRP pipe, internal diameter=37mm, external diameter=47mm, FRP bonded with vinyl resin
*** Tensile test for CFRP and GFRP bar is in accordance with BS 2872:Part10 method
4449:1997/2005.
Page 35 of 38
For personal use only. This Just-IN manuscript is the accepted manuscript prior to copy editing and page composition. It may differ from the final official version of record.
Table 2 Elastic and permanent elongation of soil nails C (CFRP) and B (steel)
Table 3 Basic soil properties in field (shear strength parameters are measured by triaxial tests)
Note : Long term strengths of ordinary and expansive grout are similar
Table 6 Detailed descriptions of soil nails for field tests in Hong Kong
Adjusted
Total Inclination to Type of soil Max. load
No. grout Length Cement
Length (m) horizontal nail (kN)
(m)
Reinforcement Type
Element Parameter
Steel Carbon Fiber Glass Fiber
Young’s Modulus
205 100 36.4
Cable (GPa)
Cable tensile
460 1096 671
strength (MPa)