See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/229028395

Field monitoring evaluation of geotextile-reinforced soil-retaining walls

Article in Geosynthetics International · January 2007

DOI: 10.1680/gein.2007.14.2.100

CITATIONS READS

156 2,321

3 authors, including:

Carlos Vinicius Santos Benjamim Jorge G. Zornberg

University of São Paulo University of Texas at Austin
12 PUBLICATIONS 178 CITATIONS 293 PUBLICATIONS 6,064 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Jorge G. Zornberg on 28 April 2015.

The user has requested enhancement of the downloaded file.

 Geosynthetics International, 2007, 14, No. 2

Field monitoring evaluation of geotextile-reinforced

soil-retaining walls
C. V. S. Benjamim1 , B. S. Bueno2 and J. G. Zornberg3
1
Ober S.A. Indústria e Comércio, Rua Herval, No. 348, CEP: 03062-000, São Paulo/SP, Brazil,
Telephone: +55 11 6090 3328, Telefax: +55 11 6693 4701, E-mail: [email protected]
2
Geotechnical Department, University of São Paulo, Av. Trabalhador Sancarlense, No. 400, CEP:
13566-590, São Carlos/SP, Brazil, Telephone: +55 16 3373 9505, Telefax: +55 16 3373 9509,
E-mail: [email protected]
3
Civil Engineering Department, University of Texas at Austin, 1 University Station C1792, Austin,
TX 78712-0280, USA, Telephone: +1 512 471 6548, Telefax: +1 512 471 6548,
E-mail: [email protected]

Received 13 February 2006, revised 11 December 2006, accepted 16 January 2007

ABSTRACT: The use of geotextiles as reinforcement in soil-retaining walls may offer benefits over
the use of other types of reinforcement, such as metallic strips or geogrids. These include ease of
construction, expediency, and significant cost reduction. However, the lack of field monitoring
results documenting the behavior of geotextile-reinforced structures, mainly regarding deformations,
has precluded widespread utilization of these systems. A better understanding of the behavior of
geotextile-reinforced soil structures could be gained by systematic monitoring of full-scale
structures, combined with a thorough material characterization program. Accordingly, eight
prototype geotextile-reinforced soil structures were built as part of a comprehensive study aimed at
analyzing their behavior. This paper presents the characteristics of the prototype wall used as
baseline for this series of structures, including construction details, instrumentation layout, and the
results of laboratory tests conducted to characterize the soils and geosynthetics involved in this
structure. The lateral displacements at the face of the wall were small, with increasing lateral
displacements observed over time, particularly in association with precipitation events during the
summer season in Brazil. The largest horizontal strains occurred towards the face, at mid-height of
the prototype wall. The use of sigmoid curves to fit displacement results within the geotextiles was
found to be very useful to define the value and the location of the peak strain in each
reinforcement layer. The location of maximum reinforcement strains within each layer was found to
be consistent with the development of a potential failure surface starting at the toe of the wall and
propagating into the soil mass. The maximum reinforcement tension predicted using the K-stiffness
method was found to compare well with reinforcement tension values obtained from monitoring
results.

KEYWORDS: Geosynthetics, Reinforced soil walls. Geotextile, Cohesive soils, K-stiffness method

REFERENCE: Benjamim, C. V. S., Bueno, B. S. & Zornberg, J, G. (2007). Field monitoring

evaluation of geotextile-reinforced soil-retaining walls. Geosynthetics International, 14, No. 2,
100–118 [doi: 10.1680/gein.2007.14.2.100]

1. INTRODUCTION displacements, has precluded their utilization in critical

Unlike in many countries, retaining walls in Brazil have structures.
often been reinforced using geotextiles instead of other In Brazil, as in many other countries, reinforced soil
reinforcement inclusions such as geogrids and metallic structures have replaced traditional retaining walls, such
reinforcements. This approach is reported to have several as concrete gravity and cantilevered walls. Soil reinforce-
advantages, such as ease of construction, expediency, and ment involves the use of tensile elements such as metallic
significant cost reduction. However, the current lack of strips, geosynthetics or wire meshes between compacted
data on the field performance of geotextile-reinforced soil layers. Modern reinforced soil technology was devel-
structures in Brazil, mainly regarding their anticipated oped in the early 1970s, when French architect Henry
1072-6349 # 2007 Thomas Telford Ltd 100
Field monitoring evaluation of geotextile-reinforced soil-retaining walls 101

Vidal invented a technique he called Reinforced Earth. shown by reinforcements under the confinement of soil,
The original system uses metallic strips connected to and the contribution of soil arching to the stability of the
concrete panels, which constitute the facing of the re- structure. Field observation of the behavior of prototype
inforced system (Vidal 1966). walls still constitutes a clear need. However, the instrumen-
Geosynthetics were first used as soil reinforcement tation of prototype walls is expensive, and the collected
inclusions in France in the early 1960s, when Rhône- field data often address only limited aspects that govern the
Poulenc’s engineers used geotextiles in geotechnical appli- overall behavior of the structures.
cations such as unpaved roads, railroads and earth dams. Studies on the behavior of geotextile-reinforced struc-
Subsequently, in countries such as the Netherlands and tures have been conducted using small-scale physical
England, geosynthetics were used as reinforcement in modeling, such as centrifuge models (e.g. Zornberg and
retaining walls, providing an alternative to the original Arriaga 2003) and numerical modeling using the finite
Reinforced Earth approach. element method (e.g. Pedroso et al. 2006). Owing to the
In Brazil, geosynthetics have been used since the early difficulties in reproducing construction procedures when
1970s, mainly in drainage applications. The first signifi- using numerical and small-scale situations, instrumented
cant reinforced soil slope involving the use of geotextiles large-scale laboratory tests have also provided valuable
as reinforcement elements (Carvalho et al. 1986) was results (e.g. Bathurst et al. 1989). Although these studies
constructed in 1984 to rebuild a failed embankment along have provided significant insight into the behavior of
Highway SP-123 in Campos do Jordão (São Paulo state). geotextile-reinforced soil structures, additional data from
The use of geosynthetics as reinforcement elements has instrumented prototype walls would continue to provide
become widely accepted worldwide. The polymeric indus- invaluable information in understanding the effect of
try has continually introduced new manufacturing techni- variables governing the wall response.
ques as well as stronger and stiffer products, which has An important set of data on the behavior of prototype
made soil reinforcement using geosynthetics more attrac- structures was reported by Christopher (1993), who eval-
tive. The use of nonwoven geotextiles as reinforcement uated 12 retaining wall and steep slope prototypes for
inclusions can provide additional advantages by allowing structures with different facings (gabion, geosynthetic,
in-plane drainage and therefore allowing the use of fine- concrete panels) and reinforcement types (geotextile,
grained soil as backfill material. metallic strips). Farrag and Morvant (2004) also con-
Most design guidelines for geosynthetic-reinforced structed full-scale reinforced test walls with low-quality
structures recommend the use of granular soils as backfill backfill in order to investigate the interaction mechanisms
material. This recommendation is prompted by concerns between various geosynthetic materials and silty clay
associated with poor drainage of fine-grained soils, which backfill. Finally, Allen et al. (2002) presented a summary
leads to both loss of soil shear strength and post-construc- of 20 well-documented geosynthetic wall case histories,
tion movements (Zornberg and Mitchell 1994). However, covering a wide variety of wall heights, surcharge load-
compacted fine-grained soils from tropical areas often ings, foundation conditions, facing and reinforcement
show very good mechanical behavior, unlike fine-grained types, and inclusion spacings.
soils from temperate climates. For example, typical Brazi- In Brazil, experimental studies using geotextiles as soil
lian soils from colluvium or residual origins show a high reinforcement have been conducted by Ehrlich et al.
percentage of fines, but low plasticity, and an overall good (1997) and Ribeiro et al. (1999). However, no studies have
mechanical behavior as compacted backfill (Cruz 1996). been reported so far on large-scale prototype walls,
The possibility of post-construction (creep) deforma- particularly considering the use of typical Brazilian soils
tions, particularly when using polypropylene (PP) geotex- as backfill material.
tiles, has been an additional concern that may have With the objective of generating a database of instru-
prevented the widespread use of geotextiles as reinforce- mented structures, eight geotextile-reinforced soil-retain-
ment inclusions. The lack of creep data for PP geotextiles ing prototype walls were constructed by the Geosynthetics
has often led to the use of creep reduction factors as high Research Group at the University of São Paulo at São
as 5. The use of such high creep reduction factors, and the Carlos. These structures were instrumented to obtain the
comparatively lower tensile strength of nonwoven geotex- internal distribution of reinforcement strains as well as the
tiles, has relegated the use of nonwoven geotextiles to overall vertical and horizontal movements. This paper
comparatively small structures subjected to conventional presents the characteristics and behavior of one of the
surcharges. structures, Prototype Wall 1, which is considered a base-
The difficulty in predicting the behavior of geotextile- line case for future comparative evaluations. This paper
reinforced soil structures using current available design summarizes the reinforcement layout, the laboratory pro-
methods has been reported extensively in the technical gram conducted to characterize the soil and reinforce-
literature. This was illustrated by Wu (1992), who re- ments, the instrumentation monitoring program, and the
quested predictions on the behavior of a highly instrumen- construction methodology.
ted reinforced soil structure. The predicted results showed a The instrumentation results include the vertical and
significant scatter when compared with the monitored horizontal movements of the wall, as well as the reinforce-
response of the prototype wall. Many factors have contrib- ment internal displacements measured both during wall
uted to such discrepancies, including the contribution of construction and up to 203 days after construction. Inter-
suction to the soil shear strength, the increased stiffness pretation of the reinforcement displacements of prototype
Geosynthetics International, 2007, 14, No. 2
102 Benjamim et al.

wall 1 allows determination of the reinforcement loads, of soil, but using different reinforcement types. Two
which is used to assess the behavior of the wall using prototype walls were constructed using the same soil and
design methods that account for the structure deformations geotextile type, but with different reinforcement layout.
(e.g. the K-stiffness method). Finally, and in order to exacerbate post-construction move-
ments, one of the prototype walls was designed using a
reduction factor (combined reduction factor due to instal-
2. OVERVIEW OF THE FIELD lation damage, creep and durability) of 1. The character-
MONITORING PROGRAM istics of the eight prototype walls built as part of that
study are summarized in Table 1.
The instrumented wall prototype investigated in this study The prototype walls were built using wrap-around
is the baseline structure of a series of eight prototype facing, with a face overlap length of 1.0 m. No seams
walls constructed by the Geosynthetics Research Group at were used during placement of the reinforcements. Place-
the University of São Paulo at São Carlos. While the focus ment of the reinforcement at the face involved the use of
of this paper is on the behavior of the baseline wall metallic supports and wood lagging boards (Figure 2),
(prototype wall 1), a brief description of the overall field which allowed a face inclination of 788 with the horizontal
monitoring program is provided in this section. (batter of 1H:5V). Some of the walls involved the use of
Figure 1 shows the general configuration of the proto- granular backfill, which was compacted using a vibratory
type walls. In order to maximize the information collected
in this field monitoring program, the prototype walls were
built in pairs (twin walls constructed back to back). The
eight geotextile-reinforced retaining walls were designed,
built and instrumented in order to investigate the effect of
soil type, geotextile type and reinforcement layout. Three
prototype wall pairs were constructed using the same type

Shaft
Wall 2 Wall 1

4.0 m 1.0 m
0.4 m
1.5 m
1.5 m
3.0 m 4.0 m 3.0 m
Figure 2. Photograph of wood lagging board supports used
Figure 1. Configuration of prototype walls during construction

Table 1. Characteristics of prototype walls

Wall Geotextile Reinforcement layouta Soil type

Product name Type Polymer

1 Ober G200 Nonwoven PET Uniform spacing SP

Sv ¼ 0.40 m
2 Ober G200 Nonwoven PET Non-uniform spacing SP
Sv ¼ 0.30 m to 0.50 m
3 Ober G250 Nonwoven PP Non-uniform spacing SM
Sv ¼ 0.30 m to 0.50 m
4 Propex 10x50 Woven PP Non-uniform spacing SM
Sv ¼ 0.30 m to 0.50 m
5 Ober G250 Nonwoven PP Uniform spacing CL
Sv ¼ 0.40 m
6 Propex 10x50 Woven PP Uniform spacing CL
Sv ¼ 0.40 m
7 Ober G150 Nonwoven PP Uniform spacing SP
Sv ¼ 0.36 m
8 Propex 10x50 Woven PP Uniform spacing SP
Sv ¼ 0.40 m

a
Reinforcement length is 3 m for all prototype walls.

Geosynthetics International, 2007, 14, No. 2

Field monitoring evaluation of geotextile-reinforced soil-retaining walls 103
100
90
friction on small-scale walls did not significantly affect
CL soil
Percentage passing (%)

80 (Nova Odessa) the behavior of their models. Similar conclusions were

70 obtained by Zornberg and Arriaga (2003), who tested
60 small-scale reinforced soil structures to failure in a
50 SM soil
(Hortolândia) geotechnical centrifuge. The centrifuge models were lined
40
30 using Mylar sheets, which proved successful in minimiz-
20 SP soil ing side friction, as observed from careful post-failure
10 (São Pedro) assessment of the models.
0 Additional evidence that the walls behaved approxi-
0.001 0.01 0.1 1 10
Grain size (mm) mately in plane-strain conditions was provided by the
ASTM D 2487 - 00
Sand
slope of the surface crack that developed in the under-
Clay Silt Gravel
Fine Medium Coarse designed wall constructed as part of this study. As shown
in Figure 4, a wide transverse crack developed parallel to
Figure 3. Particle size distributions for soils used in proto- the prototype wall face (i.e. transverse to the wood side
type walls. walls). While a curved crack pattern would have provided
evidence of side friction, the shape of the crack indicates
plate, while the other walls used fine-grained backfill, that the assumption of plane-strain conditions is reason-
which was compacted using a mechanical tamper. able.
Three different soils from tropical origin, two fine- Both woven and nonwoven geotextiles were used as
grained and one granular soil, were used as backfill reinforcement inclusions in the prototype walls. Also, both
material. The soils were tested to obtain the data needed polypropylene (PP) and polyester (PET) products were
for the design of the prototype walls and the analyses of used. This allowed comparison of wall performance for
their behavior. Figure 3 shows the particle size distribution various geotextile and polymer types. Accordingly, the
of the soils used in the prototype walls. According to the study allowed investigation of the effect of the ultimate
Unified Soil Classification System (ASTM D 2487), the tensile strength, stiffness, and creep properties on the
soils classify as a low-plasticity clay (CL), a sandy silt behavior of geotextile-reinforced walls. While comparison
(SM), and a poorly graded sand (SP). According to the of the behavior of the eight walls is beyond the scope of
MCT (Miniature, Compacted, Tropical) classification for this paper, the behavior of the baseline wall is evaluated.
fine-grained soils (Nogami and Villibor 1981), the CL soil Specifically, this paper provides the analysis of prototype
is a lateritic clayey sand, and the SM is a non-lateritic wall 1, which used SP backfill and was reinforced with
sandy soil. PET nonwoven geotextiles.
The walls were confined laterally using longitudinal
wood side walls. The walls were 4.0 m wide and 4.0 m
high. The paired reinforced walls had a total length of
3. DESCRIPTION OF THE BASELINE
10.0 m, so the total length of each wall (including a non-
PROTOTYPE WALL
reinforced zone) is 5.0 m. Geomembranes lined with
grease were used along the wood side walls in order to 3.1. Wall characteristics
minimize lateral friction between the soil and the wood Figure 5 shows a front view of prototype wall 1, which
and promote plane-strain test conditions. Direct shear tests was constructed in July 2002, using a nonwoven geotextile
were conducted to quantify the interface shear strength as reinforcement and a medium to fine sand as backfill
between the soil and the greased geomembranes. Table 2 soil. The prototype wall was built using 10 geotextile
summarizes the side wall/soil interface friction angles layers placed with a vertical spacing of 0.40 m.
measured for the soils used in this study. The test results
indicated that the interface shear strength is very small,
with a friction angle of approximately 5.08 for the fine-
grained soils and 7.68 for the medium to fine sand. The Wall facing
measured interface friction angles are smaller than those
reported by Arthur and Roscoe (1965) for dry sand in Wall side
contact with glass (
¼ 148). Their study showed that side

Table 2. Shear strength of interface between soil and

greased membranes located against test facility woodside
walls

Type of soil Interface friction angle (degrees) Wall side

Nova Odessa soil (CL) 5.2

Hortolândia soil (SM) 4.9
São Pedro soil (SP) 7.6 Figure 4. Photograph of transverse crack developed in
prototype wall 7
Geosynthetics International, 2007, 14, No. 2
104 Benjamim et al.

50

1000 h
100 h
10 h
40 1h
0.1 h
0.01 h

% of ultimate load
30

20

10

0
0 10 20 30 40 50 60
Strain (%)
Figure 5. Front view of prototype wall 1
Figure 6. Geotextile isochronous curves

3.2. Geotextile properties in the creep testing program were 10%, 20%, 30 and 50%
A short-fiber needle-punched polyester nonwoven geotex- of the ultimate geotextile tensile strength. The measure-
tile was used as reinforcement in prototype wall 1. The ments were registered using digital photography techni-
geotextile is characterized as follows: ques (Baras et al. 2002). Figure 6 shows the isochronous
curves obtained from the creep tests for the nonwoven
• mass per unit area: 204.4 g/m2 PET geotextile used in this wall.
• thickness: 1.26 mm Tensile tests (ASTM D 4595) were conducted using
• ultimate tensile strength: 8.41 kN/m three types of geotextile specimen: virgin samples, speci-
• ultimate elongation: 89.8% mens retrieved after field placement (to characterize
• offset tensile modulus (ASTM D 4595): 13.0 kN/m. construction damage), and specimens subjected to UV
exposure (approximately 200 days). Table 3 shows the
Creep tests (ISO 13341) were conducted using this results, which indicate that the retrieved specimens gave a
geotextile in order to evaluate the post-construction move- slight increase (6.5%) in tensile strength in the machine
ments observed in the prototype wall. The load levels used direction (MD) and a slight decrease (2.1%) in the cross-

Table 3. Tensile test results of nonwoven geotextiles used in prototype wall 1

Machine direction

Sample type Tensile strength Strain at failure (%) Coefficient of variation

(kN/m) COV (%)

Strength Strain

Virgin sample 4.62 94.71 14.34 9.83

Retrieved from backfill 4.92 56.16 12.70 10.67
UV exposure 2.59 33.78 12.29 14.05

Cross machine direction

Sample type Tensile strength Strain at failure (%) Coefficient of variation

(kN/m) COV (%)

Strength Strain

Virgin sample 8.12 95.85 10.42 3.57

Retrieved from backfill 7.95 54.74 8.26 15.64
UV exposure 5.10 31.68 20.89 10.34

Geosynthetics International, 2007, 14, No. 2

Field monitoring evaluation of geotextile-reinforced soil-retaining walls 105

machine direction (CMD), compared with the virgin prototype wall was constructed over a 0.40 m base layer of
sample. The results also indicate that the retrieved speci- dense sand, which was reinforced using additional non-
mens show a decreased strain at failure of 40.7% (MD) woven geotextiles.
and 66.9% (CMD), which corresponds to increased re-
inforcement stiffness of 79.6% and 90.0%, respectively. 3.4. Design overview
The exposed specimens showed a significant decrease in Prototype wall 1 was designed using a conventional
tensile strength (43.9% for MD and 37.2% for CMD), approach, consistent with that described by Mitchell and
although stiffness of the geotextiles increased after being Villet (1987) and Elias et al. (2001). However, based on
exposed 200 days to UV (57.2% for MD and 90.0% for the test results obtained as part of this study, the reduction
CMD), compared with the virgin sample. factors against creep, durability and installation damage
were taken as 1. Also, based on the results of the material
3.3. Backfill properties program, a soil friction angle of 338 was adopted while
The backfill material used in prototype wall 1 was the SP the cohesive intercept was neglected. The design assump-
soil collected from the São Pedro region (Figure 3). The tions of prototype wall 1 resulted in uniform reinforce-
maximum and minimum void ratios, emax and emin , ment spacing, Sv , of 0.4 m and a reinforcement length
according to ASTM D 4253 and ASTM D 4254, were equal to 3.0 m. The adopted facing inclination is 788,
0.70 and 0.46. The backfill soil was compacted to a target which corresponds to a face slope of 1H:5V.
relative density of 80%, which corresponds to a void ratio
of 0.51 and a dry unit weight of 1.77 kN/m3 . The backfill 3.5. Instrumentation program
was placed in the field at a moisture content of approxi- Three aspects of the prototype wall behavior were mon-
mately 5.0%. itored by field instrumentation both during and after wall
The shear strength of the soil was obtained from construction. Specifically, the monitoring program in-
consolidated-drained (CD) triaxial tests on specimens cluded:
prepared at the unit weight and water content used during
construction of the prototype wall. The shear strength is • horizontal displacements within the reinforcement
characterized by a cohesive intercept of 16 kPa and a layers measured using mechanical extensometers
friction angle of 338. (telltales);
The foundation soil where the prototype wall was • face displacements measured by external surveying;
constructed was characterized by in situ SPT (standard and
penetration test) results. Figure 7 shows the SPT profile, • vertical settlements measured by magnetic extens-
which indicates a profile with low blow counts. To ometers.
minimize settlement and external stability problems the
The instrumentation layout of prototype wall 1 is shown in
Figure 8.
Blow count – N
The mechanical extensometers consist of 0.35 mm
N 0 10 20 30 40
diameter stainless steel wires, running inside nylon tubes,
1 2
which were used to reduce friction and to protect the
2 2
wires. The wires were connected to the geotextiles using
3 2 staples. Figure 1 shows the vertical shaft, constructed at
4 4 the back of the wall, which was used to collect readings
5 4 from the mechanical extensometers. One end of the
6 6 telltale is fixed to the geotextile, and the free end is
7 6
located in the vertical shaft. The free end of the telltale is
attached to a hanging weight, which allows measurement
8 6
of the displacements (Figure 9). The free ends of the
Depth (m)

9 9
telltales were located in a shaft constructed behind the
10 10
prototype wall. Measurements were made using a digital
11 9 caliper with a resolution of 0.01 mm.
12 8 Displacements at the wall face of the prototype were
13 9 also measured during and after wall construction by the
14 12 mechanical extensometer attached to the face of each
15 12
reinforcement layer. In addition, external displacements of
16
the prototype wall face were surveyed during and after
11
construction by measuring the distance of points placed at
17 14
the center of each layer to a fixed reference point. Figure
18 29 10 shows the survey measurement of prototype wall 1 and
19 34 external reference post. Survey measurements allowed a
resolution of 1 mm. A total of five reinforcement layers
Figure 7. Results of standard penetration test in foundation were instrumented with seven mechanical extensometers
soil each, as shown in Figure 8.
Geosynthetics International, 2007, 14, No. 2
106 Benjamim et al.
Vertical extensometers

Reference post
Extensometer points

Vertical
shaft 7 6 5 4 3 2 1
Telltales Survey points
elevation 5
0.8 m
0.8 m
Telltales
elevation 4

4m
Telltales
elevation 3

0.5 m
Telltales
elevation 2
Reinforcement
0.4 m
Telltales
elevation 1
Magnets
Weight Foundation layer

80 cm

1.0 m 1.0 m 1.0 m

Figure 8. Instrumentation plan of prototype wall 1

(a) (b)

Figure 9. View of telltale free end: (a) measuring displacement; (b) telltale end

Magnetic extensometers were designed and constructed faces of the magnet. The resolution of the magnetic
as part of this study to measure vertical settlements within extensometer is 1 mm.
the backfill (Figure 11). The extensometers are composed
of magnetic plates attached to PVC pipes. Magnets were
attached to the center of each 0.20 m square plate to
4. INSTRUMENTATION RESULTS
register the position of a vertical probe. Measurements Preliminary instrumentation results obtained from proto-
were conducted by introducing an aluminum probe with type wall 1 illustrate the baseline response of the struc-
an electronic measuring device (reed switch) installed in tures constructed as part of this investigation. In addition,
its tip. When the probe approaches the magnet, an electric results of this prototype wall will be useful in conducting
circuit closes, activating a sound indicator. The measure- parametric analyses. These comparisons are beyond the
ments were taken at the beginning and ending of the scope of this paper, but include evaluation of different soil
sound indicator, which correspond to the top and bottom types (prototype walls 3 and 5), reinforcement types
Geosynthetics International, 2007, 14, No. 2
Field monitoring evaluation of geotextile-reinforced soil-retaining walls 107

initial reading reported in each figure (at time 1 hour)

represents the displacement at the end of construction.
The overall displacements reached a maximum value of
13.0 mm at mid-height of the prototype wall. The instru-
mented layers presented intermediate maximum displace-
ments of 6.0 and 8.0 mm at intermediate elevations 2
(1.2 m) and 4 (2.8 m), respectively. The top and lower
reinforcement layers showed the smallest maximum dis-
placements (2.5 mm and 5.0 mm at elevations 1 (0.4 m)
and 5 (3.6 m), respectively).
The largest horizontal displacements occurred in the
vicinity of the face of the prototype wall (see distance
from facing 0.0 m in Figures 12b to 12f), and increased
with time. Most of these displacements were observed to
take place after the rains that occur during the summer
season in Brazil. For example, during a period of rain of
1 mm/day, the displacements in layer 3 increased from 5
to 9 mm (i.e. at a rate of 0.7 mm/month). However, during
a period of rain of approximately 10 mm/day, the displace-
ment rate increased from 9 to 13 mm (i.e. at a rate of
1.0 mm/month).
In order to facilitate data interpretation, the extens-
(a) ometers were aligned in profiles parallel to the face during
installation to facilitate evaluation of the horizontal dis-
placements. Figure 13 shows the displacement distribution
patterns, using the same data as presented in Figure 12,
but illustrating the outward movements. The displacements
shown in Figure 13 were measured in relation to a fixed
location at the back of the prototype wall. Accordingly,
they represent the displacements measured directly using
the telltales. As expected, the largest displacements oc-
curred close to the wall face, reaching 13 mm at mid-
height of the prototype wall. The second profile of
extensometer points, located 0.5 m from the wall face,
shows a displacement pattern similar to the first profile
(located at the facing), with the maximum displacements
occurring at the same elevation (2.0 m).
In the two profiles closer to the prototype wall face, the
displacements decrease towards the base and to the top of
the structure, reaching negligible values at both elevations.
The subsequent profiles (from 1.0 m from the face to the
back) no longer show a pattern of maximum displacement
at mid-height of the prototype wall. Specifically, the
displacement results show significantly smaller values,
which did not exceed 2 mm at any location.

(b) 4.2. Face displacements

The face displacements measured by external survey were
Figure 10. Measurement of face displacements: (a) survey
measurement; (b) external reference post
larger than those measured using extensometers at the
facing (Figure 14). These discrepancies can be attributed
to localized displacements due to soil bulging at the
(prototype walls 7 and 8), and the internal geometry of the flexible facing, which occurred during construction. It
structures (prototype wall 2). should be noted that survey measurements were made in
the middle of each layer, whereas extensometer measure-
4.1. Horizontal displacements within the ments were made at the reinforcement layer.
reinforcements However, both external (survey) and internal (extens-
Figures 12b to 12f show the time history of the horizontal ometer) measurements showed similar patterns, with
displacements measured by extensometers along five of maximum displacement at mid-height of the wall, and
the reinforcement layers. The precipitation recorded at the decreasing values towards the base and the top of the
site during this period is also shown, in Figure 12a. The structure. The maximum displacement measured using
Geosynthetics International, 2007, 14, No. 2
108 Benjamim et al.

Aluminum probe
with ruler

PVC pipe
(φ ⫽ 1 in)

4m (b)
PVC pipe
(φ ⫽ 114 in)

Magnet

⬇ 80 cm

Reference
magnet

Concrete block

1.0 m

(a) (c)

Figure 11. Vertical displacement measurement: (a) magnetic extensometer equipment; (b) aluminum probe; (c) PVC pipe with
magnet

extensometers at the end of the construction was 5 mm, strains that occurred during construction were compara-
whereas the maximum displacement measured by external tively small, with the largest values occurring towards the
survey was 58 mm. After 203 days, the maximum dis- face of the structure. The largest strain values occurred at
placement increased to 13 mm (extensometer readings) approximately mid-height of the prototype wall, with a
and 65 mm (survey measurements), respectively. That is, maximum value calculated at the end of construction of
both survey and extensometer post-construction displace- 0.8%. Additional post-construction strains occurred during
ments showed a consistent maximum value of approxi- the weeks following construction, reaching a maximum
mately 7 mm. value of 1.3% after 203 days. The strains induced near the
Figure 15a shows the face displacements measured by face took place mainly during construction, with local
extensometers at different times after construction, and bulging observed to develop after removing the wood
Figure 15b illustrates the post-construction displacements lagging board supports.
at the face. These results show that the location of the Figure 17 shows the development of the maximum
largest displacements observed during construction (mid- strains in the instrumented reinforcement, with precipita-
height of the wall) is the same as the location of the tion data shown as a reference in the same graph. The
largest post-construction displacements. These results are precipitation data show that the prototype walls were
consistent with those obtained by Zornberg et al. (1998) subjected to heavy rain, which led to water infiltration into
using centrifuge models, who reported that the maximum the reinforced backfill. A decrease in soil suction and an
strains for steep slopes occur towards the mid-height of increase in the weight of the soil are responsible for the
the structure. slight post-construction increases in reinforcement strains.
Comparison of field post-construction strains with creep
laboratory test results indicates that the strain rates (slopes
5. EVALUATION OF REINFORCEMENT of the strain curves) registered in the prototype wall were
STRAINS AND LOADS considerably smaller than the rate obtained from labora-
tory tests. The overall strains were small, which is notable,
5.1. Preliminary evaluation of reinforcement strains particularly considering that the prototype wall was de-
The reinforcement strains were initially calculated by signed using reduction factors of 1.0. The results suggest
dividing the relative displacement between two consecu- that the method used in the prototype wall design is
tive extensometer points by the initial distance between conservative.
the extensometer points. Figure 16 shows the reinforce- However, it should be noted that other factors that may
ment strains. Although the wall was designed using have contributed to the comparatively small strains were
reduction factors of unity, the horizontal reinforcement not taken into consideration in the design. These factors
Geosynthetics International, 2007, 14, No. 2
Field monitoring evaluation of geotextile-reinforced soil-retaining walls 109

Monthly precipitation (mm)

400

300

200

100

0
1 10 100 1000 10 000
Time (hours)
(a)

Distance from facing

0.0 m 0.5 m 1.0 m 1.5 m 2.0 m 2.5m 3.0m

14
Displacement (mm)

12 Elevation 5 (3.6 m)
10
8
6
4
2
0
1 10 100 1000 10 000
(b)

14
Elevation 4 (2.8 m)
Displacement (mm)

12
10
8
6
4
2
0
1 10 100 1000 10 000
(c)

14
Displacement (mm)

12 Elevation 3 (2.0 m)
10
8
6
4
2
0
1 10 100 1000 10 000
(d)

14
Displacement (mm)

12 Elevation 2 (1.2 m)
10
8
6
4
2
0
1 10 100 1000 10 000
(e)

14
Displacement (mm)

12 Elevation 1 (0.4 m)
10
8
6
4
2
0
1 10 100 1000 10 000
(f)

Figure 12. Time record of precipitation and reinforcement displacements measured by extensometers: (a) precipitation; (b)
elevation 1; (c) elevation 2; (d) elevation 3; (e) elevation 4; (f) elevation 5
Geosynthetics International, 2007, 14, No. 2
110 Benjamim et al.
2 mm
4.0
Elevation 5 Initial
3.6 position
End of
3.2 construction
Elevation 4 7 days
2.8

Elevation (m)
2.4 15 days
Elevation 3
2.0
27 days
1.6 Face

Elevation 2 57 days
1.2
84 days
0.8
Shaft
Elevation 1
0.4 203 days

0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)

Figure 13. Data of end bearing extensometers aligned in quasi-vertical profiles

End of construction (external survey) measuring points. However, the use of raw extensometer
After 200 days (external survey) displacement data to perform these calculations often led
End of construction (extensometer)
After 200 days (extensometer) to erratic reinforcement strain distributions. This is be-
4.0 cause minor scatter in the displacement trend results in
major oscillations in the calculated strains. Consequently,
3.5 the raw extensometer displacement information was also
evaluated by fitting the data to a monotonically increasing
3.0 curve in order better to define the strain distribution
(Zornberg and Arriaga 2003).
2.5 This approach allows direct determination of the magni-
Elevation (m)

tude and location of the maximum reinforcement strain,

2.0 even in the case of very small displacements. The expres-
sion used to fit the extensometer displacements is a
1.5 sigmoid curve defined by
1
1.0 d¼ (1)
a þ be
cx

0.5 where d is the displacement of each point relative to the

face of the prototype, x is the extensometer point, e is the
0 base of natural logarithms, and a, b and c are the fitting
0 10 20 30 40 50 60 70 parameters. The main characteristics of the sigmoid func-
Displacement (mm)
tion are shown in Figure 18.
Figure 14. Face displacements measured by external survey
After fitting the displacement data to sigmoid curves,
and extensometer readings the geotextile strain distribution could be defined as the
derivative of the sigmoid function. The peak strain value
and its location from the slope face in each reinforcement
layer can be determined analytically using the parameters
include the contribution of suction in the unsaturated a, b and c that fit the displacement data. The expressions
backfill soil, the effect of soil confinement on the stiffness for maximum strain and its location are
of the nonwoven geotextile reinforcement, and the con- c
tribution of geotextile overlaps, which effectively doubled max ¼ (2)
4a
the number of geotextile layers in the vicinity of the wall

1 b
face. xmax ¼ ln (3)
c a
5.2. Use of sigmoid curves for determination of where max is the magnitude of the peak strain in each
displacement trends reinforcement and xmax is the location, measured from the
As mentioned earlier, geotextile strain values were initially slope face, where the peak strain occurs.
estimated by calculating the relative movements between Figure 19 shows the horizontal geotextile displace-
extensometers and dividing them by the distance between ments, relative to the wall face, obtained from the
Geosynthetics International, 2007, 14, No. 2
Field monitoring evaluation of geotextile-reinforced soil-retaining walls 111
4.0 4.0
End of
construction
3.5 3.5

7 days
3.0 3.0

15 days
2.5 2.5
Elevation (m)

Elevation (m)
27 days
2.0 2.0

1.5 1.5 57 days

1.0 1.0 84 days

0.5 0.5
203 days

0 0
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
Face displacement (mm) Face displacement (mm)
(a) (b)

Figure 15. Face displacements measured by extensometers: (a) Total displacements, b) Post-construction displacements

extensometers, along with the superimposed sigmoid The proximity of the potential failure surface to the
curves defined by fitting the raw displacement data. The prototype wall face can be attributed to the observed
solid points in the figure represent the displacement sagging of the facing. Accordingly, a surficial stability
registered directly with the extensometers, and the lines problem, rather than a deep shear surface, could govern
correspond to the sigmoid curves. Figure 20 shows the the performance. Yet, since this is the region of the
reinforcement strains calculated using the sigmoid fitted additional facing overlaps, there are twice as many
curve approach. Although adding significant fluctuations geotextile layers in this zone, which provide additional
to the strain distributions, preliminary calculations done stabilization at the surface.
using the raw relative movements between extensometer
points provide similar trends to those obtained by fitting 6.2. Calculation of Tmax using the K-stiffness method
the raw data to a sigmoid curve. The strain distribution The K-stiffness method (Allen and Bathurst 2003;
shows small values in the vicinity of the prototype wall Bathurst et al. 2005) was used to estimate the tensile
face, a clear strain peak, and a negligible strain towards forces in the reinforcements using the measured strain
the end of the reinforcement layer. Although the maximum values. This methodology considers the stiffness of all
strain values are very similar to those calculated directly, wall components relative to the soil stiffness to estimate
the locus of maximum strain is better defined when using the distribution and magnitude of the maximum tensile
sigmoid curves in the analysis. load, Tmax, in each reinforcement layer. The method is
semi-empirical in nature, as it was calibrated using
monitored case histories to predict the reinforcement
6. ADDITIONAL ASPECTS OF THE loads. Consequently, the results obtained in this study
MONITORED WALL PERFORMANCE provide additional information suitable for validation and/
or refinement of the K-stiffness method, particularly con-
6.1. Potential failure surface sidering that little information is available on the use of
The location of maximum tensile strain in reinforcements geotextiles as reinforcement.
has been reported to be the location of the potential failure The K-stiffness method was reported to provide a
surface (Zornberg et al. 1997). Figure 21 shows the tensile reasonably accurate prediction of wall performance under
strain distribution of prototype wall 1, with an indication working stress conditions (Allen and Bathurst 2003;
of the potential failure surfaces, as defined by the strain Bathurst et al. 2005). This method was reported to provide
calculated directly and using the sigmoid curves. The significantly more accurate estimates of reinforcement
locations of maximum reinforcement strain show evidence loads and strains in reinforced soil walls than previous
that the potential slip surface starts at the toe of the slope design methodologies.
and propagates into the soil mass, following a pattern The peak load in each reinforcement layer, Tmax, can be
consistent with that typically assumed in design. calculated using the K-stiffness method as follows:
Geosynthetics International, 2007, 14, No. 2
112 Benjamim et al.
1.6
Elevation 5 (3.6 m) 7 days
1.2 15 days

Strain (%)
27 days
0.8 57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(a)
1.6
Elevation 4 (2.8 m) End of construction
1.2 7 days
Strain (%)

15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(b)
1.6
End of construction
Elevation 3 (2.0 m)
1.2 7 days
Strain (%)

15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(c)
1.6
Elevation 2 (1.2 m) End of construction
1.2 7 days
Strain (%)

15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(d)
1.6
End of constrction
Elevation 1 (0.4 m)
1.2 7 days
Strain (%)

15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(e)

Figure 16. Reinforcement strains: (a) elevation 1; (b) elevation 2; (c) elevation 3; (d) elevation 4; (e) elevation 5

1.6 320

Elevation 1 (0.4 m)
Maximum reinforecement strain (%)

1.2 240 Elevation 2 (1.2 m)

Monthly precipitation (mm)

Elevation 3 (2.0 m)
0.8 160

Elevation 4 (2.8 m)

0.4 80 Elevation 5 (3.6 m)

Precipitation
0 0
1 10 100 1000 10 000
Time (h)

Figure 17. Maximum strains in each instrumented reinforcement

Geosynthetics International, 2007, 14, No. 2
Field monitoring evaluation of geotextile-reinforced soil-retaining walls 113

d ⫽ 1/a reinforcement density at a given layer, which is calculated

as
J
Slocal ¼ (7)
Displacement, d

Sv

a
Slocal
Ölocal ¼ (8)
Maximum strain ⫽ c/4a Sglobal
For prototype wall 1, with a vertical spacing Sv ¼
1
0.4 m, Slocal ¼ 32.50 kN/m2 and Ölocal ¼ 1.0. The wall
face batter factor, Öfb , which accounts for the influence of
d⫽0 the reduced soil weight on reinforcement loads, is deter-
mined from
Distance from slope face, x


K abh d
Figure 18. Sigmoid function used to fit extensometer dis- Öfb ¼ (9)
K avh
placement data

where Kabh is the horizontal component of the active earth

Tmax ¼ pressure coefficient accounting for wall face batter, Kavh is

0:24 the horizontal component of the active earth pressure
Sglobal
0:5Sv Kªð H þ S Þ Dtmax Ölocal Öfb Öfs 0:27 coefficient, d is a constant coefficient.
Pa Figure 23 presents the profile of Tmax values defined
(4) using the K-stiffness method. The same figure shows the
Tmax values derived by using the maximum strain values
where Sv is the tributary area, assumed equivalent to the defined using the sigmoid-fitted curves method. Based on
average vertical spacing of the reinforcement at each layer confined tests results published by Kamiji (2006) and
location;  is the friction angle of the soil; K is the lateral Gomes (1992), the confined stiffness of the tested geotex-
pressure coefficient for the reinforced backfill, determined tile was estimated as 65 kN/m, which is five times larger
approximately as K ¼ 1
sin ; H is the vertical wall than the unconfined stiffness.
height at the face; S is the average soil surcharge height Figure 23 shows that the Tmax values defined using the
above the prototype wall top; and Dtmax is a distribution measured strain results from prototype wall 1 show a
factor. Dtmax is used to estimate Tmax for each layer as a pattern consistent with those predicted using the K-
function of its depth below the prototype wall relative to stiffness method, although the actual values are smaller.
the maximum tension value within the prototype wall, However, it should be noted that suction in the unsaturated
Tmxmx (Figure 22). Öfs is the facing stiffness factor (equal backfill was not accounted for in the evaluation. Conse-
to 1.0 for wrapped faced walls), and Pa is the atmospheric quently, for practical purposes, the K-stiffness method is
pressure (101 kPa). Sglobal is the global reinforcement deemed to provide a good estimate of the values obtained
stiffness, defined as using measured results.
X
n
6.3. Vertical settlements
Ji
J ave i¼1 The vertical displacements within the backfill were
Sglobal ¼ ¼ (5) obtained by magnetic extensometers installed within the
ð H=nÞ H
reinforced zone. Figure 24 shows the time history of the
vertical displacements measured within the backfill using
where Jave is the average stiffness of all reinforcement the magnetic extensometers. The maximum settlement
layers within the entire wall section, Ji is the stiffness of measured at the end of the construction equals 5 mm, and
an individual reinforcement layer, H is the total wall is located close to the face of the prototype wall. After
height, and n is the number of reinforcement layers within 203 days the post-construction settlements increased by
the entire wall section. only 1 mm, reaching 6 mm at the magnets located at the
Considering that the same geotextile was used in all mid-height of the wall (magnets 3 and 4 in Figure 24).
the layers ( J ¼ 13.00 kN/m), the calculated Sglobal is The foundation settlements were also registered using
32.50 kN/m2 . Ölocal is the global reinforcement stiffness, vertical magnetic extensometers placed within the founda-
defined as tion soil. Figure 25 shows the settlements at the founda-


Slocal a tion level from the beginning of construction and up to 84
Ölocal ¼ (6)
Sglobal days after construction. The results indicate that the largest
displacements occurred in the middle of the reinforced
where a is a coefficient that is also a function of stiffness. zone.
The coefficient a was adopted as 1.0 for geosynthetic Large displacements of the foundation were expected
walls, as suggested by Allen and Bathurst (2003) and because of the compressible foundation soil, as quantified
Bathurst et al. (2005). by the low SPT test results (N ¼ 2). The largest displace-
The local stiffness (Slocal ) considers the stiffness and ments occurred during construction (approximately 91%
Geosynthetics International, 2007, 14, No. 2
114 Benjamim et al.

Relative displacement (mm)

14
Elevation 5 (3.6 m) End of construction
12
7 days
10
15 days
8
27 days
6
57 days
4 84 days
2 203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
Relative displacement (mm) (a)
14
End of construction
12 Elevation 4 (2.8 m)
7 days
10
15 days
8 27 days
6 57 days
4 84 days
2 203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(b)
Relative displacement (mm)

14
Elevation 3 (2.0 m) Ninth layer
12 End of construction
10 7 days
8 15 days
6 27 days
4 57 days
84 days
2 203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(c)
Relative displacement (mm)

14
Elevation 2 (1.2 m) Sixth layer
12 Ninth layer
10 End of construction
8 7 days
15 days
6 27 days
4 57 days
84 days
2
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(d)
Relative displacement (mm)

14 Fourth layer
Elevation 1 (0.4 m) Sixth layer
12
Ninth layer
10 End of construction
8 7 days
6 15 days
27 days
4
57 days
2 84 days
0 203 days
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(e)
Figure 19. Reinforcements displacements measured by extensometers: (a) elevation 1; (b) elevation 2; (c) elevation 3; (d)
elevation 4; (e) elevation 5

of the total registered displacements), reaching 21 mm in the material characterization programs, deformability data,
the center of the reinforced zone (extensometer 2), and predicted strains, and design implications of the monitored
approximately 15 mm towards both edges of the rein- results are also presented. The field monitoring program
forced zone (extensometers 1 and 3). The displacements involved measurement of vertical and horizontal displace-
did not increase significantly with time (2 mm after 200 ments within the reinforced soil mass, as well as face
days). displacements. The measurements were conducted both
during and after construction of the wall. The main
conclusions from the evaluation of the performance of the
7. CONCLUSIONS geotextile-reinforced wall are as follows.
This study presents a description of the characteristics and
construction of a prototype geotextile-reinforced wall. • Most lateral displacements occurred towards the wall
This wall is the baseline structure for a series of eight face, reaching 13 mm at mid-height of the wall, as
reinforced prototype walls constructed to assess the de- measured by horizontal extensometers. The lateral
formability of geotextile-reinforced walls. An overview of displacements were observed to increase with time,
Geosynthetics International, 2007, 14, No. 2
Field monitoring evaluation of geotextile-reinforced soil-retaining walls 115
1.6
End of construction
Elevation 5 (3.6 m) 7 days
1.2

Strain (%)
15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (cm)
(a)
1.6
End of construction
Elevation 4 (2.8 m) 7 days
1.2
Strain (%)

15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(b)
1.6
Ninth layer
Elevation 3 (2.0 m) End of construction
1.2
Strain (%)

7 days
15 days
0.8 27 days
57 days
0.4 84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(c)
1.6 Sixth layer
Elevation 2 (1.2 m) Ninth layer
1.2 End of construction
Strain (%)

7 days
0.8 15 days
27 days
0.4 57 days
84 days
203 days
0
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(d)
1.6 Fourth layer
Elevation 1 (0.4 m) Sixth layer
1.2 Ninth layer
Strain (%)

End of construction
0.8 7 days
15 days
27 days
0.4 57 days
84 days
0 203 days
3.0 2.5 2.0 1.5 1.0 0.5 0
Distance from face (m)
(e)

Figure 20. Reinforcement strains: (a) elevation 1; (b) elevation 2; (c) elevation 3; (d) elevation 4; (e) elevation 5

particularly in association with precipitation events and location of the peak strain in each reinforcement
during the summer season in Brazil. layer.
• The horizontal displacements at the face, surveyed • The location of maximum reinforcement strains
from a fixed external reference, were larger than within each layer is consistent with the development
those measured by extensometers. The external face of a potential failure surface starting at the toe of the
displacements, taken in the middle of each layer, wall and propagating into the soil mass.
incorporated localized displacements due to soil • The maximum reinforcement tension predicted using
particle rearrangement at the flexible facing (sag- the K-stiffness method compared relatively well with
ging). The location of higher displacements observed reinforcement tension values obtained using the
after construction (mid-height of the wall) is measured displacement values in prototype wall 1.
consistent with that obtained using extensometers. • The largest settlements at the foundation level
• The reinforcement strains were comparatively small. occurred in the middle of the reinforced zone.
The largest strains (0.8%) occurred at mid-height of • Overall, the geotextile-reinforced wall showed a
the prototype wall. Additional strains took place after performance that is consistent with current design
construction, reaching a maximum value of 1.3% methods and, in spite of the low reduction factors
after 203 days. considered in the design of the prototype wall,
• The use of sigmoid curves to fit displacement results resulted in comparatively small displacements during
was found to be very useful for defining the value and after construction.
Geosynthetics International, 2007, 14, No. 2
116 Benjamim et al.
Strains calculated directly Sigmoid fitted curves
End of construction After 203 days
4.0
4

Predicted Tmax
3.2

Elevation (m)
2.4

1.6

Height (m)
2

0.8

0
4.0 3.2 2.4 1.6 0.8 0 1
Face distance (m)

Figure 21. Potential slip surface, defined as locus of maxi-

mum strains
0
0 0.5 1.0 1.5
0
Tmax (kN/m)

Figure 23. Tensile loads in reinforcement layers

0.3

COV
coefficient of variation (standard deviation/
mean) (dimensionless)
Z/H
d facing batter coefficient (dimensionless)
Dtmax reinforcement load distribution factor (¼
Tmax /Tmxmx ) (dimensionless)
d displacement from sigmoid curve-fitting (m)
emax , emin maximum and minimum void ratio
0.8 (dimensionless)
H height of wall (m)
1.0 i counter (1, 2, 3, . . ., n)
0 0.2 1.0
Dtmax J, Ji tensile stiffness of reinforcement (N/m)
Jave average tensile stiffness of reinforcement
Figure 22. Distribution of Dtmax with normalized depth below layers in wall (N/m)
wall top (Allen and Bathurst 2003, Bathurst et al. 2005) K coefficient of lateral earth pressure
(dimensionless)
Kabh coefficient of horizontal active earth pressure
accounting for wall batter (dimensionless)
ACKNOWLEDGMENTS Kabv coefficient of horizontal active earth pressure
This research received financial support from Ober S.A., for ø ¼ 0 (dimensionless)
FAPESP and CAPES in Brazil. Support received by the N Standard Penetration Test blow count
third author from the National Science Foundation in the (dimensionless)
USA under Grant No. CMS-0070248 is also acknowl- n total number of reinforcement layers in wall
edged. section (dimensionless)
pa 101 kPa (atmospheric pressure)
q surcharge pressure (Pa)
S equivalent soil height of uniform surcharge
NOTATIONS pressure ( ¼ q/ª) (m)
Basic SI units are given in parentheses. Sglobal global reinforcement stiffness value (N/m2 )
Slocal local reinforcement stiffness value (N/m2 )
a, b, c sigmoid curve-fitting constants Sv tributary area for reinforcement layer
(dimensionless) (assumed equivalent to vertical spacing of
a local reinforcement stiffness coefficient reinforcement when analyses are carried out
(dimensionless) per unit length of wall) (m)

Geosynthetics International, 2007, 14, No. 2

Field monitoring evaluation of geotextile-reinforced soil-retaining walls 117
4.5

End of construction
1 mm
3.5 Extensometer 3 Extensometer 2 Extensometer 1

Magnet 5 15 days

vertical displacement (mm)

δv,max ⫽ 1 mm
2.5
Elevation (m)

Magnet 4 27 days

1.5
δv,max ⫽ 6 mm
Magnet 3
57 days

δv,max ⫽ 6 mm
0.5
Magnet 2
84 days
δv,max ⫽ 4 mm
Magnet 1 Reference Foundation
⫺0.5
4 3 2 1 0
Distance from face (m)

Figure 24. Vertical displacements in reinforced backfill, as measured by magnetic extensometers

0
Extensometer 3 Extensometer 2 Extensometer 1
Sandy soil End of construction

⫺0.5 Foundation Reinforced

base

1 mm 15 days
Elevation (m)

Displacement (mm)

27 days

57 days

16 17

84 days
22

4 3 2 1 0
Distance from face (m)

Figure 25. Vertical settlements at foundation

Tmax maximum reinforcement load in a layer max maximum reinforcement strain in layer
(N/m) (dimensionless)
Tmxmx maximum reinforcement load from all layers Öfb facing batter factor (dimensionless)
in wall (N/m) Öfs facing stiffness factor
x location along reinforcement length (m) (dimensionless)
xmax maximum displacement location (m) Ög global stiffness factor (dimensionless)
ª unit weight of soil (N/m3 ) Ölocal local stiffness factor (dimensionless)

interface friction angle (degrees)  friction angle of soil (degrees)
Geosynthetics International, 2007, 14, No. 2
 118 Benjamim et al.

REFERENCES Elias, V., Christopher, B. R. & Berg, R. R. (2001). Mechanically

Stabilized Earth Walls and Reinforced Soil Slopes: Design and
Allen, T. M & Bathurst, R. J. (2003). Prediction of Reinforcement Loads Construction Guidelines, Report No. FHWA-NHI-00-043. Federal
in Reinforced Soil Walls. Report WA-RD 522.2, Washington State Highway Administration, Washington, DC, USA, 394 pp.
Department of Transportation, Olympia, Washington. Farrag, K. & Morvant, M. (2004). Evaluation of Interaction Properties of
Allen, T. M, Bathurst, R. J. & Berg, R. R. (2002). Global level of safety Geosynthetics in Cohesive Soils: LTRC Reinforced-Soil Test Wall,
and performance of geosynthetic walls: an historical perspective. LTRC Project No. 92-4GT. State Project No. 736-99–0658.
Geosynthetics International, 9, Nos. 5–6, 395–450. Louisiana Department of Transportation and Development and
Arthur, J. R. F. & Roscoe, K. H. (1965). An examination of the edge Louisiana Transportation Research Center, 156 pp.
effects in plane-strain model earth pressure tests. Proceedings of the Gomes, R. S. (1992). Interação solo-reforço e mecanismos de ruptura em
6th International Conference on Soil Mechanics and Foundation solos reforçados com geotêxteis. PhD thesis, Escola de Engenharia
Engineering, Montreal, Canada, Vol. II, Div. 3-6, pp. 363–367. de São Carlos,USP, Universidade de São Paulo, São Carlos, Brazil,
ASTM D 2487. Standard Practice for Classification of Soils for 271 pp.
Engineering Purposes (Unified Soil Classification System). ASTM ISO (1999). ISO 13341: Geotextile and Geotextile Related Products:
International, West Conshohocken, PA. Determination of Tensile Creep and Creep Rupture Behavior.
ASTM D 4253. Standard Test Methods for Maximum Index Density and International Organization for Standardization, Geneva, Switzer-
Unit Weight of Soils Using a Vibratory Plate (Unified Soil land.
Classification System). ASTM International, West Conshohocken, Kamiji, T. S. M. M. (2006). Fluência de geotêxteis não tecidos através
PA. de ensaios confinados. MSc dissertation, Escola de Engenharia de
ASTM D 4254. Standard Test Methods for Minimum Index Density and São Carlos, Universidade de São Paulo, São Carlos, Brazil, 134 pp.
Unit Weight of Soils and Calculation of Relative Density (Unified Mitchell, J. K. & Villet, W. C. B. (1987). Reinforcement of Earth Slopes
Soil Classification System). ASTM International, West Conshohock- and Embankments. National Cooperative Highway Research
en, PA. Program Report, No. 290, Transportation Research Board, National
ASTM D 4595. Standard Test Method for Tensile Properties of Research Council, Washington, DC, USA.
Geotextiles by the Wide-Width Strip Method (Unified Soil Nogami, J. S. & Villibor, D. F. (1981). Uma nova classificação de solos
Classification System). ASTM International, West Conshohocken, para finalidades rodoviárias. Simpósio Brasileiro de Solos Tropicais
PA. em Engenharia, Rio de Janeiro, Vol. 1, pp. 30–40.
Baras, L. C. S., Bueno, B. S. & Costa, C. M. L. (2002). On the Pedroso, E. O., Bueno, B. S., Benjamim, C. V. S. & Zornberg, J. G.
evaluation of stepped isothermal method for characterizing creep (2006). Field monitoring and numerical prediction of the response
properties of geotextiles. Proceedings of the 7th International of a non-woven geotextile-reinforced wall. Proceedings of the 8th
Conference on Geosynthetics, Nice, France, Vol. 4, pp. 1515–1518. International Conference on Geosynthetics, Yokohama, Japan, Vol.
Bathurst, R. J., Benjamin, D. J. & Jarret, P. M. (1989). An instrumented 3, pp. 1129–1132.
geogrid reinforced soil wall. Proceedings of the 12th International Ribeiro, T. S. M. T., Vicenzo, M. C. Jr & Pires, J. V. (1999).
Conference on Soil Mechanics and Foundation Engineering, Rio de Comportamento do aterro de solo reforçado da encosta do
Janeiro, Vol. 2, pp. 1223–1226. Belvedere. Proceedings of the 9th Congresso Brasileiro de Geologia
Bathurst, R. J, Allen, T. M. & Walters, D. L. (2005). Reinforcement loads de Engenharia, São Pedro, Brazil, CD-ROM.
in geosynthetic walls and the case for a new working stress design Vidal, H. (1966). La Terre Armée: réalisations récentes. Annales de
method. Geotextiles and Geomembranes, 23, No. 4, 287–322. l’Institut Technique du Batiment et des Travaux Publics, 21, Nos.
Bathurst, R. J., Vlachopoulos, N., Walters, D. L., Burgess, P. G. & Allen, 259 (July)–260 (August).
T. M. (2006). The influence of facing rigidity on the performance Wu, J. T. H. (1992). Predicting performance of the Denver walls: general
of two geosynthetic reinforced soil retaining walls. Canadian report. In Proceedings of the International Symposium on
Geotechnical Journal, 43, No. 12, 1225–1237. Geosynthetic-Reinforced Soil Retaining Walls, Denver, CO, USA.
Carvalho, P. A. S., Wolle, C. M. & Pedrosa, J. A. B. A. (1986). Aterro Balkema, Rotterdam, The Netherlands, pp. 3–20.
reforçado com geotêxteis – uma opção alternativa para a Zornberg, J. G. & Arriaga, F. (2003). Strain distribution within
engenharia geotécnica. Proceedings of the 8th Brazilian Conference geosynthetic-reinforced slopes. Journal of Geotechnical and
on Soil Mechanics and Foundation Engineering, Porto Alegre, Geoenvironmental Engineering, ASCE, 129, No. 1, 32–45.
Brazil, Vol. 4, pp. 169–178. Zornberg, J. G. & Mitchell, J. K. (1994). Reinforced soil structures with
Christopher, B. R. (1993). Deformation Response and Wall Stiffness in poorly draining backfills. Part I: Reinforcement interactions and
Relation to Reinforced Soil Design. PhD thesis, Purdue University. functions. Geosynthetics International, 1, No. 2, 103–148.
Cruz, P. T. (1996). 100 Brazilian Dams. Oficina de Textos, São Paulo (in Zornberg, J. G., Sitar, N. & Mitchell, J. K. (1997). Failure of steep
Portuguese). reinforced soil slopes. Proceedings of the Geosynthetics ’97
Ehrlich, M., Vidal, D. & Carvalho, P. A. (1997). Performance of two Conference, Long Beach, CA, Vol. 1, pp. 55–72.
geotextile reinforced soil slopes. Proceedings of the International Zornberg, J. G., Sitar, N. & Mitchell, J. K. (1998). Performance of
Symposium on Recent Developments in Soil and Pavement geosynthetic reinforced slopes at failure. Journal of Geotechnical
Mechanics, Rio de Janeiro. Balkema, Rotterdam, pp. 415–420. and Geoenvironmental Engineering, ASCE, 124, No. 8, 670–683.

The Editors welcome discussion on all papers published in Geosynthetics International. Please email your contribution to
[email protected] by 15 October 2007.

Geosynthetics International, 2007, 14, No. 2

View publication stats