See discussions, stats, and author profiles for this publication at: [Link]

net/publication/324832136

MODELLING OF PULLOUT RESISTANCE OF CONCRETE ANCHOR BLOCK

EMBEDDED IN COHESIONLESS SOIL

Thesis · May 2016

CITATIONS READS

2 121

1 author:

Rowshon Jadid
North Carolina State University
7 PUBLICATIONS   16 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Pullout resistance of anchor block View project

All content following this page was uploaded by Rowshon Jadid on 29 April 2018.

The user has requested enhancement of the downloaded file.

 MODELLING OF PULLOUT RESISTANCE OF CONCRETE
ANCHOR BLOCK EMBEDDED IN COHESIONLESS SOIL

by
Rowshon Jadid

A thesis submitted to the Department of Civil Engineering, Bangladesh University of

Engineering and Technology, Dhaka, in partial fulfilment of the degree of Master of Science in
Civil and Geotechnical Engineering.

DEPARTMENT OF CIVIL ENGINEERING

BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY

May, 2016
iii
iv
 TABLE OF CONTENTS

DECLARATION............................................................................................................. iv
TABLE OF CONTENTS ................................................................................................ v
LIST OF FIGURES .......................................................................................................vii
LIST OF TABLES .......................................................................................................... ix
NOTATION ...................................................................................................................... x
ACKNOWLEDGEMENT .............................................................................................. xi
ABSTRACT ....................................................................................................................xii
CHAPTER 1: INTRODUCTION
1.1 General ................................................................................................................ 1
1.2 Background and Present State of the Problem .................................................... 1
1.3 Objective of the Research ................................................................................... 3
1.4 Organization of the Thesis .................................................................................. 3
CHAPTER 2: LITERATURE REVIEW
2.1 General ................................................................................................................ 5
2.2 Previous Experimental Investigations ................................................................. 5
2.2.1 Anchor plate ............................................................................................ 6
2.2.2 Anchor block ........................................................................................... 6
2.3 Previous Theoretical Investigation .................................................................... 10
2.3.1 Ovesen and Stromann (1972) ................................................................ 10
2.3.2 NAVFAC DM 7.02 (U.S. Navy, 1986) ................................................. 13
2.3.3 BS 8006 (British Standard, 1995).......................................................... 14
2.3.4 Ghaly (1997) .......................................................................................... 15
2.3.5 Bowles (1997)........................................................................................ 15
2.3.6 Naser (2006) .......................................................................................... 16
2.4 Previous Numerical Investigations ................................................................... 17
2.5 Factors Affecting the Pullout Capacity ............................................................. 20
2.6 Concluding Remarks ......................................................................................... 21
CHAPTER 3: THEORITICAL ANALYSIS OF ANCHOR BLOCK
3.1 General .............................................................................................................. 22
3.2 Principle of the Proposed Method ..................................................................... 22
v
 3.2.1 Geometry of the failure surface ............................................................. 23
3.2.2 Forces acting on anchor block ............................................................... 25
3.2.3 Friction between soil and structure ........................................................ 26
3.3 Proposed Analysis for Anchor Block Capacity ................................................ 27
3.3.1 Computation of soil reaction at the bottom of the wedge...................... 27
3.3.2 Computation of pullout capacity of anchor block ................................. 28
3.3.3 Computation of forces acting on the anchor block ................................ 29
3.4 Comparison with the Experimental Results ...................................................... 31
3.5 Break-Out Factors ............................................................................................. 35
3.6 Concluding Remarks ......................................................................................... 38
CHAPTER 4: NUMERICAL ANALYSIS OF ANCHOR BLOCK
4.1 General .............................................................................................................. 39
4.2 Problem Definition ............................................................................................ 39
4.3 Finite Element Model ........................................................................................ 40
4.3.1 Modelling of anchor block .................................................................... 43
4.3.2 Modelling of cohesionless soil .............................................................. 44
4.3.3 Interface behaviour ................................................................................ 44
4.3.4 Modulus of elasticity of sand................................................................. 44
4.3.5 Dilatancy angle of cohesionless soil ...................................................... 45
4.4 Results and Discussions .................................................................................... 45
4.4.1 Mesh sensitivity analysis ....................................................................... 45
4.4.2 Effect of aspect ratio .............................................................................. 46
4.4.3 Effect of ground water table .................................................................. 48
4.4.4 Effect of embedment depth ratio ........................................................... 50
4.5 Failure Modes of Soil ........................................................................................ 52
4.6 Remarks on Mohr-Coulomb Model .................................................................. 54
CHAPTER 5: CONCLUSIONS AND RECOMMENDATIONS
5.1 General .............................................................................................................. 57
5.2 Conclusions ....................................................................................................... 57
5.3 Recommendations for Future Study.................................................................. 59
REFERENCES ............................................................................................................... 61
APPENDIX ..................................................................................................................... 66

vi
 LIST OF FIGURES

Figure 1.1 Anchored retaining earth wall (Khan and Sikder, 2004). ............................................. 2
Figure 2.1 Problem notation: (a) Front view, (b) Side view of an anchor. .................................... 6
Figure 2.2 Test arrangement for passive pressure load tests (Duncan and Mokwa, 2001) ........... 8
Figure 2.3 Computed and measured load-deflection curves for passive pressure load tests
(Duncan and Mokwa, 2001) .......................................................................................................... 8
Figure 2.4 Load-displacement curves for anchor block in sand of different moisture conditions
(Naser, 2006). ................................................................................................................................ 9
Figure 2.5 Variation of ultimate pullout resistance with angle of internal friction of soil
(Mostofa, 2013). .......................................................................................................................... 10
Figure 2.6 Basic case: continuous vertical anchor in granular soil (Das, 2007).......................... 11
Figure 2.7 Variation of with (Ovesen and Stromann, 1972). ........................... 11
Figure 2.8 Strip case: vertical anchor (Das, 2007) ...................................................................... 11
Figure 2.9 Actual case for row anchors (Das, 2007) ................................................................... 12
Figure 2.10 Variation of with (Ovesen and Stromann,
1972). ........................................................................................................................................... 12
Figure 2.11 Effect of anchor location relative to the wall (adapted from NAVFAC DM 7.02,
1986). ........................................................................................................................................... 13
Figure 2.12 Active and passive earth pressure calculation for continuous anchor wall located
between rupture surface and slope at friction angle (adapted from NAVFAC DM 7.02, 1986). 13
Figure 2.13 Design criteria for deadman anchorage (adapted from NAVFAC DM 7.02, 1986).
..................................................................................................................................................... 14
Figure 2.14 Complete set of forces, not including , acting on any anchor block (Bowles,
1997). ........................................................................................................................................... 16
Figure 2.15 Free body diagram of anchor block (Naser, 2006). .................................................. 16
Figure 2.16 Failure pattern and zones of plastic yielding for rough anchor plates by Rowe and
Davis (1982). ............................................................................................................................... 18
Figure 2.17 Failure pattern and zones of plastic yielding for rough anchor plates by Merifield
and Sloan (2006). ......................................................................................................................... 19
Figure 2.18 Failure surface in front of a square anchor slab (150 mm x 150 mm) embedded in
sand at as observed by Hueckel (1957). ...................................................................... 20

vii
Figure 3.1 Structure and soil movement on the verge of pullout failure of an anchor block
(Duncan and Mokwa, 2001). ....................................................................................................... 23
Figure 3.2 Idealized failure zone in front of single anchor with acting forces. (a) Elevation, (b)
Plan section of an anchor block. .................................................................................................. 24
Figure 3.3 Structure and soil movements for heavy anchor block (Duncan and Mokwa, 2001).27
Figure 3.4 Determination of weight of assumed failure wedge. (a) Elevation, (b) Plan section of
wedge. .......................................................................................................................................... 29
Figure 3.5 Break-out factors for anchor block in cohesionless soil (for ). ................. 37
Figure 3.6 Break-out factors for anchor block in cohesionless soil (for ). ............... 37
Figure 4.1 Problem definition (a) Front view, (b) Side view of an anchor. ................................ 39
Figure 4.2 Geometry of Finite Element model. ........................................................................... 40
Figure 4.3 Distribution of nodes and stress points in a 15-node wedge element. ....................... 41
Figure 4.4 Distribution of elements around the anchor block. .................................................... 41
Figure 4.5 Mesh sensitivity analysis. ........................................................................................... 45
Figure 4.6 Load-displacement curve for the study of the effect of aspect ratio ( ). ............... 46
Figure 4.7 Effect of aspect ratio (L/B) on pullout capacity of anchor. ........................................ 47
Figure 4.8 Effect of water table and correction factors. .............................................................. 48
Figure 4.9 Load-displacement curves for different locations of ground water table. .................. 49
Figure 4.10 Variation of correction factors with the location of ground water table. ................. 50
Figure 4.11 Load-displacement curves for different embedment depth ratio. ............................. 51
Figure 4.12 Comparison of break-out factors obtained from FE analyses with the proposed
method along with other methods suggested by different authors for =35o. ............................ 52
Figure 4.13 Failure modes and zones of plastic yielding for anchor block in cohesionless soils
as predicted by PLAXIS for =3 and =35o (a) vertical section and (b) at ground surface. 53
Figure 4.14 Velocity fields during collapse at mid-level of anchor block for =3 and =35o.
..................................................................................................................................................... 53
Figure 4.15 Load-displacement curves as predicted by PLAXIS for the experimental studies by
Duncan and Mokwa (2001) ......................................................................................................... 54
Figure 4.16 Load-displacement curves as predicted by PLAXIS for the experimental studies by
Naser (2006). ............................................................................................................................... 55
Figure 4.17 Load-displacement curves as predicted by PLAXIS for the experimental studies by
Mostofa (2013) ............................................................................................................................ 55

viii
 LIST OF TABLES

Table 2.1 Laboratory model tests and field tests on vertical anchor plate in cohesionless soil
(Modified from Merifield and Sloan, 2006) .................................................................................. 7
Table 2.2 Laboratory model tests and field test on vertical anchor block in cohesionless soil. .... 7
Table 2.3 Numerical studies on vertical anchors in cohesionless soil (Modified from Merifield
and Sloan, 2006) .......................................................................................................................... 17
Table 3.1 Anchor block and soil parameters. .............................................................................. 32
Table 3.2 Comparison of theoretical predictions of pullout capacity with experimental results
for vertical anchor block. ............................................................................................................. 33
Table 3.3 Cumulative frequency distribution of errors for vertical anchor block. ...................... 33
Table 3.4 Mean absolute percentage error (MAPE) of different methods. ................................. 34
Table 4.1 Geometry and mechanical properties in FE modelling. .............................................. 43
Table 4.2 Anchor block and soil parameters to investigate the effect of aspect ratio. ................ 46
Table 4.3 Correction factors for different locations of ground water table. ................................ 48
Table 4.4 Pullout resistance of anchor block for different embedment ratio. ............................. 52
Table 4.5 Anchor and soil parameters for experimental studies.................................................. 56

ix
 NOTATION

Height of anchor block

Modulus of elasticity of concrete anchor block

Initial stiffness of soil

Unloading/reloading stiffness of soil

Secant stiffness of soil

Finite element

Depth of embedment of anchor block

Embedment depth ratio of anchor block

Hardening soil model

Failure point at load-displacement curve

Length of anchor block

Aspect ratio of anchor block

Ultimate pullout capacity of anchor block

Power for stiffness stress dependency

Mohr-Coulomb model

Mean Absolute Percentage Error

Break-out factor

Interface reduction factor

Thickness of anchor block

 angle of side flanks of the wedge

 angle of Rankine passive failure surface with the vertical face of block

angle of friction between soil and top surface of the block

angle of friction between soil and bottom surface of the block

angle of friction between soil and side surface of the block

x
 ACKNOWLEDGEMENT

I would like to express my sincere gratitude to my supervisor Dr. Md. Zoynul Abedin,
Professor, Department of Civil Engineering, BUET. He often went beyond the call of duty in
encouraging, participation and supporting the goals of the research and providing his expertise.
His devotion for guidance and constant encouragement strongly supported me to complete the
present research work in this manner.

I would also like to thank Dr. Abdul Muqtadir, Dr. Abu Siddique and Dr. Md. Abu Taiyab for
their valuable time as members of my advisory committee.

I would like to extend my thanks to all my friends who helped me during my study. Finally, I
greatly appreciate my family for their enduring love and support.

xi
 ABSTRACT

Anchor block is a specially designed concrete member intended to withstand pullout or

thrust forces from backfill material of an anchored earth retaining wall by passive
resistance of soil in front of the block. Present study describes the theoretical and
numerical investigations into the behaviour of an anchor block in cohesionless soil. While
analysing theoretically, a passive wedge of soil was assumed to develop in front of the
anchor block due to the pullout force exerted from retaining wall via rebar. The
equilibrium condition of the wedge was employed to propose a new analytical method to
estimate the pullout capacity of an anchor block embedded in shallow depth. The
comparison of the proposed theoretical predictions with the existing theoretical and
experimental studies shows that, the proposed method provides a better estimate of the
pullout capacity. Furthermore, the theoretical results are facilitated with charts which may
be used in hand calculations to obtain an estimate of anchor capacity for most frequently
used shapes, cube and square with half of height/length of block as thickness. The use of
these charts is illustrated by worked examples.

Finite Element (FE) analysis was also conducted using PLAXIS to investigate the effect of
anchor shape, embedment depth and ground water table on pullout resistance of anchor
block. The effect of anchor shape was found to be considerable for small aspect ratio of
anchor block, whereas the influence of ground water table was substantial only when the
ground water table was located anywhere between the ground surface and the base of the
anchor block. Based on numerical analysis, an empirical correlation was developed to
determine the correction factor that accounts the effect of ground water table. The
proposed analytical method showed very close agreement with FE analysis for shallow
embedment depth of anchor block. In addition, the failure modes of the soil body were also
observed. These findings may be useful to all those dealing with civil engineering projects
and research works on anchored retaining earth wall.

xii
 1 CHAPTER 1
INTRODUCTION

1.1 General

In many instances, the design of civil engineering structures requires that the foundation
systems withstand the horizontal pullout forces. In such circumstances, an efficient and
economic design solution may be obtained through the use of tension members. These
tension members, which are denoted as earth anchors, are typically attached with the
retaining structure and embedded in the soil to sufficient depth so that they can resist
pullout forces with safety. They are generally used to transmit tensile forces from a
structure to the foundation soil thus generating passive support to structures like bulkheads,
sheet piles and retaining walls. Their pullout capacity is obtained considering shear
strength and dead weight of the surrounding soil mass. The use such anchors in retaining
structure may reduce the construction cost as much as 43 to 64% (Khan and Sikder, 2004).
However, inadequate anchorage systems constitute most of the causes of failure of sheet-
pile walls (Daniel and Olson, 1982; Sowers and Sowers, 1973). Various types of anchor
include anchor plates and beams (deadman), tie backs, anchor plates, anchor beams
supported by batter piles, and block anchors. According to Bowles (1997), block anchors
are cast-in-place or precast concrete members as shown in Figure 1.1, that may be square
or rectangular in section with necessary length to develop adequate passive resistance.
Anchor block can be installed by excavating the ground to the required depth, placing the
anchor, and then backfilling with soil. When used as a support for retaining structures,
anchors are installed in excavated trenches and connected to tie rods that may be driven or
placed through augered holes. This type of anchor is of interest in the present study.

1.2 Background and Present State of the Problem

In Bangladesh, two anchored earth walls have been implemented; one in Kalyanpur, Dhaka
and the other in BCIC industrial area, Narayanganj for establishing a substation for Dhaka
Electric Supply Authority (DESA) in 2002 and 2003 respectively, where anchored systems
were designed according to the guidelines provided by BS 8006 (1995). This code states
1
that the ultimate pullout resistance of an anchor block is four times the passive pressure
force acting on anchor block (ignoring the insignificant amount of resistance offered by
rebar). Recently, Mostopha (2013) conducted extensive laboratory tests to investigate the
pullout capacity of concrete anchor block located at different horizontal distances from
yielding boundary wall. He found that the maximum resisting force of concrete anchor
block is always less than four times the Rankine’s passive thrust acting on the anchor
block, and recommended to use the passive resistance coefficient less than 4 to ensure safe
design. Although no unsatisfactory performance has been reported from those projects, an
arbitrarily chosen higher factor of safety was used while designing to ensure safety
(Mostofa, 2013).

Figure 1.1 Anchored retaining earth wall (Khan and Sikder, 2004)

Literature reveals that many researches have been conducted on the capacity of vertical
anchors especially for anchor plate including that by Hueckel (1957), Ovesen and
Stromann (1972), Neely, et al. (1973), Das (1975), Akinmusuru (1978), and Ghaly (1997)
etc. However, only few studies were found for block anchor, Bowles (1997), Duncan and
Mokwa (2001), Naser (2006) and Mostofa (2013).

Bowles’s (1997) calculated passive earth resistance against anchor blocks using
conventional lateral earth pressure theories, and is applicable for long and continuous wall
structure. According to Ovesen (1964), the passive earth pressure against short structures
(e.g. anchor block) is higher than those predicted by conventional theories, and the
difference can be quite significant. Thus, Bowles’s (1997) model yields lower value than
2
the actual pullout capacity of anchor block. Duncan and Mokwa (2001), Naser (2006) and
Mostofa (2013) conducted experimental studies on anchor block.

It is clear that the majority of past research has been experimentally based and, as a result,
current design practices are largely based on empiricism. In contrast, very few thorough
theoretical and numerical analyses have been performed to determine the ultimate pullout
loads of anchor block. Of the numerical studies that have been presented in the literature,
few can be considered as rigorous. In addition, hardly any researches were conducted on
the effect of ground water table.

To address these issues, this research work aims to propose a new theoretical approach to
calculate the ultimate pullout capacity of an anchor block embedded in cohesionless soil. A
finite element analysis will be conducted using PLAXIS software to substantiate the
proposed theoretical method. Moreover, an effort will be taken to make comparison
between different geotechnical models that can be used to design anchor block. The
comparison is useful, because in the daily practice of geotechnical engineering many
discussions arise on which model is most suitable in which situation. This is also
mainly due to the differences in results between different models. Sharp designs can reduce
cost massively.

1.3 Objective of the Research

The present study aims to achieve the following objectives:

i. To propose a new theoretical method which can be used to predict the ultimate
pullout capacity of anchor block embedded in cohesionless soil.
ii. Finite element analysis using PLAXIS software to investigate the effect of anchor
shape, embedment depth and ground water table on pullout resistance of anchor
block.
iii. To make comparison between different geotechnical models with laboratory/field
tests, so that suitable method or methods with greater accuracy and computationally
ease can be adopted for different conditions.

1.4 Organization of the Thesis

The thesis consists of five chapters. Chapter 1 is an introduction that includes the problem
statement and objective of this study along with the thesis organization. Chapter 2 presents
the literature review including the experimental, theoretical and numerical studies
3
conducted on anchor block and plate etc. Chapter 3 introduces the derivation of theoretical
approach to calculate the pullout capacity of anchor block. The proposed theoretical
approach is also compared with the existing theoretical and experimental studies.
Chapter 4 presents the numerical analyses on anchor block, which mainly focuses on the
effect of anchor shape, embedment depth and ground water table on pullout resistance.
Additionally, different constitutive models are compared to assess their suitability at
different state of affairs. The main conclusions drawn from the theoretical and numerical
analyses are presented in Chapter 5, along with the recommendations for future work.

4
 2 CHAPTER 2
LITERATURE REVIEW

2.1 General

In order to provide a satisfactory background of subsequent discussions, a summary of

research into vertical anchor plate/block behaviour is presented. A comprehensive
overview on the topic of anchors is given by Das (1990). Although the research work aims
to focus on the analysis of concrete anchor block in cohesionless soil, emphasis on vertical
anchor plate will be equally given, because most of the time anchor block is designed using
the principle of vertical anchor plate (Das, 2007). Research into the behaviour of soil
anchors can take one of three forms, namely experimental-based, theoretical-based and
numerical–based studies. The brief summary of existing research herein has been separated
based on this distinction. No attempt is made to present a complete bibliography of all
research; rather a more selective overall summary of research with greatest relevance to the
current study is presented. In addition, contributions made to the behaviour of circular,
multiple underreamed, or multihelix anchors have not been reviewed.

2.2 Previous Experimental Investigations

Although there are no entirely adequate substitutes for full-scale field testing, tests at the
laboratory scale have the advantage of allowing close control of at least some of the
variables encountered in practice. In this way, trends and behaviour patterns observed in
the laboratory can be of value in developing an understanding of performance at larger
scales. In addition, observations made in laboratory testing can be used in conjunction with
mathematical analyses to develop semi-empirical theories. These theories can then be
applied to solve a wider range of problems. In this section, a brief discussion will be
presented on anchor plate. Whereas the experimental studies on anchor block will be
discussed elaborately in a separate subsection, since the results of these studies will be
utilised in the subsequent chapters.

5
2.2.1 Anchor plate

Experimental investigations on plate anchor behaviour have generally adopted one of the
two approaches, namely, conventional methods under “normal gravity” conditions or
centrifuge systems. Both the methods have advantages and disadvantages, and these must
be borne in mind when interpreting the results from experimental studies of anchor
behaviour. Numerous investigators have performed model tests in an attempt to develop
semi-empirical relationships that can be used to estimate the capacity of anchor plates in
cohesionless soil. This is evidenced by the number of studies shown in Table 2.1.
Illustrations of notation used in this table are presented in Figure 2.1.

The works prior to 1970 have not been presented in Table 2.1. This includes the field and
(or) model testing of vertical square and rectangular anchors by Hueckel (1957), Smith
(1962), Smith [Link]. (1965) etc. In the majority of these earlier studies, a failure mechanism
was assumed and the pullout capacity was then determined by considering the equilibrium
of the soil mass around the anchor and contained by the assumed failure surface.

(a) (b)

Figure 2.1 Problem notation (a) Front view, (b) Side view of an anchor

2.2.2 Anchor block

In contrast of previous researches on anchor plates, only few experimental studies were
found for anchor block, Duncan and Mokwa (2001), Naser (2006) and Mostopha (2013).
Table 2.2 summarizes the experimental studies conducted on anchor block. Duncan and
Mokwa (2001) carried out passive pressure load tests at the Virginia Tech field test site at
Kentland Farms near Blacksburg.

6
 Table 2.1 Laboratory model tests and field tests on vertical anchor plate in cohesionless soil (Modified from Merifield and Sloan, 2006)
Source Type of testing Anchor shape Anchor Friction Anchor H/B
size (mm) angle (°) roughness (°)
Neely et. al. (1973) Chamber Square; rectangular 50.8 38.5 21 1–5
Das (1975) Chamber Square; circular 38–76 34 ? 1–5
Akinmusuru (1978) Chamber Strip; rectangular; square; circular; L/B = 2,50 24; 35 ? 1–10
10
Ovesen (1981) Centrifuge; field Square 20 29.5–37.7 ? 1–3.39
Rowe and Devis (1982) Sand Chamber Square; rectangular; L/B = 1–8.75 51 32 ? 1-8
Dickin and Leung (1983, 1985) Centrifuge Square; rectangular; strip 25; 50 41a Polished, 29 1–8; 1–13
chamber
Hoshiya and Mandal (1984) Sand chamber Square; rectangular; L/B = 2, 4, 6 25.4 29.5 ? 1–6
Murray and Geddes (1989) Sand chamber Square; rectangular; L/B = 1–10 50.8 43.6, dense 10.6 1–8
Dickin and King (1997) Centrifuge Rectangular; L/B = 7.8 25 37.3-46.1 ? 1-12
a
Mobilized plane strain friction angle, φ′mp

Table 2.2 Laboratory model tests and field test on vertical anchor block in cohesionless soil
Source Type of Anchor Block Anchor Block Friction Anchor H/B
testing shape size (mm) angle (°) roughness (°)
Duncan and Mokwa (2001) Field Rectangular; L/B = 1.7 1100 x 1900 x 900 50 6 1
Naser (2006) Chamber Square 150 x 150 x 150 43.5 7.3-11.2 2
Mostofa (2013) Chamber Square 150 x 150 x 75 37.2- 44.8 ? 3.2

7
The tests were performed using the test arrangements shown in Figure 2.2. Using a pile
group for reaction, horizontal loads were applied to a concrete anchor block 3.5-ft high,
6.3-ft long, and 3.0-ft thick (1100 x 1900 x 900 mm). Figure 2.3 shows comparisons of
measured and computed load-deflection curves, using the log spiral value of Pu corrected
for 3D effects proposed by Ovesen (1964).

Figure 2.2 Test arrangement for passive pressure load tests (Duncan and
Mokwa, 2001)

Figure 2.3 Computed and measured load-deflection curves for passive pressure load
tests (Duncan and Mokwa, 2001)
8
Naser (2006) performed laboratory pullout tests on 150  150  150 mm concrete block
anchors embedded in sand at a depth of 300 mm. The sand was deposited in a chamber of
dimensions 1200  800  600 mm using a pluviation method to ensure a uniform and
reproducible density. He compared the experimental results with the analytical calculations
by Rankine, Coulomb, and log spiral theories. Figure 2.4 shows typical the load-
displacement relationship for block anchor embedded in sand at dry, wet and saturated
conditions.

Figure 2.4 Load-displacement curves for anchor block in sand of different moisture
conditions (Naser, 2006)

Mostofa (2013) investigated the pullout capacity of concrete anchor block located at
different distances from yielding boundary wall and embedded in air dry sand, using small
scale laboratory experimental works. The experimental setup consisted of a large tank
made of fibre glass sheets and steel framing system. A series of tests was carried out in the
tank to investigate the load-displacement behaviour of anchor block. Figure 2.5 shows the
variation of ultimate resistance with angle of internal friction of soil. He also proposed the
passive resistance coefficient which should be multiplied with the passive resistance in
front of the anchor block to obtain the pullout capacity of anchor block.

More specific details of these studies will be provided in the later chapters of this thesis,
when several previous theoretical studies are compared with the new theoretical approach
proposed in the current study.

9
 Figure 2.5 Variation of ultimate pullout resistance with angle of internal friction of
soil (Mostofa, 2013)

2.3 Previous Theoretical Investigation

The previous theoretical studies post-1970 for vertical anchors, which are commonly used
by practicing engineers, will be discussed elaborately in this section. As mentioned earlier,
the theoretical methods of calculating pullout capacity of vertical anchor plate are
commonly employed for anchor block also (Das, 2007). Thus, no effort has been made
here to distinguish between anchor plate and anchor block design method. Previous
theoretical studies of anchors in sand have typically utilized simple analytical approaches
such as limiting equilibrium, cavity expansion, and limit analysis. In the limit equilibrium
method, a failure surface is assumed along with a distribution of stress along that surface.
Equilibrium conditions are then considered for the failing soil mass, and an estimate of the
collapse load is obtained.

2.3.1 Ovesen and Stromann (1972)

Ovesen and Stromann (1972) proposed a semi-empirical method for determining the
ultimate resistance of anchors in sand. When an anchor slab has a height of B, equal to the
depth of embedment H and is continuous (Figure 2.6), the ultimate resistance per unit
length of anchor, can be calculated from Eq. (2.1).

(2.1)

10
 Figure 2.6 Basic case continuous vertical anchor in granular soil (Das, 2007)

Where is the friction angle between anchor slab and soil, is the active earth pressure
coefficient with = , and is the passive earth pressure coefficient. The term
of Eq. (2.1) can be obtained from Figure 2.7 using the value of
and the angle of internal friction . Where W is the
effective weight per unit length of anchor slab.

Figure 2.7 Variation of with (Ovesen and Stromann, 1972)

Figure 2.8 Strip case vertical anchor (Das, 2007)

11
For a continuous strip anchor (Figure 2.8) of height B (which is less than the depth of
embedment, H), the ultimate resistance per unit length is corrected as:

(2.2)

Where is the ultimate resistance for strip case, = 19 for dense sand and 14 for loose
sand. In practice, the anchor plates are placed in a row with center-to-center spacing
(Figure 2.9). The ultimate resistance of each anchor of length L is
(2.3)

Where, is the equivalent length and can be obtained from Figure 2.10.

Figure 2.9 Actual case for row anchors (Das, 2007)

Figure 2.10 Variation of with (Ovesen and

Stromann, 1972)

12
2.3.2 NAVFAC DM 7.02 (U.S. Navy, 1986)

A step-by-step design procedure of vertical anchors can be found in the design manual of
Naval Facilities Engineering Command (NAVFAC DM 7.02, 1986) by U.S. Navy.
According to NAVFAC DM 7.02 (1986), the anchor blocks should be placed outside the
surface making an angle equal to angle of friction of backfill soil, with the horizontal as
shown in Figure 2.11. The anchor blocks locating between line and Rankine’s failure
surface will generate partial passive resistance. Design of anchor blocks using active and
passive earth pressures in accordance with the NAVFC DM 7.02 (1986) are briefly
illustrated in Figure 2.12 and Figure 2.13. A factor of safety 2 against pullout resistance is
suggested in this method.

Figure 2.11 Effect of anchor location relative to the wall (adapted from NAVFAC DM
7.02, 1986)

Figure 2.12 Active and passive earth pressure calculation for continuous anchor wall
located between rupture surface and slope at friction angle (adapted from NAVFAC
DM 7.02, 1986)
13
Figure 2.13 Design criteria for deadman anchorage (adapted from NAVFAC DM 7.02,
1986)

2.3.3 BS 8006 (British Standard, 1995)

BS 8006 (1995) recommends to use passive resistance coefficient while caculating the
pullout resistance of an anchor block. This is based on the fact that, the conditions at the
ends of the structures are quite different from those at the centre, which has significant
influence on the passive resistance. Ovesen (1964) found that the passive earth pressure
against short structures is higher than those predicted by conventional theories (Rankine
and Coulomb theories), and the difference can be quite significant. Hansen (1966)
developed a method for correcting the results of conventional pressure theories for shape
(or 3-D) effects. For short anchors, the ultimate resistance should be multiplied by a
coefficient (M) to account for 3-D effects. For a plate anchor passive resistance coefficient
(M) is given as:

(2.4)

Where, and are the coefficients of passive and active earth pressure respectively,

- and ,and is the center-to-center distance between two

anchors. The anchor block geometry used in Eq. (2.4) can be found in Figure 2.1. The
above equation considers both the embedment factor ( ) and the shape factor ( ). The
value of is 0.0 for long and continuous anchor, and is 1.0 for single short anchor.
According to the BS 8006 (1995) and Jones (1996), the horizontal pullout resistance of an

14
anchor block is 4 times the passive pressure force acting on anchor block (ignoring the
insignificant amount of resistance offered by rebar) i.e. passive resistance coefficient,
M = 4 is suggested. A factor of safety 2.5 to 3 is used in this method. However, the
experimental studies by Mostofa (2013) indicated that this coefficient is always less than 4.

2.3.4 Ghaly (1997)

Ghaly (1997) used the results of 104 laboratory tests, 15 centrifugal tests, and 9 field tests
to propose an empirical correlation for the ultimate resistance of single anchors, where unit
weight and internal friction angle of soil ranged from 14 to16 kN/m3 and 34o to 38.5o
respectively. The embedment depth ratio of the anchor plate varied between 1 and
4. This data was incorporated in a generalized from to predict the ultimate horizontal
pullout resistance of anchor plates in terms of the influencing parameters. The correlation
can be expressed as:

(2.5)

Where is the area of the anchor = .

2.3.5 Bowles (1997)

Based on horizontal equilibrium analysis of lateral earth pressures and frictional

resistances acting on an anchor block as depicted in Figure 2.14, Bowles (1997) proposed a
general equation to determine the horizontal pullout resistance of anchor as follows:

(2.6)

Where and are the passive and active thrust acting on the anchor block respectively,
and the frictional resistances ( and ) can be determined when the angle of friction
between soil and concrete ( and ) acting at the top and bottom of the anchor block as
shown in Figure 2.14 is known. For maximum efficiency, Bowles (1997) suggested to
locate the anchor block such that the Rankine passive zone in front of the anchor block
should be completely outside the Rankine active zone behind the retaining wall. A factor
of safety 1.2 to 1.5 is suggested for this method.

15
 Figure 2.14 Complete set of forces, not including , acting on any anchor block
(Bowles, 1997)

Figure 2.15 Free body diagram of anchor block (Naser, 2006)

2.3.6 Naser (2006)

Naser (2006) analysed pullout capacity of an anchor block using limit equilibrium
approach (Figure 2.15). The ultimate pullout capacity of block anchor block ( ) was
obtained from the equilibrium of forces acting on the block by summing them along the
horizontal direction and multiplying the lateral earth pressures (passive and active) by the
3-D correction factor M, to yield the following equation.

(2.7)

16
Where, , and are the effective friction forces at the top, bottom and at two side of
the block, is the effective horizontal passive thrust and is the effective horizontal
active thrust. The 3-D correction factor (M) may be calculated from Eq.(2.4). He found that
the pullout capacity of block anchor with Rankine’s theory (1857), corrected for the 3-D
effect with the contribution of friction showed a close agreement with experimental results.

2.4 Previous Numerical Investigations

Although there are a variety of experimental results in the literature, very few rigorous
numerical analyses have been performed to determine the pullout capacity of anchors in
sand (Merifield and Sloan, 2006). Most of those numerical analyses were performed for
strip anchor. On the other hand, to the knowledge of the author, hardly any effort has been
made so far to conduct numerical analysis on anchor block. Although it is essential to
verify theoretical solutions with experimental studies wherever possible, results obtained
from laboratory testing alone are typically problem specific. Since the cost of performing
laboratory tests on each and every field problem combination is prohibitive, it is necessary
to be able to model soil pullout resistance numerically for the purposes of design. A
summary of some important previous numerical studies post-1980 on vertical anchors is
provided in Table 2.3.

Table 2.3 Numerical studies on vertical anchors in cohesionless soil (Modified from
Merifield and Sloan, 2006)
Source Analysis method Anchor Anchor Friction H/B
shape roughness angle (°)
Rowe and Elastoplastic finite Strip Smooth 0–45 1–8
Davis (1982) element
Hanna et al. Limiting equilibrium Strip; ? All All
(1988) inclined
Murray and Limit analysis – upper Strip; Smooth; 43.6 1–8
Geddes (1989) bound inclined rough
Basudhar and Limit analysis – lower Strip Rough; 32; 35; 38 1–5
Singh (1994) bound smooth
Merifield and Limit analysis – upper Strip Rough 20-40 1-10
Sloan (2006) and lower bound

17
The most complete numerical study first appears to be that by Rowe and Davis (1982).
They described a theoretical assessment of anchor plates in sand which considered the
effect of anchor plate embedment, friction angle, dilatancy, and initial stress state and
anchor plate roughness for vertical anchor plates. Their theoretical solution was based on
an elasto-plastic finite element analysis using a soil structure interaction theory. The sand
was assumed to have a Mohr-Coulomb failure criterion. The theoretical data were
presented in the form of design charts which could be used in hand calculations to obtain
an estimate of anchor plate capacity for a wide range of anchor plate geometries and sand.
The failure pattern and zones of plastic yielding obtained from their numerical analysis is
shown in Figure 2.16. The finite element method has also been used by Vemeer and
Sutjiadi (1985), Tagaya et al. (1983, 1988) and Koutsabeloulis and Griffiths (1989), and
Sakai and Tanaka (1998) for horizontal anchor plates. However, limited results of these
studies were presented in literature.

Figure 2.16 Failure pattern and zones of plastic yielding for rough anchor plates by
Rowe and Davis (1982)

Tagaya et. al. (1983, 1988) conducted two dimensional plane strain and axisymmetric
finite element analyses by the constitutive law of Lade and Duncan (1975). Scale effects
for circular anchor plates in dense sand were investigated by Sakai and Tanaka (1998) by a
constitutive model for a non-associated strain hardening-softening elasto-plastic material.
18
Koutsabeloulis and Griffiths (1989) investigated the trap door problem by the initial stress
finite element method. Both plane strain and axisymmetric works were conducted. The
researchers concluded that an associated flow rule has little effect on the collapse load for
strip anchor plates but a significant effect (30%) for circular anchor plates. Large
displacements were observed for circular anchor plates prior to collapse.

Upper and lower bound limit analysis techniques have been used by Murray and Geddes
(1987, 1989) and Basudhar and Singh (1994) to estimate the capacity of vertical strip
anchor plates. Basudhar and Singh (1994) obtained estimates with a generalized lower
bound procedure based on finite element method and non-linear programming similar to
that of Sloan (1988). The solutions of Murray and Geddes (1987, 1989) were obtained by
typically constructing kinematic ally admissible failure mechanisms (upper bound).
Merifield et al. (2006) presented the results of a rigorous numerical work to estimate the
ultimate capacity load for vertical anchor plate in cohesionless material. Rigorous bounds
have been obtained using two numerical procedures that are based on finite element
method of the upper and lower bound of limit analysis. For comparison purposes,
numerical and theoretical results of the break-out factor have also been obtained by the
more conventional displacement finite element method. Results are presented in the
familiar form of break-out factors based on various soil strength profiles and geometries
and are compared with existing numerical and empirical solutions. The failure pattern and
zones of plastic yielding obtained from their numerical analysis is shown in Figure 2.17.

Figure 2.17 Failure pattern and zones of plastic yielding for rough anchor plates by
Merifield and Sloan (2006)
19
2.5 Factors Affecting the Pullout Capacity

The ultimate pullout capacity of an anchor block depends on several factors. Some of the
most important governing factors are-

(i) Embedment depth ratio

(ii) Aspect ratio or length-to- height ratio
(iii) Shear strength parameters of the soil (soil friction angle, and cohesion, ); and
(iv) The angle of friction at the anchor-soil interface,

Figure 2.18 Failure surface in front of a square anchor slab (150 mm x 150 mm)
embedded in sand at as observed by Hueckel (1957)

The pullout capacity of an anchor is primarily derived from the passive force imposed by
the soil in front of the anchor slab. If the embedment depth ratio of the anchor is
relatively small, at ultimate pullout load on the anchor the passive failure surface
developed in soil in front of the anchor will intersect the ground surface. This is referred to
as shallow anchor condition. Figure 2.18 shows the failure surface in front of a shallow
square plate anchor (that is, = ) embedded in sand as observed by Hueckel (1957). At
greater embedment ratios, the local shear failure in soil will take place at ultimate load, and
these anchors are called deep anchor (Das, 1990). The aspect ratio also affects the pullout
capacity of an anchor. If the aspect ratio of an anchor is large enough, the anchor can be
considered as strip anchor. Two-dimensional plane strain case is applicable for strip
anchor. However, for square anchor three-dimensional analysis is required as the boundary
conditions at the ends of failure surface is quite different than the strip anchor. The
numerical studies by Rowe and Davis (1982) and Merifield and Sloan (2006) show that
20
the interface roughness has a significant effect on pullout capacity of strip anchor.
Changing the interface roughness from perfectly rough to perfectly smooth can lead to a
reduction in the anchor capacity by as much as 65% (Merifield and Sloan, 2006). More
specific details of these factors will be discussed in the later chapters of this thesis.

2.6 Concluding Remarks

Literature review reveals that most of past research has been carried out experimentally
and, as a result, current design practices are largely based on empiricism. Consequently,
the results obtained from these empirical methods are generally case specific. Thus, a
generalized analytical approach is required to resolve this problem. Reported literature on
the effect of ground water table on pullout capacity are also very few. This is another
important design consideration that needs to be addressed. Although, many geotechnical
models are available regarding the design of vertical anchors, with today’s most used
models, comparisons between the models and measurements are very rare. The
comparison is useful as because in practice of geotechnical engineering many
discussions arise on which model is most suitable in which situation. This is also
mainly due to the differences in results between different models. This thesis, as such, is
intended to make a contribution to the design of anchored retaining earth wall by
conducting an extensive investigation on these issues.

21
 3 CHAPTER 3
THEORITICAL ANALYSIS OF ANCHOR BLOCK

3.1 General

From literature review, it is quite evident that many researches have been conducted on the
capacity of vertical anchors especially with anchor plate. However, studies on anchor block
are limited compared to the anchor plate (Naser, 2006). The theoretical methods of
calculating pullout capacity of vertical anchor plate are commonly employed for anchor
block also (Das, 2007). On the contrary, this study shows the limitations of these methods
while using them to calculate the pullout resistance of anchor block. Consequently, an
effort has been taken to propose an analytical method to calculate the pullout capacity of
anchor block, which is presented in the first part of the chapter. Later, the predictions of
the proposed method are compared with the laboratory and field test results available in the
literature along with other methods to assess the suitability of these methods.

3.2 Principle of the Proposed Method

In this method, a passive wedge model is employed to derive analytical expression to

calculate the ultimate pullout capacity of anchor block in cohesionless soil. The passive-
wedge model is based on the principle that normal stresses will increase in front of the
anchor block due to the pullout force exerted from retaining wall via rebar. The increasing
stresses cause the soil to move up, and a passive wedge will develop as schematized in
Figure 3.1. It is assumed that the resistance of the soil will be provided by the normal force
acting perpendicular to the bottom of the failure plane and the friction forces along the side
planes of the failure wedge. The normal force, self-weight of failure wedge and the friction
forces acting on the wedge can be determined from geometry of the wedge and properties
of the soil. Thus, the ultimate resistance of the soil then can be derived from the
equilibrium of the wedge. Passive wedge model was successfully used for the analysis of
anchor plate by Dickin and Leung (1985), for the analysis of laterally loaded pile by Gabr
and Borden (1990), Mirzoyan (2007) and Cheng et. al. (2014).

The proposed method is restricted to shallow laid anchor block in cohesionless soil. For
shallow anchors, the embedment ratio is such that, failure surface reaches the ground
22
surface at limit equilibrium; whereas for deep anchors, the embedment ratio is such that,
failure surface does not reach the ground surface at limit equilibrium (Das, 1990).

Figure 3.1 Structure and soil movement on the verge of pullout failure of an anchor
block (Duncan and Mokwa, 2001)

3.2.1 Geometry of the failure surface

The assumed failure surfaces in the proposed method and the forces acting in the failure
mechanism are illustrated in Figure 3.2. Three assumptions are made on the geometry of
failure surface.

(i) The passive failure surface in front of the anchor block is planer and reaching up to
the top level of anchor block. Zone ABC in Figure 3.2(a) is considered as a
Rankine passive zone.
(ii) The soil located above the top of the anchor is a simple surcharge, .
(iii) Failure planes around anchor radiate outwards. Thus, the soil wedge has an
extension with two side flanks, which have an angle as shown in Figure 3.2(b).

The first assumption was made for the simplicity of the analysis, although the actual failure
surface follows very irregular pattern and reaches up to the ground surface for shallow
anchor. The second assumption neglected the shearing resistance of soil above the top of
the anchor block, but the effect of soil weight above the base is considered by
superimposing an equivalent surcharge. Neely et. al. (1973) also neglected shearing
resistance of soil above anchor for the analysis of pullout resistance of strip anchor, while
Tarzaghi (1948) neglected soil resistance above footing level for the analysis of bearing
capacity of soil. Soil resistance of cohesionless soil is generally calculated from peak angle
of internal friction , that is obtained from plain strain shear tests or triaxial tests.
However, the failure of the soil mass is rather progressive, and the applicable value of is

23
moblized friction angle ( ), which is generally less than the peak value (Dickin and
Leung, 1985; Das, 1990). Therefore, neglecting shear resistance above anchor level,
somewhat balance the overestimation of pullout capacity due to the use of obtained
from test results. The third assumption was made from soil’s dilatancy property.

(a)

(b)
Figure 3.2 Idealized failure zone in front of single anchor with acting forces.
(a) Elevation, (b) Plan section of an anchor block

Many soils (e.g. dense sand) dilate during plastic deformation and this characteistic of the
soil may have a significant effect on anchor behaviour. Dilatancy during plastic
24
deformation tends to cause the soil in front of the anchor to lock up and it is necessary for
an extensive plastic region to develop before there is sufficient freedom for collapse to
occur (Rowe and Davis, 1982). Laboratory test reults and numerical analysis also
confirmed the formation of side wedges caused by dilation of soil (Rowe and Davis, 1982;
Dickin and Leung, 1983). As an approximation Reese, Cox, and Koop (1974) suggested to
use , where, is the angle of internal friction of soil. Mirzoyan (2007) conducted
field tests on laterally loaded pile and found that is about 75% of the angle of internal
friction.

3.2.2 Forces acting on anchor block

Figure 3.2 shows the elevation and plan section of a short anchor block embedded in sand,
with all forces acting on it and with friction considered. For an anchor block with the
dimensions height ( ), width ( ), thickness ( ) and the depth of embedment below soil
surface ( ); the acting forces are as follows:

= normal force at the bottom of the wedge

= friction force at the bottom of the wedge
= normal force at the side of the wedge
= friction force at the side of the wedge
= normal force at the bottom of the anchor block
= friction force at the bottom of the block
= friction force at the top of the block
= friction force at two sides of the block
= resultant active thrust
= weight of the wedge
= weight of the surcharge above wedge
= weight of the block anchor
= weight of soil above the block anchor
= total weight of the block-wedge system
= + + +

25
3.2.3 Friction between soil and structure

Anchor blocks are relatively light structure and are not supported by rigid structures. The
upward components of the passive resistance is large enough to cause upward movement
of the anchor block, and the soil and structure move together as illustrated in Figure 3.1
(Duncan and Mokwa, 2001). Thus, the weight of the anchor block and the weight of soil
above it should be considered for the vertical equilibrium of wedge (Duncan and Mokwa,
2001; Naser, 2006). Shear displacement between soil and structure is not expected to occur
in the vertical direction of plane AB, Figure 3.2 (a); accordingly, interface friction is
neglected along this direction. However, frictional resistance between soil and structure
develops in the horizontal direction, e.g. , , in Figure 3.2 (a), since the pullout force
from retaining wall causes horizontal shear displacement between soil and anchor block.

The magnitude of this type of resistance depends on the angle of friction between soil and
structure, The maximum possible value of depends on the roughness of the interface
and properties of the soil. It is convenient to characterize values of in terms of the
ratio , where is the angle of internal friction of soil. The value of varies
between 0 and 1 depending upon surface roughness, mean particle size of sand and method
of installation (CFEM, 2006; Tiwari, et al., 2010). Potyondy (1961) conducted interface
shear tests on a variety of structural materials and soils and found smallest value of
= 0.76 for concrete. As an approximation, Singh (1967) suggested following
values (1) = 1/3 for smooth structure (wall), (2) = 2/3 for ordinary
retaining wall, (3) = 3/4 for rough walls with well-drained backfill, and (4) = 0
when backfill is subjected to vibrations. The range of values of between fine sand and
concrete is 15-25o, Das (1995) and Bowles (1997); whereas Naser (2006) found the value
of as 38o. More recently, Gireesha and Muthukkumaran (2011); Tiwari and
Al-Adhadh (2014) conducted numerous laboratory tests to study soil-structure frictional
resistance, and observed that varies between 0.76-0.79 and 0.69-0.94 respectively
between concrete and sand depending upon the relative density and degree of saturation of
soil.

Some amount of relative shear displacement across the interface is required to mobilize
interface friction. The amount of relative shear displacement required to mobilize the full
strength of the interface is not large, typically not more than 0.1 – 0.25 in. (2.5 – 6 mm).
Smaller relative displacements across the interface will result in only partial mobilization

26
of the interface friction. Therefore, will be less than or equal to (Duncan and
Mokwa, 2001).

In other circumstances, the heavy anchor block (e.g. strip anchor) moves horizontally
while the soil moves both horizontally and upward. As a result of the upward movement of
the soil with respect to the structure, there is an upward shear force on the structure and a
downward shear force on the soil. In the case of a rough anchor, significant shear stresses
will develop at the anchor interface in response to this upward movement. These shear
stresses are resisted by the interface and contribute significantly to the anchor capacity
(Merifield and Sloan, 2006; Kame, et al., 2012).

Figure 3.3 Structure and soil movements for heavy anchor block (Duncan and
Mokwa, 2001)

3.3 Proposed Analysis for Anchor Block Capacity

3.3.1 Computation of soil reaction at the bottom of the wedge

The ultimate pullout capacity of block anchor can be obtained from the equilibrium of
forces acting on the block and wedge illustrated in Figure 3.2. First, we will consider the
vertical equilibrium of the block and wedge to determine the normal force at the bottom of
the wedge .

Thus, Force equilibrium in the vertical direction:

(3.1)
Substituting into Eq. (3.1):

(3.2)
Solving Eq. (3.2) for :

(3.3)

27
3.3.2 Computation of pullout capacity of anchor block

Force equilibrium in the horizontal direction:

 
(3.4)
Substituting into Eq. (3.4):
 
(3.5)
Rearranging the Eq. (3.5) for ultimate pullout capacity of anchor block ( ):
  
(3.6)
Now, substituting the value of from Eq. (3.3) into Eq. (3.6):

 

(3.7)
After simplification:

  (3.8)

Where,

(3.9)

Naser (2006) observed that, anchor thickness contributed to the pullout capacity through
friction forces. This contribution was not significant as compared to the passive resistance.
Jones (1996) also reported that the passive resistance has the greatest influence on pullout
capacity of anchor block, and it derives from the passive wedge in front of the anchor.
Thus, neglecting frictional forces and active thrust from Eq. (3.8), one may obtain
simplified and computationally easy form of Eq. (3.8) as follow:

(3.10)

The accuracy and the applicability of Eq. (3.8) and Eq. (3.10) will be discussed in the later
sections of this chapter.

28
 (a)

(b)
Figure 3.4 Determination of weight of assumed failure wedge. (a) Elevation, (b)
Plan section of wedge

3.3.3 Computation of forces acting on the anchor block

Equation (3.8) or (3.10) can be used to determine the pullout capacity of anchor block in
chohesionless soil. As mentioned earlier, total weight in Eq. (3.8) consists of four
weight components of block and wedge. These weight components along with other
contributing forces in Eq. (3.8) can be determined as follows:
Weight of wedge, Ww:
Referring to Figure 3.4, a small elemental strip of thickness at a height of from the
base of the block is considered. At this depth level the width and the length of the

29
elemental strip can be expressed as and  respectively. Thus, the
elemental weight, of the wedge can be obtained as follow
 (3.11)
Integrating of Eq. (3.11) over to gives

 (3.12)


(3.13)


(3.14)

Weight of the surcharge above wedge, Wq:

 (3.15)

Or,  (3.16)
Weight of the block anchor, Wb:
(3.17)
Weight of soil above the block anchor, WS:
(3.18)
Normal force at the side of the wedge, Fsn:
(3.19)


Friction force at the side of the wedge, Fsf :
(3.20)
Resultant active thrust, Pa:
(3.21)

Friction force at the top of the block, Ft:

(3.22)
Friction force at the bottom of the block, Fb:
(3.23)
Friction force at two sides of the block, Fs:
(3.24)

30
Where,
 = effective angle of internal friction of soil
 = effective unit of soil
= effective unit weight of concrete block = 23.6 kN/m3
 = angle of side flanks of the wedge (Reese, et al., 1974)
 = angle of Rankine passive failure surface with the vertical face of
 anchor block = 45o+
= coefficient of active lateral earth pressure =

= coefficient of lateral earth pressure at rest (Jaky ,

1948)
= angle of friction between soil and top surface of the block
= angle of friction between soil and bottom surface of the block
= angle of friction between soil and side surface of the block

The value or formulae indicated with the above parameters are used while calculating the
pullout capacity of anchor block using Eq. (3.8) or Eq. (3.10). It was reported that the
anchor block accelerates in the vertically upward direction on the verge of failure (Naser,
2006; Kame, et al., 2012). Consequently, friction force at the bottom of the block ( ) is
assumed to be zero (as ). On the other hand, = =20o is preferred to determine
and respectively (Bowles, 1997).

3.4 Comparison with the Experimental Results

The anchor and soil parameters of experimental studies conducted by Duncan and
Mokwa (2001), Naser (2006) and Mostofa (2013) are presented in Table 3.1. The
experimental results of these studies as shown in the second column of Table 3.2 have been
used to validate the proposed equations. In addition, pullout capacity were calculated using
other theoretical approaches, e.g. Ovesen and Stromann (1972), NAVFAC DM 7.02
(1986), BS 8006 (1995), Ghaly (1997), Bowles (1997) and Naser (2006) and the
percentage errors of these methods with respect to experimental results are also shown in
the same table. Rankine’s lateral earth pressure theory was preferred when required during

31
calculation process. For a better understanding of the relative predictive accuracy of the
proposed method, a cumulative frequency distribution of data corresponding to the
percentage error is presented in Table 3.3. Absolute values of the percentage errors were
used during frequency distribution, and their average values for each method, which is
statistically defined as Mean Absolute Percentage Error (MAPE), were calculated and
shown in Table 3.4 in ascending order.

Table 3.1 Anchor block and soil parameters

t Soil Exp.
Author 3
(mm) (mm) (mm) (mm) (deg.) (kN/m ) type type
Duncan and 1100 900 1100 1900 50.0 21.20 Gravel Field
Mokwa (2001)
300 150 150 150 43.5 17.40 Sand Lab.
Naser (2006) 300 150 150 150 43.5 19.60 Sand Lab.
1
300 150 150 150 43.5 10.90 Sand Lab.
475 75 150 150 37.2 14.74 Sand Lab.
Mostofa (2013) 475 75 150 150 43.9 15.62 Sand Lab.
475 75 150 150 44.8 17.51 Sand Lab.
1
Submerged unit weight of soil calculated as= (20.7-9.8) kN/m3= 10.90 kN/m3.

Although rigorous conclusion may not be drawn from limited number of test data (e.g. 7
test results used in this case), this comparison imparts some useful information regarding
the suitability of different methods at different conditions. For example, when compared to
experimental value reported by Duncan and Mokwa (2001), the methods proposed by
NAVFAC DM 7.02 (1986), BS 8006 (1995), Ghaly (1997) give errors more than 50%
(Table 3.2). Duncan and Mokwa (2001) measured anchor block bearing against compacted
gravel backfill near ground surface.

Naser (2006) performed one laboratory test on 100% saturated sandy soil to observe the
effect of degree of saturation on pullout resistance of anchor. The error with respect to this
test result for methods proposed by Ovesen and Stromann (1972), NAVFAC DM 7.02
(1986) and BS 8006 (1995) is also more than 50%. The remaining 5 test results were
conducted on poorly graded unsaturated sand. Each of the method discussed here, except
proposed method, gives error more than 50% at least twice (Table 3.3).

32
 Table 3.2 Comparison of theoretical predictions of pullout capacity with experimental results for vertical anchor block

Proposed Proposed Ovesen and

Exp. NAVFAC DM BS 8006 Ghaly Bowles Naser
Method Method Stromann
results 7.02 (1986) (1995) (1997) (1997) (2006)
Authors Eq. (3.8) Eq. (3.10) (1972)
Pu Pu % Pu % Pu % Pu % Pu % Pu % Pu % Pu %
(kN) (kN) error (kN) error (kN) error (kN) error (kN) error (kN) error (kN) error (kN) error
Duncan and Mokwa (2001) 410.0 412.0 0.5 396.0 -3.4 427.0 4.1 768.0 87.3 735.0 79.2 190.0 -53.6 233.0 -43.2 297.0 -27.6
1.3 1.3 0.0 1.2 -7.7 1.9 46.1 1.8 38.5 1.9 46.2 1.0 -23.1 0.6 -53.8 1.3 0.0
Naser (2006) 2.3 1.5 -34.8 1.4 -39.1 2.0 -13.0 2.0 -13.0 2.2 -4.3 1.1 -52.2 0.7 -69.6 1.5 -34.8
0.71 0.8 14.3 0.7 0.0 1.5 114.3 1.1 57.1 1.2 71.4 0.6 -14.3 0.4 -42.9 0.8 14.3
1.9 1.2 -36.8 1.1 -42.1 2.2 15.8 1.5 -21.1 2.2 15.8 2.1 10.5 0.6 -68.4 1.8 -5.3
Mostofa (2013) 2.1 1.9 -9.5 1.7 -10.5 3.1 47.6 3.0 42.9 3.1 47.6 1.8 -14.3 0.9 -57.1 3.6 71.4
2.3 2.2 -4.3 2.0 -13.0 3.5 52.2 3.8 65.2 3.6 56.5 1.9 -17.4 1.0 -56.5 4.4 91.3
1
Pullout resistance at 100% saturated condition

Table 3.3 Cumulative frequency distribution of errors for vertical anchor block
Proposed Proposed
Ovesen and NAVFAC DM BS 8006 Ghaly Bowles Naser
Method Method
Stromann (1972) 7.02 (1986) (1995) (1997) (1997) (2006)
Absolute Eq. (3.8) Eq. (3.10)
% error Cum. Cum. Cum. Cum. Cum. Cum. Cum. Cum.
Frequ- Frequ- Frequ- Frequ- Frequ- Frequ- Frequ- Frequ-
Frequ- Frequ- Frequ- Frequ- Frequ- Frequ- Frequ- Frequ-
ency ency ency ency ency ency ency ency
ency ency ency ency ency ency ency ency
0-10 4 4 3 3 1 1 0 0 1 1 0 0 0 0 2 2
10-20 1 5 2 5 2 3 1 1 1 2 4 4 0 0 1 3
20-30 0 5 0 5 0 3 1 2 0 2 1 5 0 0 1 4
30-40 2 7 1 6 0 3 1 3 0 2 0 5 0 0 1 5
40-50 0 7 1 7 2 5 1 4 2 4 0 5 2 2 0 5
>50 0 7 0 7 2 7 3 7 3 7 2 7 5 7 2 7

33
 Table 3.4 Mean absolute percentage error (MAPE) of different methods

Sl. No. Authors MAPE

1 Proposed Method Eq. (3.8) 14.3
2 Proposed Method Eq. (3.10) 16.5
3 Ghaly (1997) 26.5
4 Naser (2006) 35.0
5 Ovesen and Stromann (1972) 41.9
6 BS 8006 (1995) 45.9
7 NAVFAC (1986) 46.4
8 Bowles (1997) 55.9

Ovesen and Stromann (1972) overestimated the test results in most of cases. They
considered friction between wall and soil during upward movement of passive wedge,
which in turn contributes favourably in pullout capacity. However, anchor block moves
together with the passive wedge resulting no shear displacement between wall and passive
wedge (Duncan and Mokwa, 2001). Thus, the test results are smaller than their predictions
for anchor block. BS 8006 (1995) also overestimated test results in 6 out of 7 cases
significantly. BS 8006 (1995) assumes that pullout resistance of anchor is 4 times the
passive resistance of soil, while experimental studies by Mostofa (2013) indicated that this
coefficient is always less than 4.

In other circumstances, Bowles (1997) method underestimated the each test results
significantly. He used conventional earth pressure theories assuming implicitly that the
conditions at all cross sections along the length of a structure are the same, ignoring the
influence of the different conditions at the ends of the structure which contributes greatly
on pullout resistance of anchor.

Ghaly’s (1997) empirical equation underestimated test results twice by more than 50%
(Table 3.2) for relatively denser soil (Table 3.1). He used the results of 104 laboratory tests,
15 centrifugal tests, and 9 field tests to propose this empirical correlation, where unit
weight and internal friction angle of soil ranged from 14 to16 kN/m3 and 34o to 38.5o
respectively. It is expected that more deviation from the range of test parameters used to
derive the empirical correlation causes more errors. This is probably the reason for highest
error corresponding to the test results of Duncan and Mokwa (2001), where unit weight and
internal friction angle of soil were 21.2 kN/m3 and 50o respectively.

34
The proposed method, Eq. (3.8) or Eq. (3.10) gives absolute error in the range, 0 to 15% in
5 out of 7 cases. For remaining cases, it underestimates the actual results by 30% to 45%. It
indicates the better predictive capability of the proposed method as compared to the other
methods for varied parameters of soil and anchor. The mean absolute percentage error
(MAPE), as shown in Table 3.4, is lowest as compared to the other methods discussed here.

The prediction of pullout capacity by Eq. (3.10) is slightly smaller than Eq. (3.8). Thus, the
frictional resistance offered by thickness of the anchor block and side wedge is
insignificant compared to the passive resistance that derived from the weight of passive
wedge. Although the accuracy level of both proposed equation does not differ much,
Eq. (3.10) is preferable for its simplicity and offers quick estimation of pullout capacity of
anchor block using charts which will be presented in the later section of this chapter.

3.5 Break-Out Factors

It is always convenient to use graphical representation of a mathematical model that

consists of many variables or parameters. An effort has been made here to develop charts;
those may be used to calculate the pullout capacity of anchor block instead of using
Eq. (3.10) directly. For this purpose, a dimensionless parameter called break-out factor,
is introduced that may be defined as the ratio of and .

(3.25)

Let us consider a cube shaped anchor block (i.e. ), for which the weight
components of block and wedge can be expressed as:

(3.26)

 (3.27)

(3.28)

(3.29)

Summing up Eq. (3.26), Eq. (3.27), Eq. (3.28) and Eq. (3.29) to obtain W



(3.30)

35
Now, substituting the value of from Eq. (3.30) into Eq. (3.25) for :


(3.31)
Similarly, for , the break-out factor will be:


(3.32)

From Eq. (3.31) and Eq. (3.32), it is observed that is the function of and only,
and the graphical representation of these equations yield a series of straight lines as
depicted in Figure 3.5 and Figure 3.6 respectively. It is worth to mention that unit weight of
concrete was assumed as 1.5 times unit weight of soil in Eq. (3.28). However, such
assumption causes insignificant amount of error. For example, = 63.45 (for =5,
=45o, and =23.6 kN/m3) when ; while for similar
condition =65.86 for . Thus, the later value is greater by only 3.8% than
the previous value.

It is to be noted that while using Figure 3.5 and Figure 3.6, liner interpolation may be
performed for intermediate values of . Also, submerged unit of soil should be used when
the degree of saturation is 100% or, water table is located at ground surface.

Example 1 = 5 m, = = =1 m, =45o, =18 kN/m3

From Figure 3.5, =65 for =5 and =45o
Hence, = =(65)(18)(13)=1170 kN
Example 2 = 5 m, = =2 =1 m, =43o, =18 kN/m3, Water table at ground surface.
Interpolating for = 43o and =5, Figure 3.6 yields = 51
3 3
= = (18-9.8) kN/m =8.2 kN/m
Hence, = =(51)(8.2)(13)=418 kN

36
 150
f' = 55o
125

100
50o
N= Pu /( B3)

75
45o

50 40o
35o
30o
25 25o

0
1 2 3 H/B 4 5

Figure 3.5 Break-out factors for anchor block in cohesionless soil (for )

150

f' = 55o
125

100
N= Pu /( B3)

50o
75
45o
50
40o
35o
25 30o
25o

0
1 2 3 H/B 4 5

Figure 3.6 Break-out factors for anchor block in cohesionless soil (for )

37
3.6 Concluding Remarks

Proposed analytical method was first formulated considering all possible contributing
forces in the anchor block and wedge. Subsequently, it was simplified neglecting the
comparatively less influencing parameters intended for ease. Furthermore, simplified
method was facilitated with charts for quick estimation of the pullout resistance of anchor
block for most frequently used shapes, cube shaped and square shaped with half of
height/length of block as thickness. In most of the cases, the predictions of this method
show a very close agreement with experimental results. The prediction appears to be better
than other methods for varied condition of soil. The effect of ground water table can also be
taken into consideration while calculating the pullout resistance. Although the proposed
analysis uses equilibrium of passive wedge and ignores the shearing above the anchor
block, it is improved with a proper selection of angle and . Since limited field and
laboratory observations were available in the literature for fair comparison of the different
methods, sufficient field observations are required to develop confidence among the users
about the suitability of the proposed method along with other methods discussed here.

38
 4 CHAPTER 4
NUMERICAL ANALYSIS OF ANCHOR BLOCK

4.1 General

In the previous Chapter, proposed theoretical method was verified with limited number of
experimental studies. Since the cost of performing experimental tests on each and every
field problem combination is excessive, it is necessary to perform numerical analysis for
economy, time and safety. Moreover, the results obtained from experimental testing alone
are generally problem specific. This is particularly the case in geomechanics where we are
dealing with a highly nonlinear material which often displays pronounced scale effects. As
a result, it is often difficult to extend the findings from experimental data to full scale
problems with different material or geometric parameters. Therefore, an effort has been
made here to study the pullout resistance of anchor block numerically, which substantiates
the theoretical method proposed in the last chapter. This chapter mainly focuses on study of
the effects of aspect ratio, ground water table and embedment depth ratio on pullout
resistance of anchor block. However, it is essential to approximate the behaviour of soil as
realistic as possible. If the input parameters are sufficiently precise, then the output results
of the numerical analysis are usable.

Figure 4.1 Problem definition (a) Front view, (b) Side view of an anchor

4.2 Problem Definition

An anchor block of length , height and thickness installed in sand with an embedment
depth of as schematized in Figure 4.1 was simulated in this study. During the

39
installation, the soil in the vicinity of the anchor block can be disturbed. However, the
effects of disturbance on capacity were not considered in this study; instead, the
simulations were performed for a wished-in-place anchor block. The anchor block was
loaded at the centre of passive side as shown in Figure 4.1 (b). The effects of aspect ratio
was observed using five different sizes of anchor block, and the ground water level
is located eight different levels for a given anchor block to study the impact of ground
water table on pullout capacity. Additionally, the anchor block is placed at nine different
depths to investigate the variation of pullout capacity of anchor block with embedment
ratio . The properties of anchor block and soil are presented in the following
sections.

4.3 Finite Element Model

The 3D Finite Element (FE) program, PLAXIS 3D Foundation (Version 1.1), was used to
represent the soil-structure system. The system is symmetrical with respect to the z-axis.
The soil is assumed isotropic and homogenous with Mohr-Coulomb yield criterion and is
defined by the cohesion value , angle of internal friction , and dilatancy angle 

Figure 4.2 Geometry of Finite Element model

The selected dimensions of the model were large enough to prevent any restriction of
propagation of stresses and strains. Although analyses would take more time, the stress and
strain distribution in the soil because of different pullout loads in different positions can be
observed precisely. Considering the stress and strain propagation, dimensions of the model

40
to avoid any restrictions at the boundaries were computed through trial and error. Hence, a
homogeneous and isotropic soil domain with a width of 20 m, height of 20 m, and depth of
20 m was modelled (Figure 4.2).

Figure 4.3 Distribution of nodes and stress points in a 15-node wedge element

The wedge elements used in the 3D Foundation program consists of 15 nodes. The
distribution of nodes over the elements is shown in Figure 4.3. Adjacent elements are
connected through their common nodes. During a finite element calculation, displacements
( , and ) are calculated at the nodes. Nodes may be pre-selected for the generation
of load-displacement curves.

Figure 4.4 Distribution of elements around the anchor block

41
As mentioned earlier, all of the boundaries were considered far enough to prevent stress
reflection. In fact, the dimensions of the model have been selected in a way that the stress
and displacement gradients would decrease and become zero by getting close to the
boundaries. To obtain more accurate results, elements were kept very small near the anchor
block, increasing gradually in size and moving away from the block (Figure 4.4).

PLAXIS automatically imposes a set of general fixities to the boundaries of the geometry
model. These conditions are generated according to the following rules:

(i) Vertical model boundaries with their normal in x-direction (i.e. parallel to the y
plane) are fixed in x-direction ( = 0) and free in y- and z-direction, as shown in
Figure 4.2.

(ii) Vertical model boundaries with their normal in z-direction (i.e. parallel to the x-y-
plane) are fixed in z-direction ( = 0) and free in x- and y-direction.
(iii) Vertical model boundaries with their normal neither in x- nor in z-direction (skew
boundary lines in a work plane) are fixed in x- and z-direction ( = = 0) and
free in y-direction.
(iv) The model bottom boundary is fixed in all directions ( = = = 0).
(v) The 'ground surface' of the model is free in all directions.

Load-displacement curves can be used to visualise the relationship between the applied
loading and the resulting displacement of a certain point in the geometry. In PLAXIS
output-program, the x-axis relates to the displacement of a particular node, and the y-axis
contains data relating to load level. The latter is related with the value of ΣMstage in the
following way: Applied load = Total load applied in previous phase + ΣM stage times (Total
load applied in current phase minus Total load applied in previous phase). At the start of a
staged construction calculation, the multiplier that controls the staged construction process,
ΣMstage, is zero and this multiplier is stepwise increased to 1.0. When ΣMstage has reached
the value of 1.0, the current phase is finished. However, if a staged construction calculation
is not properly finished, i.e. the multiplier ΣMstage is less than 1.0 at the end of a staged
construction analysis. To obtain the load-displacemnt curve, first the ΣMstage - displacement
curve was generated in the output program. Then, the data from the curve was copied to a
spreadsheet program. The Applied load was calculated simply multiplying the ΣMstage with

42
the input-load in the loading phase, as the Total load applied in previous phase was zero in
this case. A sample calculation is provided in the APPENDIX, where the applied load and
the corresponding displacement was determined from Finite Element analysis of the
experimental study conducted by Naser (2006).

4.3.1 Modelling of anchor block

An anchor block of 0.4 m length, 0.4 m height and 0.4 m thickness was modelled first.
Analyses were also performed for different lengths to observe the effect of aspect ratio on
pullout capacity. By modelling the block as elastic-perfectly plastic material and also as
rigid body, it was found that the pullout capacity does not vary significantly with these
modelling techniques. However, the FE model with the block as a rigid body is
computationally very efficient. Therefore, the block was considered as a rigid body in the
FE analyses. The geometry and the properties of the anchor block used in the study are
presented in the Table 4.1. This geometry along with the soil parameters shown in Table
4.1 is referred as “base case” in the following sections.

Table 4.1 Geometry and mechanical properties in FE modelling

Anchor Block Length ( ) 0.4 m

Height ( ) 0.4 m
Thickness ( ) 0.4 m
Modulus of elasticity ( ) 2 x 107 kN/m2
Poisson’s ratio ( ) 0.2
Unit weight of anchor block( ) 23.6 kN/m3
Cohesionless Soil Angle of internal friction ( ) 35o
Angle of dilation () 5o
Young’s modulus ( ) 60,000 kN/m2
Poisson’s ratio ( ) 0.3
Cohesion ( )1 0.5 kN/m2
Unit weight of soil ( ) 15.73 kN/m3
Interface reduction factor ( ) 0.5
1
Small cohesion is required to be defined in PLAXIS FE analysis. For sand, in this study a very small
value of c′= 0.5 kN/m2 was used.

43
4.3.2 Modelling of cohesionless soil

The cohesionless soil was modelled using the built-in Mohr-Coulomb model available in
the PLAXIS FE software. While loading, cohesionless soil was treated as drained
homogeneous material. The unit weight and the angle of internal friction of soil were
assumed as 15.73 kN/m3 and 35o respectively for most of the cases. The properties of the
soil used in the FE analysis are shown in Table 4.1.

4.3.3 Interface behaviour

In PLAXIS, the soil/anchor interaction is modelled by choosing a suitable value for the
strength reduction factor ( ) in the interface, which is defined as =
. Where, and are the friction angle for interface and soil respectively. This factor
relates the interface strength (anchor friction) to the soil strength (friction angle and
cohesion). The value varies between 0 and 1 depending upon surface roughness,
mean particle size of sand and method of installation (CFEM, 2006; Tiwari, et al., 2010). In
general, for real soil-structure interaction the interface is weaker and more flexible than the
associated soil layer, which means that the value of should be less than 1. Suitable
values for for the case of the interaction between various types of soil and structures
in the soil can be found in the literature. In this study, friction angle for interface ( ) was
assumed as 20o (Das, 1995; Bowles, 1997). Thus, the approximate value of interface
reduction factor ( ) was calculated as 0.5 for =35o.

4.3.4 Modulus of elasticity of sand

The Young’s modulus of sand, , can be expressed as a function of mean effective stress,
, as, = (Janbu, 1963; Hardin and Black, 1966); where, and are
two material parameters, is the atmospheric pressure = (100 kPa). The displacement
of an anchor block due to pullout force may be obtained from FE model. This displacement
depends considerably on the value of modulus of elasticity of soil, . Since the study of
such displacement is out of scope of the present research work, no attempt has been taken
to vary with , rather a constant value of = 60 MPa was used.

44
4.3.5 Dilatancy angle of cohesionless soil

The angle of dilation ( ) controls an amount of plastic volumetric strain developed during
plastic shearing and is assumed constant during plastic yielding. The value of
corresponds to the volume preserving deformation while in shear. For sands, the angle of
dilation depends on the angle of internal friction. In this study, the value of dilation angle
was estimated using PLAXIS 3D reference manual (version 1) which recommends to
determine the value of dilation angle as 30o for non-cohesive soils (sand, gravel)
with the angle of internal friction, 30o.

50
H/B=3

40
Pullout Force (kN)

30

20

Coarse Mesh (3204 elements)

10
Medium dense mesh (7632 elements)
Fine mesh (18272 elements)
0
0 5 10 15 20 25
Total Displacement (mm)

Figure 4.5 Mesh sensitivity analysis

4.4 Results and Discussions

4.4.1 Mesh sensitivity analysis

In general, smaller FE mesh yields more accurate results but computationally expensive.
For efficient modelling, small elements are used near the block. The size of the elements is
increased with increase in distance from the block as shown in Figure 4.4. Similarly, the
element size is increased with distance from the bottom of the block. To select the optimum
mesh, several trial analyses are conducted with different mesh sizes. The force-
displacement curves for three different sizes of mesh are shown Figure 4.5 for the base case

45
(Table 4.1) with the embedment depth ratio, =3. As shown in Figure 4.5, the
calculated pullout force is smaller with fine mesh than that of with coarse mesh. In this
study, the fine mesh was selected to perform the analyses as it is more accurate, although it
is recognized that it is computationally expensive.

Table 4.2 Anchor block and soil parameters to investigate the effect of aspect ratio

  N
3
(m) (m) (m) (m) (deg.) (kN/m ) (kN) =Pu/( BLt)
1.2 0.4 0.4 0.4 35 15.73 28 1 27.8
1.2 0.4 0.4 0.8 35 15.73 50 2 24.8
1.2 0.4 0.4 1.2 35 15.73 70 3 23.2
1.2 0.4 0.4 2.0 35 15.73 112 5 22.3
1.2 0.4 0.4 3.6 35 15.73 200 9 22.1

4.4.2 Effect of aspect ratio

While the soil properties influence the pullout capacity significantly, the size and shape of
the anchor has also considerable influence on ultimate pullout resistance of an anchor block
(Das, 1990; Rowe and Davis, 1982). The length ( ) of the anchor block is gradually
increased from the base case to observe the effect of anchor shape on ultimate capacity.
However, the other geometries of the anchor block and soil parameters remain same as
base case, which is presented in Table 4.2.

Figure 4.6 Load-displacement curve for the study of the effect of aspect ratio ( )
46
The variation of pullout force with total displacement for embedment depth ratio =3,
along the direction of pulling is shown Figure 4.6. In this study, the load displacement
curves do not show any peak; rather the pullout force is increasing very gradually after
plastic yielding. Therefore, ultimate capacity is defined by adopting failure points, as
shown in Figure 4.6, after which the load-displacement diagrams become practically linear
(Neely, et al., 1973; Das, 1990). This method was also utilized by Dickin and King (1997)
and Ghaly (1997) and they found more consistent results with experimental studies than
other alternative methods.

The impact of aspect ratio ( ) on ultimate capacity for an anchor block with = 3, is
presented in Figure 4.7. Here, the pullout capacity is expressed as a dimensionless
parameter called break-out factors ( ) similar to the last chapter, which is defined as the
ratio of and . Figure 4.7 indicates that decreasing aspect ratio results increasing
break-out factor (relative to =9) of 26%, 12%, 5% and 1% for ratios of 1, 2, 3 and
5 respectively. Hence, the effect of shape is considerable for ≤ 2 and is of insignificant
importance for ≥ 5. This is attributed to fact that the boundary conditions at the ends
of a square structure are quite different from those at the center, which contributes greatly
on pullout resistance. As the aspect ratio increases, the influence of boundary conditions
decreases, and plain strain results (2D analysis) will be more applicable for large aspect
ratios. Figure 4.7 also indicates that a rectangular anchor block with ≥ 5 should be
treated as strip anchor block.

30
H/B = 3

28
N=Pu /( BLt)

26

24

22

20
1 2 3 4 5 6 7 8 9
Aspect Ratio, L/B

Figure 4.7 Effect of aspect ratio (L/B) on pullout capacity of anchor

47
4.4.3 Effect of ground water table

The effect of ground water table can be considered in the analytical method, Eq. (3.8) and
Eq. (3.10), derived in the last chapter, when ground water table is located at ground
surface. The use of submerged unit weight ( ) instead of bulk unit weight of soil ( ) in the
analytical expression is sufficient to account the effect of water table. However, ground
water table may be located at any level in the soil layer. Consequently, an effort has been
made here to account the influence of ground water table located at any position.

The analysis was performed for the base case, where the anchor block is placed at 1.2 m
depth, i.e. =3. The depth of ground water table is denoted by , as shown in Figure 4.8.
All the analyses were conducted for the locations of ground water table at = 0, 0.17,
0.33, 0.5, 0.67, 0.83 and 1.00 as presented in Table 4.3. In one circumstances, the water
table is kept well below (20 m depth) so that water table has practically no influence on
pullout capacity of anchor block, which is referred as dry condition in this section. The
pullout capacity for different locations of ground water table is determined from load-
displacement curves as depicted in Figure 4.9 and is presented in the Table 4.3.

Figure 4.8 Effect of water table and correction factors.

Table 4.3 Correction factors for different locations of ground water table
Pu,GWT Pu,dry Pu,GWT/ Pu,dry
(mm) (mm) (kN) (kN)
0.0 1.2 0.00 14.0 28.0 0.50
0.2 1.2 0.17 18.0 28.0 0.64
0.4 1.2 0.33 22.5 28.0 0.80
0.6 1.2 0.50 25.5 28.0 0.91
0.8 1.2 0.67 27.0 28.0 0.96
1.0 1.2 0.83 27.5 28.0 0.98
1.2 1.2 1.00 28.0 28.0 1.00
Dry cond. 1.2 28.0

48
 35 Dry condition
H/B=3 1.00
0.83
30 0.67
0.50

Pullout Force (kN) 25 0.33

0.17
20
z/H=0
15

10
Failure Points
5

0
0 5 10 15 20 25
Total Displacement (mm)

Figure 4.9 Load-displacement curves for different locations of ground water table

The pullout resistance of anchor block, when ground water table is located at the base of
the block is same as dry condition in this case (Table 4.3). Thus, the ground water table
may not have essentially any influence on pullout resistance of anchor block for the case
≥ . However, the effect of ground water table should be considered, when the water
table is anywhere between the ground surface and the base of the anchor block, that is,
0≤ ≤ . The correction factor for water table, defined as the ratio of pullout resistance
with ground water table between 0 ≤ ≤ to that of dry condition, is determined and
presented in Table 4.3. Figure 4.10 shows the variation of correction factor with . In
this case, the correction factor was found as 0.5 when water table is at ground surface i.e.
=0 and the factor is 1 when water table is at the base of anchor block, i.e. =1. The
variation of correction factor from ground surface to the base of anchor block is not linear,
rather concave downward as shown in Figure 4.10. Therefore, an empirical equation is
proposed here, which may be used roughly to determine the correction factor for water
table. The correction factor should be multiplied with the pullout resistance at dry condition
to account the ground water table effect. The considerations on ground water table can be
summarized as follow:

Case I: No correction is needed for ≥ .

Case II: Correction should be applied for 0 ≤ ≤ . Correction factor may be calculated
from Eq. (4.1).
49
 (4.1)
Where, is correction factor for water table and is the depth ratio at
ground water level.

1.00

0.75
Pu,GWT/Pu,dry

Pu,GWT/Pu,dry= -0.6 (z/H)2 + 1.1(z/H) + 0.5

0.50

0.25
0.0 0.2 0.4 0.6 0.8 1.0
z/H
Figure 4.10 Variation of correction factors with the location of ground water table

4.4.4 Effect of embedment depth ratio

In the last chapter, we have seen that the analytical expression, Eq. (3.10) has been
facilitated with charts (Figure 3.5 and Figure 3.6) for most frequently used shapes, cube
shaped and square shaped with half of height/length of block as thickness. It is observed
from Figure 3.5 and Figure 3.6 that the theoretical break out factor, varies linearly with
the embedment depth ratio ( ) for a given value of angle of internal friction of soil ( ).
However, Eq. (3.10) was verified with several test results from literature, which were
mostly conducted at shallow depth ( < 4). The failure mechanism of shallow anchors
and deep anchors are not similar. For shallow anchors, the embedment ratio is such that,
failure surface reaches the ground surface at limit equilibrium; whereas for deep anchors,
the embedment ratio is such that, failure surface does not reach the ground surface at limit
equilibrium (Das, 1990). Thus, the suitablity of Eq. (3.10) for greater embedment ratio
( ≥ 4) should be confirmed.

Accordingly, FE analysis with PLAXIS was conducted for the base case, where the anchor
block is placed at embedment depth ratio ( ) of 1 to 9. The pullout capacity for different
embedment ratio is determined from load-displacement curves as depicted in Figure 4.11
50
and is presented in the Table 4.4. The break-out factors ( ) were calculated from pullout
capacity of anchor block (Table 4.4) and compared with the proposed method, Eq. (3.10)
along with other methods suggested by different authors as shown in Figure 4.12.

The FE analyses show that the relationship between the break-out factor and the
embedment depth ratio is not linear as depicted in Figure 4.12, instead increases faster at
greater embedment depth ratio. The proposed theoretical method fits better with the FE
analysis results than other methods suggested by Naser (2006), Ghaly (1997),
BS 8006 (1995) and Ovesen and Stromann (1972) for ≤ 5. In section 4.6, it will be
shown that, the Mohr-coulomb model somewhat overestimates the actual result. Inspite of
this fact, the smaller prediction of the proposed method than the FE analyses is significant
for greater embedment ratio ( > 5). For exmple, it predicts nearly half as FE analysis
result for =9. On the other hand, the methods proposed by BS 8006 (1995) and Naser
(2006) show good agreement with the FE anlysis for greater embedment ratio. Thus, it is
recommended that the proposed theoretical method, Eq. (3.10) should be used for shallow
depth i.e. ≤ 5.

Figure 4.11 Load-displacement curves for different embedment depth ratio

51
 Table 4.4 Pullout resistance of anchor block for different embedment ratio.
B   =
3
(m) (m) (m) (m) (deg.) (kN/m ) kN Pu/( 
0.4 0.4 0.4 0.4 35 15.73 8 1 7.9
0.8 0.4 0.4 0.4 35 15.73 17 2 16.9
1.2 0.4 0.4 0.4 35 15.73 26 3 25.8
1.6 0.4 0.4 0.4 35 15.73 35 4 34.8
2 0.4 0.4 0.4 35 15.73 50 5 49.7
2.4 0.4 0.4 0.4 35 15.73 65 6 64.6
2.8 0.4 0.4 0.4 35 15.73 85 7 84.4
3.2 0.4 0.4 0.4 35 15.73 110 8 109.3
3.6 0.4 0.4 0.4 35 15.73 140 9 139.1

600
Mohr Coulomb-PLAXIS
Proposed Method: Eq. (3.10)
500 Naser (2006)
Ghaly (1997)
BS 8006 (1995)
400
Ovesen and Stromann (1972)
N=Pu /( B3)

300

200

100

0
1 2 3 4 5 6 7 8 9 10
H/B

Figure 4.12 Comparison of break-out factors obtained from FE analyses with the
proposed method along with other methods suggested by different authors for =35o

4.5 Failure Modes of Soil

The complex nature of vertical anchor behaviour, due to the lack of symmetry, has led to a
great deal of uncertainty regarding the likely failure patterns at collapse. As a result, it is
difficult to predict the capacity of vertical anchors using existing approaches that require
52
assumptions regarding the shape of the failure surface (Merifield and Sloan, 2006). Such
approaches include the limit equilibrium method, the method of characteristics, and any
analytical upper bound method.

(a) (b)

Figure 4.13 Failure modes and zones of plastic yielding for anchor block in
cohesionless soils as predicted by PLAXIS for =3 and =35o (a) vertical section
and (b) at ground surface

Figure 4.14 Displacement fields during collapse at mid-level of anchor block for
=3 and =35o

A distinct advantage of the numerical analysis used is that the form of the failure
mechanism can be obtained automatically without any assumptions being made in advance.
The observed mode of failure and velocity diagram at collapse of an anchor block is shown
in Figure 4.13 and Figure 4.14 respectively. In general, it was found that failure could be
characterized by the development of a small active failure zone immediately behind the
anchor, and an extensive passive failure zone in front of the anchor as depicted in
53
Figure 4.13 (a) and (b). It was also found that upward movement of soil wedge is
accompanied by lateral deformation extending out from the anchor edge (Figure 4.13 (b)
and Figure 4.14). As the anchor is pulled horizontally, the material in front of the anchor
tends to lock up as it attempts to dilate during deformation. As a consequence, to
accommodate the rigid soil wedge movement, the observed plastic zone is forced to extend
a large distance laterally outwards into the soil mass. This soil locking phenomenon was
noticed by (Rowe, 1978) who found that soil dilatancy had a significant effect on the
observed plastic region at failure, which in turn resulted in substantial variations in the
predicted anchor break-out factor. The condition of soil above the anchor block is very
uncertain. Two small non-plastic zones were observed to develop above the anchor block
as shown in Figure 4.13 (a).

Figure 4.15 Load-displacement curves as predicted by PLAXIS for the experimental

studies by Duncan and Mokwa (2001)

4.6 Remarks on Mohr-Coulomb Model

The experimental studies conducted by Duncan and Mokwa (2001), Naser (2006) and
Mostofa (2013) were simulated using Mohr-Coulomb (MC) model in PLAXIS. The pullout
capacity of anchor block is determined from load-displacement curve as shown in Figure
4.15, Figure 4.16 and Figure 4.17, and presented in Table 4.5 along with anchor and soil
parameters used for experimental studies. From Table 4.5 it is seen that, the predicted
results from MC model overestimates the test results by 16 to 47 %. This is probably due to
the fact that MC model is an elastic perfectly-plastic model. It cannot simulate non‐linear
and inelastic behaviour of soil properly. In this case, the failure points for pullout capacity

54
were obtained at the end of linear portion of load-displacement curves, whereas the
experimental studies by Neely et al. (1973) showed that the same state of soil is achieved at
lower stress level. Hereby, the soil behaves stiffer than in reality and the numerical results
obtained from the MC model are usually higher than the experimental results. The PLAXIS
3D results of Nimityongskul (2010), obtained with the soil model MC model, also
overestimated the stiffness of the load displacement curve, while he studied the effects of
soil slope on lateral capacity of piles in cohesive soils.

Figure 4.16 Load-displacement curves as predicted by PLAXIS for the experimental

studies by Naser (2006)

Figure 4.17 Load-displacement curves as predicted by PLAXIS for the experimental

studies by Mostofa (2013)

55
 Table 4.5 Anchor and soil parameters for experimental studies

Pullout Capacity (kN)

Author
(mm) (mm) (mm) (deg.) (kN/m3) Exp. FE- %
(mm)
MC error
Duncan and
1100 900 1100 1900 50.0 21.20 410 602 47
Mokwa (2001)
300 150 150 150 43.5 17.40 1.3 1.7 31
Naser (2006) 300 150 150 150 43.5 19.60 2.3 2.9 26
300 150 150 150 43.5 20.71 0.7 0.9 29
475 75 150 150 37.2 14.74 1.9 2.2 16
Mostofa (2013) 475 75 150 150 43.9 15.62 2.1 2.6 24
475 75 150 150 44.8 17.51 2.3 3.0 30
1
Saturated unit weight

In spite of this limitation, MC model was adopted in this study, since it is computationally
very efficient and requires only five basic input parameters (Young's modulus, ; Poisson's
ratio, ; cohesion, ; friction angle, ; and dilatancy angle, ). Moreover, all the useful
findings of this numerical study were made on the shape of the curves, which were plotted
using non-dimensional parameters (Figure 4.7, Figure 4.10 and Figure 4.12). Thus, the
qualitative trends of the curves were more important than the quantitative value of pullout
capacity for this study.

56
 5 CHAPTER 5
CONCLUSIONS AND RECOMMENDATIONS

5.1 General

This chapter summarizes the major findings through this study. The main objective of this
study was theoretical and numerical analyses of pullout capacity of anchor block embedded
in cohesionless soil. The effect of ground water table on anchor capacity was also studied
and can be considered during design. The numerical analyses were conducted using Mohr-
Coulomb constitutive model in PLAXIS. Of course, this research work could not capture
all the field problems regarding anchor block design. However, it is expected that this
research will make an important contribution to the design of anchored retaining earth wall
in cohesionless soil used as backfill.

5.2 Conclusions

The following conclusions can be drawn from the results presented in this research work:

(i) The proposed theoretical method, Eq. (3.8) and Eq. (3.10) as presented below, can be
used to determine the pullout capacity of an anchor block embedded in cohesionless
soil. Equation (3.8) was first formulated considering all possible contributing forces
in the anchor block and wedge. Subsequently, it was simplified as Eq. (3.10)
neglecting the comparatively less influencing parameters- forces acting at the side of
the wedge, frictional resistance between concrete and soil and active thrust on block
intended for ease. The effect of ground water table can be considered only when the
water table is located at ground surface. The use of submerged unit weight ( )
instead of bulk unit weight of soil ( ) in the analytical expression is sufficient to
account the effect of water table.

  (3.8)

(3.10)

57
(ii) Equation (3.10) is facilitated with charts for quick estimation of the pullout
resistance of anchor block for most frequently used shapes, cube shaped and square
shaped with half of height/length of block as thickness. It is observed from those
charts that the theoretical break out factor, varies linearly with the embedment
depth ratio ( ) for a given value of angle of internal friction of soil ( ).

(iii) The results obtained from the experimental studies in the literature were compared
with the predictions from the proposed method and existing methods. The
comparisons between the proposed method and experimental results appear to be
better than other methods for varied condition of soil. The Mean Absolute
Percentage Error (MAPE) was found to be smaller, 14.3% and 16.5% for Eq. (3.8)
and Eq. (3.10) respectively, than other methods considered in this study.

(iv) The proposed theoretical method shows very close agreement with FE analysis for
embedment depth ratio, ≤ 5. However, it underestimates the break-out factors
significantly for greater embedment ratio ( > 5) and, predicts nearly half as FE
analysis result for =9. Thus, it is recommended that the proposed theoritical
method should be used for shallow depth only i.e. should be less than or equal
to 5.

(v) From FE analysis, it was found that the ground water table may not have essentially
any influence on pullout resistance of anchor block when it is located at or below the
base of the block. However, the effect of ground water table should be considered,
when the water table is anywhere between the ground surface and the base of the
anchor block. In such cases, a correction factor may be calculated from Eq. (4.1) as
shown below, and should be multiplied with the pullout capacity determined from
the proposed theoretical method.

(4.1)

(vi) The size and shape of the anchor has also considerable influence on ultimate pullout
resistance of an anchor block. FE analysis shows that the effect of shape is
considerable for aspect ratio, ≤ 2 and is of insignificant importance for ≥ 5.
As the aspect ratio increases, the influence of boundary conditions decreases, and
58
 plain strain results (2D analysis) will be more applicable for large aspect ratios.
Hence, the rectangular anchor block with ≥ 5 should be treated as strip anchor
block.

(vii) Mohr-Coulomb model was found to overestimate the experimental results

considerably. Despite of this limitation, Mohr-Coulomb model was adopted in this
study, since all the findings were made on the shape of the curves, which were
plotted using non-dimensional parameters. Thus, the qualitative trends of the curves
were more important than the quantitative value of pullout capacity for this study.

(viii) From FE analysis, it was found that failure could be characterized by the
development of a small active failure zone immediately behind the anchor, and an
extensive passive failure zone in front of the anchor. It was also found that upward
movement of soil wedge is accompanied by lateral deformation extending out from
the anchor edge. However, the condition of soil above the anchor block is very
uncertain. Two small non-plastic zones were observed to develop above the anchor
block.

5.3 Recommendations for Future Study

During the course of theoretical and numerical analyses, there was always an urge to
expand the scope of the study in order to gather some more information and to achieve
some better results. Moreover, as this is frequently adopted method during design and
construction of retaining wall, opportunities for future researches are numerous. Some of
these future research prospects are recommended below-

(i) Although the cohesionless soil is always preferred as a backfill material in general,
sometimes cohesive soil may be encountered during backfilling. In such
circumstances, design of anchor block should be made on cohesive soil, which may
be studied.

(ii) Very high retaining wall may require multi-staged anchor system. The anchor in the
lowest stage may be embedded at deeper depth. In that case, new theoretical
approach may be developed to study the behaviour of deep anchor.

(iii) Sometimes, the displacement corresponding to the ultimate capacity is required for
59
 design purposes. This can be obtained from the load-displacement curves. However,
the load-displacement curves obtained from Mohr-Coulomb model in this study are
not reliable, since the soil behaves stiffer than in reality. Therefore, rigorous field
tests may be conducted to obtain actual load-displacement curves, or advanced
constitutive model like Hardening-Soil (HS) model may be adopted which is
formulated in the framework of hardening plasticity.

(iv) The ultimate capacity is achieved when anchor block is kept far away from the active
failure surface behind the retaining wall. On the other hand, sometimes space
constraints allow the anchor block to place near failure surfaces. In such case, partial
passive resistance develops in front of the anchor block that may be studied.

(v) The effect of dynamic loading (e.g. earthquake or, liquefaction) on pullout capacity of
anchor block may be investigated.

Thus, it is recommended for future study to work on these areas to capture all the problems
encountered in the field during the design and construction of anchor block. Finally, it is
expected that the present study will be useful to all those dealing with civil engineering
projects and research works on anchored retaining wall. This research will also be useful to
those who are involved in the development of standards on the determination of horizontal
pull out capacity of anchor block.

60
 REFERENCES

Akinmusuru, J. O. (1978). Horizontally Loaded Vertical Plate Anchors in Sand. Journal of the
Geotechnical Engineering Division, ASCE, 104(2), 283-286.
Basudhar, P. K., and Singh, D. N. (1994). A Generalized Procedure for Predicting Optimal
Lower Bound Break-out Factors of Strip Anchors. Géotechnique, 44(2), 307-318.
Bowles, J. E. (1997). Sheet-pile Walls Cantilevered and Anchored. In Foundation Analysis and
Design (5th ed., pp. 778-780). The McGraw-Hill Companies, Inc., Singapore.
BS 8006. (1995). Strengthened/Reinforced Soils and Other Fills, Section 6.6. In British
Standard (p. 66). London.
CFEM. (2006). Canadian Foundation Engineering Manual (4th ed.). Richmond, BC, Canada
Canadian Geotechnical Society.
Daniel, D. E., and Olson, R. E. (1982). Failure of an Anchored Bulk-Head. Journal of
Geotechnical Engineering Division, ASCE, 108(10), 1318-1327.
Das, B. M. (1990). Vertical Plate Anchors. In Earth Anchors (pp. 84-134). Amsterdam,
Netherlands Elsevier.
Das, B. M. (1995). Principles of Foundation Engineering (3rd ed.). California Books/Cole
Publishing Company.
Das, B. M. (2007). Sheet Pile Walls. In Principles of Foundation Engineering (6th ed., pp. 452-
459). Thomson Brooks/Cole, Ontario, Canada.
Das, B. M., and Seeley, G. R. (1975). Load-Displacement Relationships for Vertical Anchor
Plates. Journal of the Geotechnical Engineering Division, American Society of Civil
Engineers, 101(GT7), 711-715.
Dickin, E. A., and King, J. W. (1997). Numerical Modelling of the Load-Displacement
Behaviour of Anchor Walls. Computers and Structures, 63(4), 849-858.
Dickin, E. A., and Leung, C. F. (1983). Centrifuge Model Tests on Vertical Anchor Plates.
Journal of Geotechnical Engineering, ASCE, 109(12), 1503-1525.
Dickin, E. A., and Leung, C. F. (1985). Evaluation of Design Methods for Vertical Anchor
Plates. Journal of Geotechnical Engineering, ASCE, 111(4), 500-520.
Duncan, M., and Mokwa, R. (2001). Passive Earth Pressures Theories and Test. Journal of
Geotechnical and Geoenvironmental Engineering, ASCE, 127(4), 248-257.
Gabr, M. A., and Borden, R. H. (1990). Lateral Analysis of Piers Constructed on Slopes. Journal
61
 of Geotechnical Engieering, 116(12), 1831-1850.
Ghaly, A. M. (1997). Load-Displacement Prediction for Horizontally Loaded Vertical Plates.
Journal of Geotechnical and Geoenvironmental Engineering, American Society of Civil
Engineers, 10.1061/(ASCE)1090-0241(1997)1231(74).
Gireesha , T., and Muthukkumaran, K. (2011). Study on Soil-Structure Interface Strength
Property. International Journal of Earth Sciences and Engineering, 4(6), 89-93.
Hanna, A. M., Das, B. M., and Foriero, A. (1988). Behaviour of Shallow Inclined Plate Anchors
in Sand. (B. M. Das, Ed.) Geotechnical Special Publication 16, ASCE, 54-72.
Hansen, J. B. (1966). Resistance of Rectangular Anchor Slab. Danish Geotechnical Institute, 21,
12-13.
Hardin, B. O., and Black, W. L. (1966). Sand Stiffness Under Various Triaxial Stresses. Journal
of the Soil Mechanics and Foundations Division, 92(2), 27-42.
Hoshiya, M., and Mandal, J. N. (1984). Some Studies of Anchor Plates in Sand. Soils and
Foundations, 24(1), 9-16.
Hueckel, S. (1957). Model Tests on Anchoring Capacity of Vertical and Inclined Plates.
Proceeding,4th Internation Conference on Soil Mechanics and Foundation Engineering,
2, pp. 203-206. London, England Butterworths Scientific Publications.
Hueckel, S., Kwasniewski, J., and Baran, L. (1965). Distribution of Passive Earth Pressure on
the Surface of a Square Vertical Plate Embedded in Soil. Proceeding, 6th International
Conference on Soil Mechanics and Foundation Engineering, 2, pp. 381-385. Montreal,
Canada.
Jaky , J. (1948). Pressure in silos. 2nd ICSMFE, 1, pp. 103-107. London.
Janbu, N. (1963). Soil Compressibility as Determined by Oedometer and Triaxial Test., 1, pp.
19-25. Wiesbaden, Germany.
Jesmani, M., Kamalzare, M., and Nazari, M. (2013). Numerical Study of Behavior of Anchor
Plates in Clayey Soils. International Journal of Geomechanics, 13(5), 502-513.
Jones, C. (1996). Earth Reinforcement and Soil Structures. Thomas Telford Services Ltd,
London, UK.
Kame, G. S., Dewaikar, D. M., and Choudhury , D. (2012). Pullout Capacity of a Vertical Plate
Anchor Embedded in Cohesionless Soil. Earth Science Research, 1(1), 27-56.
Khan, A. J., and Sikder, M. (2004). Design Basis and Economic Aspects of Different Types of
Retaining Walls. Journal of Civil Engineering, IEB, CE 32(1), 17-34.
Kötter, F. (1903). Die Bestimmung des Drucks an gekrümmten Gleitflächen, eine Aufgabe aus
62
 der Lehre vom. 229-233. Berlin Univ.-Bibliothek Frankfurt am Main.
Koutsabeloulis, N. C., and Griffiths, D. V. (1989). Numerical Modelling of the Trap Door
Problem. Géotechnique, 39(1), 77-89.
Kumar, J., and Sahoo, J. P. (2012). Upper Bound Solution for Pullout Capacity of Vertical
Anchors in Sand Using Finite Elements and Limit Analysis. International Journal of
Geomechanics, 12(3), 333-337.
Lade, P. V., and Duncan, J. M. (1975). Elasto-Plastic Stress-Strain Theory for Cohesionless Soil.
Journal of Geotechnical Engineering Division, ASCE.
Lin, C., Han, J., Bennett, C., and Parsons, R. L. (2014). Analysis of Laterally Loaded Piles in
Sand Considering Scour Hole Dimensions. Journal of Geotechnical and
Geoenvironmental Engineering, ASCE, 140(6), 1-13.
Merifield, R. S., and Sloan, S. W. (2006). The Ultimate Pullout Capacity of Anchors in
Frictional Soils. Canadian Geotechnical Journal, 43, 852-868.
Mirzoyan. (2007). Lateral Resistance of Piles at the Crest of Slopes in Sand. Master’s Thesis,
Brigham Young University, Department of Civil and Environmental Engineering, Provo,
Utah, United States.
Mostofa, M. G. (2013). Horizontal Pullout Resistance of Concrete Anchor Blocks in Sand
Backfill. Master’s Thesis, Bangldesh University of Engineering and Technology,
Department of Civil Engineering, Dhaka, Bangladesh.
Murray, E. J., and Geddes, J. D. (1987). Uplift of Anchor Plates in Sand. Journal of
Geotechnical Engineering, ASCE, 113(3), 202-215.
Murray, E. J., and Geddes, J. D. (1989). Resistance of Passive Inclined Anchors in Cohesionless
Medium. Géotechnique, 39(3), 417-431.
Naser, A. S. (2006). Pullout Capacity of Block Anchor in Unsaturated Sand. Unsaturated Soils
2006 (GSP 147), Proceedings of the Fourth International Conference on Unsaturated
Soils. (pp. 403-414). Arizona ASCE.
NAVFAC DM 7.02. (1986). Foundations and Earth Structures . In Naval Facilities Engineering
Command (pp. 91-92). Alexandria, Virginia.
Neely, W. J., Stuart, J. G., and Graham, J. (1973). Failure Loads of Vertical Anchor Plates in
Sand. Journal of the Geotechnical Engineering Division, ASCE, 99(9), 669-685.
Nimityongskul, N. (2010). Effects of Soil Slope on Lateral Capacity of Piles in Cohesive Soils.
PhD Thesis, Department of Civil and Construction Engineering, Oregon State
University, Oregon, USA.
63
Niroumand, H., and Kassim, A. (2010). Analytical and Numerical Study of Horizontal Anchor
Plates in Cohesionless Soils. Electronic Journal of Geotechnical Engineering, 15(C),
281-292.
Ovesen, N. K. (1964). Passive Anchor Slabs, Calculation Methods and Model. Danish
Geotechnical Institute, 4, 5-39.
Ovesen, N. K. (1981). Centrifuge Tests of the Uplift Capacity of Anchors. Proceedings of the
10th International Conference on Soil Mechanics and Foundation Engineering, 1, pp.
717-722. Stockholm, A.A. Balkema, Rotterdam, The Netherlands, 15–18 June 1981.
Ovesen, N. K., and Stromann, H. (1972). Design Methods of Vertical Anchor Slabs in Sand.
Proceedings, Specialty Conference on Performance of Earth and Earth-Supported
Structures, American Society of Civil Engineers, 2.1, pp. 1481-1500.
Potyondy, J. G. (1961). Skin Friction between Various Soils and Construction Materials.
Géotechnique, 11(1), 339-353.
Reese, L. C., Cox, W. R., and Koop, F. D. (1974). Analysis of Laterally Loaded Piles in Sand.
Sixth Annual Offshore Technology Conference. 2, pp. 473-480. Houston, Texas OTC
2080.
Rowe, R. K. (1978). Soil Structure Interaction Analysis and Its application to the Prediction of
Anchor Behaviour, PhD Thesis, University of Sydney, Department of Civil Engineering,
Sydney, Australia
Rowe, R. K., and Davis, H. (1982). The Behaviour of Anchor Plates in Sand. Géotechnique,
32(1), 25-41.
Sakai, T., and Tanaka, T. (1998). Scale Effect of a Shallow Circular Anchor in Dense Sand. Soils
and Foundations, 38(2), 93-99.
Samui, P., and Sitharam, T. G. (2009). Pullout Capacity of Small Ground Anchors A Relevance
Vector Machine Approach. Geomechanics and Engineering, 1(3), 259-262.
Singh, A. I. (1967). Soil Engineering in Theory and Practice. New York Asia Publishing House,
Inc.
Sloan, S. W. (1988). Lower Bound Limit Analysis Using Finite Elements and Linear
Programming. International Journal for Numerical and Analytical Methods in
Geomechanics, 12, 61-67.
Smith, J. E. (1962). Deadman Anchorages in Sand. Washington, D.C. U.S. Naval Civil
Engineering Laboratory.
Song, Z. (2008). Pullout Behaviour of Suction Embedded Plate Anchors in Clay. PhD Thesis,
64
 Curtin University of Technology, Department of Civil and Construction Engineering,
Perth, Australia.
Sowers, G. B., and Sowers, G. F. (1973). Failures of Bulkhead and Excavation Bracing. Civil
Engineering, ASCE, 37(1), 72-77.
Tagaya, K., Scott, R. F., and Aboshi, H. (1988). Pullout Resistance of Buried Anchor in Sand.
Soils and Foundations, 28(3), 114-130.
Tagaya, K., Tanaka, A., and Aboshi, H. (1983). Application of Finite Element Mehod to Pullout
Resistance of Buried Anchor. Soils and Foundations, 23(3), 91-104.
Terzaghi, K. (1948). Soil Mechanics in Engineering Practice. New York Wiley.
Tiwari, B., and Al-Adhadh, A. R. (2014). Influence of Relative Density on Static Soil–Structure
Frictional Resistance of Dry and Saturated Sand. Geotechnical and Geological
Engineering, 32, 411-427.
Tiwari, B., Ajmera, B., and Kaya, G. (2010). Shear Strength Reduction at Soil Structure
Interaction. GeoFlorida 2010 Advances in Analysis, Modeling and Design (pp. 1747-
1756). Orlando, Florida, United States ASCE.
Vermeer, P. A., and Sutjiadi, W. (1985). The Uplift Resistance of Shallow Embedded Anchors.
Proceedings of the 11th International Conference on Soil Mechanics and Foundation
Engineering, 12–16 August, 4, pp. 1635-1638. San Francisco.
Zhao, L., Li, L., Yang, F., and Liu, X. (2011). Joined Influences of Nonlinearity and Dilation on
the Ultimate Pullout Capacity of Horizontal Shallow Plate Anchors by Energy
Dissipation Method. International Journal of Geomechanics, 11(3), 195-201.

65
 6 APPENDIX

Load-displacemnt curve obtained from PLAXIS 3D analysis:

The experimental study conducted by Naser (2006) for =19.6 kN/m3 was modelled, and
the output results obtained from FE analysis, i.e. displacement and ΣMstage, is presented
in column 1 and 2 of the following table. In this case, the input-load in the loading phase
was 5000 kN, which was large enough to cause ultimate collapse of the soil body before
ΣMstage reaching to unity. The applied load level (column 4) for a particular displacemnt
was calculated multiplying ΣMstage (column 2) with 5000 kN. The applied load (column
4) was plotted against displacemnt (column 3) to obtain load-displacemnt curve, which is
illustrated in Figure 4.16 (i).

Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)

1 2 3 4
0 0 0 0
1.61E-05 4.55E-05 0.016072 0.227417
1.99E-05 5.42E-05 0.019874 0.270851
2.38E-05 6.26E-05 0.023804 0.313036
2.78E-05 7.04E-05 0.027797 0.352179
3.58E-05 8.54E-05 0.03578 0.426838
4.39E-05 9.95E-05 0.04386 0.497479
5.19E-05 0.000113 0.0519 0.564809
5.99E-05 0.000126 0.059867 0.628676
6.77E-05 0.000138 0.067724 0.689357
7.16E-05 0.000144 0.071602 0.718197
7.94E-05 0.000155 0.079385 0.775166
8.72E-05 0.000166 0.087165 0.83008
0.000103 0.000187 0.102678 0.933642
0.00011 0.000196 0.110283 0.978809
0.000118 0.000205 0.117877 1.023429
0.000125 0.000213 0.125486 1.067468
0.000133 0.000222 0.133112 1.109569
0.000141 0.00023 0.140728 1.149241
0.000156 0.000245 0.155944 1.227336
0.000186 0.000274 0.186301 1.367598
0.000194 0.000279 0.193665 1.394102
0.000201 0.000284 0.200963 1.419568
0.000208 0.000289 0.208294 1.445574
0.000216 0.000294 0.215688 1.472328
66
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.00023 0.000305 0.230499 1.525453
0.000245 0.000315 0.245288 1.574718
0.000275 0.000334 0.274802 1.671803
0.000304 0.000351 0.304455 1.757048
0.000334 0.000367 0.333925 1.836391
0.000363 0.000381 0.363234 1.905428
0.000393 0.000394 0.392658 1.968475
0.000396 0.000395 0.396293 1.972569
0.0004 0.000395 0.399886 1.976232
0.000403 0.000396 0.403452 1.980478
0.000411 0.000398 0.41057 1.990287
0.000425 0.000403 0.424848 2.012689
0.000439 0.000408 0.439359 2.039374
0.000454 0.000414 0.454149 2.06848
0.000458 0.000416 0.458196 2.077762
0.000462 0.000418 0.46229 2.087787
0.000466 0.000419 0.466339 2.097434
0.000474 0.000423 0.474266 2.114281
0.000482 0.000426 0.482066 2.128928
0.000497 0.000431 0.497425 2.153886
0.000513 0.000435 0.512745 2.177054
0.00052 0.000437 0.520292 2.185466
0.000528 0.000439 0.527741 2.193315
0.000535 0.00044 0.535191 2.201808
0.00055 0.000444 0.550114 2.219539
0.00058 0.000451 0.580157 2.257281
0.000611 0.000459 0.611049 2.296514
0.000642 0.000466 0.642163 2.331124
0.000673 0.000473 0.673313 2.364056
0.000704 0.000479 0.704485 2.393564
0.000736 0.000484 0.73561 2.420641
0.000767 0.000489 0.766669 2.445688
0.000798 0.000494 0.797668 2.469252
0.000829 0.000498 0.828622 2.491688
0.00086 0.000503 0.85956 2.512992
0.00089 0.000507 0.890492 2.532957
0.000921 0.00051 0.921416 2.551411
0.000952 0.000514 0.952302 2.568348
0.000983 0.000517 0.983104 2.583822
0.001014 0.00052 1.013816 2.59847
0.001044 0.000523 1.044469 2.612708
0.001075 0.000525 1.075103 2.626699
67
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.001106 0.000528 1.105725 2.640236
0.001136 0.000531 1.136347 2.653127
0.001167 0.000533 1.166977 2.665227
0.001198 0.000535 1.197618 2.676522
0.001228 0.000537 1.228299 2.687293
0.001259 0.000539 1.258964 2.697435
0.00129 0.000541 1.289578 2.70698
0.00132 0.000543 1.320147 2.716024
0.001351 0.000545 1.350681 2.72466
0.001381 0.000547 1.381185 2.732949
0.001412 0.000548 1.411667 2.740942
0.001473 0.000551 1.472557 2.756098
0.001533 0.000554 1.533378 2.77057
0.001594 0.000557 1.594139 2.784238
0.001655 0.000559 1.654849 2.797129
0.001716 0.000562 1.715521 2.809348
0.001776 0.000564 1.776147 2.821063
0.001837 0.000567 1.836794 2.832913
0.001897 0.000569 1.897438 2.844283
0.001958 0.000571 1.958071 2.855044
0.002019 0.000573 2.018682 2.865136
0.002079 0.000575 2.079264 2.874469
0.00214 0.000577 2.139806 2.883085
0.002261 0.00058 2.260864 2.900148
0.002382 0.000583 2.381842 2.915723
0.002442 0.000585 2.442293 2.923121
0.002563 0.000587 2.563044 2.936719
0.002684 0.00059 2.683691 2.950225
0.002744 0.000591 2.744113 2.957042
0.002865 0.000594 2.86483 2.970081
0.002985 0.000596 2.985386 2.982303
0.003106 0.000599 3.105838 2.994048
0.003226 0.000601 3.226038 3.005632
0.003286 0.000602 3.285986 3.01098
0.003346 0.000603 3.345955 3.015185
0.003466 0.000605 3.465866 3.023466
0.003585 0.000606 3.585439 3.031414
0.003705 0.000608 3.704785 3.040188
0.003824 0.00061 3.824006 3.049159
0.003943 0.000612 3.943081 3.058054
0.004062 0.000613 4.061923 3.066269
0.004181 0.000615 4.180519 3.074383
68
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.004299 0.000617 4.299338 3.082572
0.004418 0.000618 4.418256 3.09067
0.004537 0.00062 4.537251 3.098741
0.004656 0.000621 4.656353 3.106654
0.004776 0.000623 4.775514 3.114224
0.005014 0.000626 5.013556 3.128326
0.005132 0.000627 5.132227 3.133658
0.005251 0.000628 5.25081 3.138076
0.005369 0.000628 5.369392 3.141883
0.005488 0.000629 5.487982 3.145545
0.005607 0.00063 5.606579 3.149721
0.005725 0.000631 5.725085 3.154124
0.005962 0.000632 5.962004 3.162169
0.006021 0.000633 6.021176 3.164496
0.00608 0.000633 6.080362 3.1667
0.006199 0.000634 6.19881 3.171005
0.006317 0.000635 6.317316 3.175235
0.006436 0.000636 6.436127 3.17942
0.006674 0.000637 6.67392 3.187118
0.006912 0.000639 6.911743 3.194463
0.00715 0.00064 7.149528 3.201781
0.007179 0.000641 7.179271 3.202866
0.007209 0.000641 7.209008 3.203593
0.007268 0.000641 7.268468 3.204472
0.007387 0.000641 7.38738 3.205739
0.007506 0.000641 7.506348 3.20742
0.007744 0.000642 7.744401 3.21225
0.007983 0.000644 7.982516 3.218376
0.008042 0.000644 8.042221 3.221031
0.008102 0.000645 8.101986 3.223708
0.008222 0.000646 8.221503 3.228667
0.00846 0.000647 8.460103 3.235986
0.008579 0.000648 8.579121 3.238647
0.008698 0.000648 8.698012 3.240514
0.008936 0.000649 8.935729 3.243963
0.008965 0.000649 8.965482 3.244627
0.008995 0.000649 8.99526 3.245204
0.009055 0.000649 9.054829 3.246017
0.009174 0.000649 9.173949 3.247294
0.009412 0.00065 9.411933 3.250248
0.00965 0.000651 9.64962 3.254519
0.009887 0.000652 9.887186 3.260602
69
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.010125 0.000653 10.12451 3.26594
0.010243 0.000654 10.24301 3.268165
0.010361 0.000654 10.36146 3.270262
0.01048 0.000654 10.47994 3.27206
0.010717 0.000655 10.71692 3.275383
0.010836 0.000655 10.8356 3.277163
0.010895 0.000656 10.89506 3.277956
0.011014 0.000656 11.01394 3.279541
0.011252 0.000657 11.25157 3.282662
0.011489 0.000657 11.48892 3.285975
0.011726 0.000658 11.72604 3.289493
0.011755 0.000658 11.75544 3.290502
0.011785 0.000658 11.78479 3.291473
0.011843 0.000659 11.84349 3.292959
0.011961 0.000659 11.96091 3.295069
0.012079 0.000659 12.07866 3.295765
0.012197 0.000659 12.19671 3.296236
0.012433 0.00066 12.43347 3.29851
0.01267 0.00066 12.67026 3.301213
0.012789 0.000661 12.78853 3.302966
0.013025 0.000661 13.02503 3.306245
0.013498 0.000662 13.4978 3.31216
0.013734 0.000663 13.73401 3.315441
0.013793 0.000663 13.79294 3.316188
0.013852 0.000663 13.85182 3.316852
0.01397 0.000664 13.96951 3.31798
0.014205 0.000664 14.20481 3.320142
0.014323 0.000664 14.32278 3.321373
0.014559 0.000665 14.55887 3.32387
0.014795 0.000665 14.79498 3.326326
0.015267 0.000666 15.26692 3.331393
0.015503 0.000667 15.50272 3.333447
0.015739 0.000667 15.73866 3.336545
0.015975 0.000668 15.97462 3.339348
0.016211 0.000668 16.21063 3.340934
0.016447 0.000668 16.44667 3.341975
0.016919 0.000669 16.91857 3.344037
0.01739 0.000669 17.39029 3.346562
0.017508 0.00067 17.50836 3.348005
0.017626 0.00067 17.62648 3.349372
0.017863 0.00067 17.86257 3.351567
0.018334 0.000671 18.33412 3.354849
70
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.018569 0.000671 18.56949 3.356124
0.018687 0.000671 18.68706 3.356857
0.018922 0.000672 18.92206 3.358432
0.01904 0.000672 19.03952 3.359351
0.019274 0.000672 19.27432 3.361112
0.019392 0.000672 19.39171 3.362027
0.019626 0.000673 19.62645 3.363791
0.019744 0.000673 19.74384 3.364584
0.019979 0.000673 19.97862 3.366104
0.020213 0.000674 20.21333 3.367593
0.020331 0.000674 20.3307 3.368339
0.020565 0.000674 20.56541 3.369746
0.0208 0.000674 20.80005 3.371114
0.021035 0.000675 21.0346 3.372515
0.021152 0.000675 21.15186 3.373349
0.021386 0.000675 21.38635 3.375029
0.021621 0.000675 21.62078 3.376584
0.02209 0.000676 22.08951 3.379606
0.022558 0.000677 22.55792 3.382556
0.022675 0.000677 22.67509 3.383296
0.022792 0.000677 22.79233 3.383869
0.023027 0.000677 23.02677 3.384932
0.023495 0.000677 23.49538 3.387398
0.023612 0.000678 23.61216 3.388229
0.023729 0.000678 23.72889 3.389097
0.023846 0.000678 23.84556 3.389907
0.024079 0.000678 24.07882 3.391469
0.024195 0.000678 24.19536 3.39203
0.024312 0.000679 24.31191 3.392503
0.024428 0.000679 24.42845 3.392924
0.024661 0.000679 24.66148 3.393782
0.024894 0.000679 24.89443 3.394761
0.025127 0.000679 25.12732 3.395825
0.025593 0.00068 25.59296 3.398006
0.025709 0.00068 25.70924 3.398568
0.025826 0.00068 25.82552 3.399156
0.026058 0.00068 26.05805 3.40033
0.026523 0.000681 26.52296 3.402671
0.027452 0.000681 27.4521 3.40744
0.027916 0.000682 27.91616 3.409786
0.028032 0.000682 28.03209 3.41024
0.028148 0.000682 28.14799 3.410639
71
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.02838 0.000682 28.37975 3.411401
0.028843 0.000683 28.84313 3.412958
0.029306 0.000683 29.30629 3.414745
0.029538 0.000683 29.53785 3.416046
0.030001 0.000684 30.00083 3.418561
0.030926 0.000685 30.92628 3.423283
0.031389 0.000685 31.38862 3.425138
0.031851 0.000685 31.85076 3.426512
0.032775 0.000686 32.77465 3.429287
0.033698 0.000686 33.69808 3.432304
0.033756 0.000687 33.7557 3.432857
0.033813 0.000687 33.81331 3.433263
0.033928 0.000687 33.92849 3.433948
0.034159 0.000687 34.1588 3.435177
0.034389 0.000687 34.38903 3.436203
0.034619 0.000687 34.61931 3.437212
0.03508 0.000688 35.07994 3.439182
0.03531 0.000688 35.31037 3.440119
0.035771 0.000688 35.77122 3.441757
0.036002 0.000688 36.00156 3.44224
0.036462 0.000689 36.46219 3.443177
0.036923 0.000689 36.92261 3.444269
0.037383 0.000689 37.38279 3.445669
0.037613 0.000689 37.61276 3.446549
0.038073 0.00069 38.07251 3.448349
0.038532 0.00069 38.53203 3.450068
0.038991 0.00069 38.9913 3.451862
0.03945 0.000691 39.45028 3.453857
0.039909 0.000691 39.90918 3.455803
0.040827 0.000692 40.82676 3.459517
0.041285 0.000692 41.28537 3.461322
0.041515 0.000692 41.51457 3.462127
0.041973 0.000693 41.97281 3.463377
0.042431 0.000693 42.43098 3.464573
0.04266 0.000693 42.65989 3.465207
0.043118 0.000693 43.11768 3.466312
0.044033 0.000694 44.03332 3.46846
0.044949 0.000694 44.94933 3.470837
0.045408 0.000694 45.40761 3.472249
0.045637 0.000695 45.63684 3.473069
0.045866 0.000695 45.86614 3.473874
0.046325 0.000695 46.32489 3.47525
72
Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.047243 0.000696 47.24268 3.477677
0.048161 0.000696 48.16061 3.480033
0.04862 0.000696 48.6196 3.481153
0.049079 0.000696 49.07864 3.48215
0.049538 0.000697 49.53774 3.482985
0.050456 0.000697 50.456 3.48457
0.051374 0.000697 51.37429 3.486133
0.051833 0.000697 51.83334 3.487127
0.052063 0.000698 52.06281 3.487677
0.052522 0.000698 52.52177 3.488657
0.05344 0.000698 53.43975 3.490365
0.054358 0.000698 54.35759 3.492043
0.055275 0.000699 55.27518 3.494008
0.055734 0.000699 55.73383 3.495085
0.056651 0.000699 56.651 3.497172
0.05688 0.0007 56.88023 3.497706
0.057339 0.0007 57.33867 3.498667
0.057797 0.0007 57.79713 3.499515
0.058256 0.0007 58.25558 3.500307
0.058485 0.0007 58.4848 3.500687
0.058599 0.0007 58.59941 3.500841
0.058829 0.0007 58.82863 3.501093
0.059287 0.0007 59.28708 3.501582
0.059746 0.0007 59.74554 3.502152
0.060204 0.000701 60.20399 3.502756
0.060662 0.000701 60.66243 3.503413
0.061121 0.000701 61.12084 3.50415
0.061579 0.000701 61.57921 3.504947
0.062038 0.000701 62.03756 3.505779
0.062496 0.000701 62.49588 3.506628
0.062954 0.000701 62.95418 3.507463
0.063412 0.000702 63.41246 3.508266
0.063871 0.000702 63.87072 3.509032
0.064787 0.000702 64.78718 3.510466
0.065245 0.000702 65.24536 3.511129
0.065704 0.000702 65.70352 3.511719
0.06662 0.000703 66.61981 3.512796
0.067078 0.000703 67.0779 3.513359
0.067536 0.000703 67.53598 3.513911
0.067994 0.000703 67.99403 3.514447
0.068452 0.000703 68.45206 3.51498
0.06891 0.000703 68.91006 3.515523
73
 Displacement (m) ΣMstage Displacemnt (mm) Applied Load (kN)
1 2 3 4
0.069368 0.000703 69.36802 3.516086
0.069826 0.000703 69.82594 3.516674
0.070284 0.000703 70.28383 3.517278
0.070742 0.000704 70.74168 3.517885
0.071199 0.000704 71.19949 3.518485
0.071657 0.000704 71.65728 3.519087
0.072115 0.000704 72.11503 3.51969
0.07303 0.000704 73.03047 3.520894
0.073946 0.000704 73.94582 3.522088
0.074861 0.000705 74.8611 3.523262
0.075776 0.000705 75.77625 3.524363
0.076234 0.000705 76.23381 3.52486
0.076691 0.000705 76.69137 3.525268

74

View publication stats