Citation
Sandbag structural stability analysis

Material Information

Title:
Sandbag structural stability analysis
Creator:
Pearson, Walter Charles
Publication Date:
Language:
English
Physical Description:
vii, 306 leaves : illustrations ; 29 cm

Subjects

Subjects / Keywords:
Flood dams and reservoirs -- Design and construction ( lcsh )
Retaining walls -- Design and construction ( lcsh )
Flood dams and reservoirs -- Design and construction ( fast )
Retaining walls -- Design and construction ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaf 305).
General Note:
Submitted in partial fulfillment of the requirements for the degree, Master of Science, Civil Engineering.
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Walter Charles Pearson.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
31508823 ( OCLC )
ocm31508823
Classification:
LD1190.E53 1994m .P43 ( lcc )

Full Text
SANDBAG STRUCTURAL STABILITY ANALYSIS
by
Walter Charles Pearson, Jr.
M.S.P.E. Colorado School of Mines, 1968
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
1994


CONTENTS
Chapter
1. Introduction.......................................................... 1
1.1 Problem Statement.................................................. 1
1.2 Research Objective................................................. 3
1.3 Engineering Significance............................................4
1.4 Method of Research................................................. 5
2. Review of Literature ...................................................9
2.1 Sandbags for Flood Control..........................................9
2.2 Geotextiles for Sandbags ......................................... 16
2.2.1 Geotextile Materials....................................... 17
2.2.2 Material Properties........................................ 19
2.2.3 Designing with Geotextiles................................. 24
3. Properties of Geotextiles as Sandbag Materials .......................28
3.1 Material Properties .............................................. 29
3.1.1 Mechanical Properties...................................... 29
3.1.2 Hydraulic Properties ...................................... 33
3.1.3 Degradation Properties .................................... 34
3.2 Direct Shear Test Program and Results ............................ 37
3.2.1 Test Procedure ............................................ 37
3.2.2 Test Apparatus .............................................42
v


3.2.3 Geosynthetics Tested .......................................49
3.2.4 Black Split-Film Woven Polypropylene....................... 51
3.2.5 Woven Natural Jute......................................... 63
3.2.6 White Woven Split-Film Polypropylene....................... 69
3.2.7 Nonwoven Needle-Punched Geotextile ........................ 72
3.2.8 Summary.................................................... 82
4. Properties of Sand and Pore Pressure Distribution ................... 83
4.1 Index Properties................................................. 83
4.2 Pore Pressure Distribution ..................................... 88
5. Stability Analysis.................................................... 100
5.1 Development of Analysis Models................................... 101
5.1.1 Factor of Safety Sliding................................ 102
5.1.2 Factor of Safety Overturning ........................... 110
5.1.3 Factor of Safety Slumping............................... 116
5.1.4 Bearing Capacity Implications ............................ 125
5.2 Analyses of Conventional Sandbag Geotextiles..................... 125
5.2.1 Black Split-Film Woven Polypropylene...................... 127
5.2.2 Woven Natural Jute........................................ 135
5.2.3 White Woven Split-Film Polypropylene...................... 142
5.3 Analysis of Alternative Geotextile .............................. 148
5.4 Water Velocity Normal to the Structure............................ 149
5.5 Comparison of Conventional Geotextiles............................ 155
6. Discussion of Analyses Results........................................ 160
VI


6.1 Configuration Implications
161
6.2 Cost Analysis .............................................. 165
7. Summary and Conclusions.......................................... 171
7.1 Summary .................................................... 172
7.2 Conclusions ................................................ 178
7.3 Recommendations for Future Study ........................... 181
Appendix
A. Geotextile Properties........................................... 183
B. Calibration Curves Direct Shear Test Apparatus............... 185
C. Direct Shear Test Results Raw Data........................... 186
D. Curve Fit Analyses Spreadsheets................................. 218
E. Factor of Safety Spreadsheets .................................. 248
References ......................................................... 305
vii


1. Introduction
1.1 Problem Statement
The annual and incessant problem of flooding is conventionally responded to with the
age-old and proven "sandbag". The recent flooding problems in the mid-continent of the United
States amply demonstrates a need for an improved approach to flood abatement and the
subsequent repairs to earth structures. During these times of crisis all resources,including the
finite human resources, are over taxed. The opportunity exists for implementation of innovative
approaches that more efficiently and timely utilize the human and physical resources available
to construct flood control earth structures.
The current "sandbag" technique is to utilize a semi-standard 14 inches by 26 inches
woven polypropylene or natural fiber bag. These bags can hold up to 50 pounds of dry sand and
are used untied. The bags are stacked in a "brick" like fashion and/or cross-stacked for
additional stability. Over 125,000,0001 bags were used in the recent 1993 flooding in the mid-
United States. They are individually filled by hand and are often non-uniform in dimensions and
weight. This lack of uniformity and associated structural integrity are both functions of fill time
Estimate from telephone conservation with Corps of Engineers fall 1993.
1


and the remaining energy of the labor force. To enhance the structural integrity and safety,
rational construction and analysis procedures are needed.
Literature research indicates that the design of sandbag structures is based on the time
tested techniques dating back as far as the Egyptian Pharaohs. Flood abatement has been a
recurrent problem throughout documented time, however, references for the analysis of sandbag
structural stability or alternative configurations appear to be quite limited. A record has not been
found of a rational safety analysis for the potential of sliding, overturning, and slumping of
sandbag earth structures. It was determined that the standard guideline is 2.5 feet of base width
for each 1 foot of barrier height2, a semi-pyramidal shape based on the dimensions of the
sandbag. Structural designs using these techniques have been adapted for individual needs as
Figure 1.1 Typical Sandbag Structure
Flood Fighting with Sandbags
U.S. Army Corps of Engineers
The absence of rational analysis procedures and documentation of such analysis
2 U.S. Army Corps of Engineers, Flood Fighting with Sandbags
2


provided the opportunity for investigations into this realm. The potential for development of cost
effective improvements to the conventional structural design or alternative configurations
appeared reasonable for pursuit. The reduction, even if minor, in the base to height ratio would
result in significant reduction in material cost, labor requirements, and construction time. The
improvements in geotextiles further enhances the potential for benefits to be derived from
research on this subject. Individually, sandbags are inexpensive and simple of function,
however, when 125,000,000 are involved the opportunity for cost benefit is great.
1.2 Research Objective
The objectives of this proposal were threefold, first to investigate the structural stability
of the conventional sandbag structures, second to investigatealtemativestructural configurations
utilizing conventional sandbag geotextiles, and third to investigate structural configurations
utilizing an alternative geotextile. The goal of this proposal was to define combinations of
structural configurations and sandbag materials that are more efficient with respect to time, labor
and cost.
The attainment of these objectives required three intermediate objectives:
0 determination of the geotextile to geotextile internal resistance to sliding, ie. the coefficient
of friction3, to, and adhesion4, ca;
A constant proportionality factor, relating normal stress and the corresponding critical
shear stress, at which sliding starts between two surfaces. ASTM D 5321-92
4 The shearing resistance between two geotextiles under zero externally applied forces.
ASTM D 5321-92
3


0 determination of average filled sandbag physical properties, and;
0 development of a rational analysis procedure for sandbag structures, conventional and
alternative.
Geotextile to geotextile shear properties were not readily available in the literature or
from manufacturers due to the lack of economic incentive to define them. This requires a direct
shear testing program using nonconventional and expensive test equipment. Sand properties
are abundant, however, the encapsulating of the sand in the sandbag necessitated defining
certain basic properties through elementary test procedures. The rational analysis procedure
had to model the field, real world, conditions as effectively as possible and include the potential
for sliding, overturning, slumping, and evaluation of required foundation bearing capacity. These
intermediate objectives had to be met prior to pursuit of the primary thesis objectives.
1.3 Engineering Significance
The time tested sandbag approach to flood abatement and other uses appears to be as
prudent an immediate solution to emergency situations today as it was in past centuries.
However, there has been a significant change in the construction of such structures, ie. the cost
of resources to construct these structures especially the labor and materials. The thesis
objective was to define the best engineered application of this most dependable solution to
emergency situations.
A sandbag structures stability is a function of:
0 height-to-base dimensions, cross-sectional configuration;
4


0 stacking pattern, usually a brick-like pattern;
0 weight of the filled sandbag, packed sand density, ie. net bag weight;
0 the bag-to-bag friction angle, fi;
0 the bag-to-bag mechanical interlock or adhesion;
0 forces expected on the structure as water height, saturation, and normal water velocity, and;
0 foundation bearing capacity.
This thesis incorporates the investigation of these seven interrelated parameters from
both a theoretical and an empirical perspective. The goal was to define a rational approach to
evaluate the factor of safety of alternative configurations with respect to sliding, overturning and
slumping. The determined evaluation procedure includes the ability to analyze alternative
sandbag materials, configurations, fill materials, and expected conditions in a timely procedure
that can be utilized in an emergency situation.
1.4 Method of Research
The attainment of the intermediate objectives was accomplished prior to the
development of the rational analysis procedure. The basic sandbag physical properties were
defined using a structurally sound vessel of known volume to determine the volumetric density
of sand packed by dumping, hand tamping with a 10 pound weight, and shaking and hand
tamping. The sand packed vessel was then weighed and the vessel weight subtracted for
determination of net sand pack density. The trials were repeated for saturated sand to enable
comparison of dry and saturated (submerged) sand densities.
Commercially available black woven split-film polypropylene sandbags were supplied
5


by Nicolon Corporation. These are made using Mirafi 500X geotextile, a 5.5 ounce/yard
material. The sandbags were then packed per the Corps of Engineers Flood Fighting with
Sandbags manual and weighed. These were then compared with the prior determined densities.
These defined densities were then used in the developed rational analysis procedure.
The geotextile to geotextile interfacial shear resistance properties were defined utilizing
ASTM D 5321-92, Standard Test Method for Determining the Coefficient of Soil and
Geosynthetic or Geosynthetic and Geosynthetic Friction by Direct Shear Method5. The Bureau
of Reclamation, Denver office, made available their test apparatus located at the Denver Federal
center in Lakewood, Colorado, for this testing. Three geotextiles currently used for sandbags
were tested in both the cross and long bag directions as defined in more detail in Chapter 3,
Section 3.2.1 and Section 3.2.2. These properties are not available from literature or from
manufacturers. A fourth geotextile, a nonwoven needle-punched material not currently used for
sandbag construction, was also tested for analysis as a potential alternative to the conventional
geotextiles. It was tested in the machine and cross machine directions. The geotextiles tested
are defined in Table 1.1.
These direct shear test results were then analyzed to define the best numerical fit of
shear strength as a function of normal stress. Linear, hyperbolic and parabolic analyses were
conducted. The analysis of this test data is presented in Chapter 3, Sections 3.4 to 3.7. The
This test method is under the jurisdiction of ASTM Committee D-35 on Geosynthetics
and is the direct responsibility of Subcommittee D35.01 on Mechanical Properties.
Current edition approved Oct. 15, 1992, Published December 1992.
6


conventional soil interpretation and that most often used for geosynthetic to soil or to other
geosynthetic analysis is the linear fit or Mohr-Coulomb Failure Criteria. The best numerical fit
to the test data was determined to vary with the geotextile. The best fit for each individual
geotextile is utilized in the rational analysis of Chapter 5. A comparison of the fi angles and
adhesion values from linear analysis are presented in Chapter 5. for comparative purposes.
Manufacturer Product Identification Geotextile Type Test Direction Supplier
Mirafi 500X black woven polypropylene machine & cross machine Nicolon
Langston Bags Jute sandbag woven natural jute cross & long bag Langston Bags
Unknown N/A white woven Polypropylene cross bag Corps of Engineers
Nicolon S800 nonwoven needle punched machine & cross machine Nicolon
Table 1.1 Tested Geotextiles
The rational analysis of the structures, conventional and alternative, was accomplished
via spreadsheets developed on Lotus. These were developed to analyze the structural
configurations and geotextiles for the potential of sliding, overturning, and slumping and to define
the minimum foundation bearing capacity required to support the structural configurations. The
theory and assumptions behind each analysis model are presented in Chapter 5. Section 5.1.
These models were developed to be user friendly and functional in time constrained situations.
The developed analysis models were used to design efficient and sound structural
7


configurations for each geotextile based on the previously defined properties. The structural
designs were then compared on a cost, time to construct and required materials basis. The
thesis concludes with recommended configurations for each geotextile and suggestions for
further research.
8


2. Review of Literature
2.1 Sandbags for Flood Control
An extensive effort was made to find references pertaining to the history of sandbags,
the application of sandbags for structures and the analysis of sandbag structures. The reference
search included four Corps of Engineer libraries, the university librarys computer assisted search
system, CARL, which allowed a search of all regional libraries, and hiring a Science and
Engineering Bibliographer, Ms. Mara Sprain, to assist in searching commercial library data
bases. It also included communication with the Corps of Engineers Vicksburg, Mississippi, office
who supplied the materials available to them pertaining to the construction of flood abatement
and ammunition storage structures. The reference search concluded disappointingly, finding no
documentable history of sandbag usage and no rational models to evaluate the structural
integrity of sandbag structures.
The Corps of Engineers directed the search to a gentleman, Mr. Guy Osterneck, in
Raleigh, North Carolina, who was somewhat of a sandbag historian. Mr. Osterneck passed away
two years prior to my contact and with him most of the knowledge he had accumulated.
However, an employee of Mr. Osterneck, Mr. Ben Morris, was able to confirm some of the
conclusions arrived at in the reference search about the history of sandbag utilization and
9


structural configuration.
The size and shape of the individual sandbag is the one and only basis of the design of
sandbag structures. A flat 14 inch by 26 inch sandbag filled to two thirds full results in a sandbag
of approximate dimensions 4 inches tall, 10 inches wide, and 17 inches long. If these are
stacked in a brick like configuration with each new layer, in the vertical direction, offset one half
a sandbag width the result is a pyramidal cross-sectional configuration with 2.5 times the height
in basal width. The structural shape is a result of the unit component dimensions.
The sandbag size is a function of weight, the weight an average low level construction
worker of past centuries could reasonably carry for an extended period of time. This weight
range was then and is now around 40 pounds. Varying from 30 pounds if the sand is dry and
the sandbag half full to just over 50 pounds if the sandbag is two thirds full and the sand
saturated. Historical managers of construction used knowledge from the agricultural managers
that purveyance workers can more easily carry grain in bags of rectangular shape, ie. the long
direction vertical and the short direction horizontal. This knowledge resulted in the flat sandbag
dimension of 14 inches wide by 26 inches long and the base to height ratio of 2.5:1.
Mr. Morris also indicated that, per his understanding, sandbags have been utilized for
temporary construction almost continuously since humans invented cloth. Sandbags have
provided structural elements for shelters, roads and other structures, used for protection in battle,
and utilized in flood control efforts since before recorded time. They are size wise today much
the same as they were in the beginning. Only the materials used to make them have changed,
10


and considering jute as a major material for modern sandbags, this too has not changed all that
much.
Through the Corps of Engineers Vicksburg office copies of currently available manuals
for flood fighting and construction of ammunition structures were obtained. These manuals are:
0 Flood Fighting with Sandbags U.S. Army Corps of Engineers North Pacific Division
A 36 page introductory manual for constructing sandbag flood abatement structures.
0 Flood Emergency Handbook U.S. Army Corps of Engineers Vicksburg District
December 1991
A 45 page manual defining flood fighting methods in emergency situations including the
utilization of sandbags.
0 Geosvnthetic Reinforced Barricades for Ammunition Storage U.S. Army corps of
Engineers Waterways Experiment Station Vicksburg October 1992
This manual defines construction methods using sandbags and other geosynthetic
materials for the safe storage of ammunition.
The Flood Fighting with Sandbags manual is a straightforward introductory instruction
primer for construction of flood abatement structures with sandbags. It instructs, in detail, on the
filling of individual bags making the point that the sandbags are to be left untied for tighter
placement of the individual bags and, therefore, reduced potential for seepage through the
structure.
The main thrust of the manual is the construction of the pyramidal configuration of the
11


structure. The manual details the placement of individual sandbags. The sandbags are to be
lapped end over end and offset one half a sandbag length. It is pointed out that the net effective
length of a sandbag is one foot due to this lapping. This stacking procedure results in three
sandbags in height equaling about one foot, ie. four inches height per sandbag.
The structure is to be set back a minimum of one foot from the levee water side edge.
The height to base width ratio is one foot of height to two and a half feet of base width. Each
sandbag is to be tamped into place by walking on the sandbags (Figure 1.1). A formula is
presented to calculate the number of sandbags required per foot of height for a foot of structure
length:
N = number of sandbags
(3*H+9*H2)
N----------------- H = height of structure
The manual goes on to present options in fill materials, methods, locations, and the
alternative uses of sandbags in flood fighting. There are sections on levee failure, ringing sand
boils, and use with other materials as straw bales. This is a manual typically made available for
training in emergency situations.
The Flood Emergency Handbook is a "standing operating procedures" manual
describing procedures successfully utilized in prior flood fighting efforts. It details the typical
situations that a supervisor may encounter during a flood event and provides guidance in prudent
responses to specific situations. Mention is made of utilizing sacked earth to abate topping of
levees and sand boils. This is a generalized guidance document which includes numerous
12


figures detailing the construction of specific flood control structures. However, this is not a
technical record or engineering design manual. This document provided an understanding of
the general and widespread use of sandbags in flood control but does not answer to the primary
objectives of the thesis.
The third manual, Geosvnthetic Reinforced Barricades for Ammunition Storage, is
focused mainly on the use of sheet geosynthetics and expanded geocell mattress in structure
construction. It does point to the long-standing use of sandbags and mentions how they are
utilized. Here too there is nothing encompassing the focus of the thesis but as background the
material was useful.
Additionally, a manual entitled Effectiveness of Expedient Levee-Raising Structures. U.S.
Army Corps of Engineers Waterways Experiment Station Vicksburg (April 1988), was located.
This manual presents testing conducted to "define the static and dynamic load limits beyond
which selected existing US Army Corps of Engineers (Corps) designs of expedient levee-raising
structures will fail". This was a full scale testing procedure conducted on the Big Black River 10
miles southeast of Vicksburg, Mississippi. A test basin was constructed for controlled testing of
potato ridges, earth-filled sacks, plastic grid with sand fill, plywood flashboard, plywood
flashboard with earth backing, planking flashboard with earth fill backing, and mud boxes with
earth fill. These tests included two, four and six foot structures and testing was both static and
dynamic. The emphasis of these tests with earth filled sacks was to determine the potential for
leaching of fill material from sacks, the seepage through the filled sacks, and structural
deterioration of the sack material due to wave action.
13


The earth filled sacks were used in the two and four foot high structures. The structure
was constructed using the CELMK Flood Emergency Handbook (1975). Loosely woven burlap
sacks were used for the two foot structure. A six foot wide base with a two and a half foot crown
configuration was used for the two foot tall structure. A 0.2 foot trench was cut as a key seat for
the structure and to prevent piping. Two earth fills were used, clayey silt and clay gravel. The
four foot height structure had dimensions of 12 foot wide at the base and 2.5 feet wide at the
crown. Looselywoven burlap, woven polypropylene, and spun woven polypropylene sacks were
used in the four foot structure tests. Again the structure was reset 0.2 feet into the levee as the
two foot structure was.
The two foot test sections were exposed for 240 hours at 0.5 foot static differential head,
and 24 hours each at 1.0 and 1.5 foot static differential heads. The clayey silt filled sacks were
spongy but the gravel filled sacks remained solid. Seepage through the structure was minimal
and no leaching of fill material was experienced. The structure was exposed to two to three
inches of rain during the static testing.
The sections were then exposed to wave action for 19 hours at the 1.0 foot water level
over a period of 191 hours total time. The silt filled sack experienced excessive leaching and the
gravel bags some leaching. Finally an exposure of 19 hours of wave action at the 1.3 foot water
level over a time period of 384 hours was conducted. This resulted in rapid erosion of silt filled
sacks during the first 6 hours but this slowed throughout the remainder of the test. Seepage
through the structure was never a problem. The structure was exposed to 4 to 5 inches of snow
and 0.5 inches of ice during the dynamic testing.
14


The four foot structures were constructed and left exposed 3 months prior to testing.
During this time 20 inches of rain fell on the structure and the burlap sacks showed noticeable
deterioration. During testing the structure was exposed to 750 hours of 1.1 to 2.0 foot static
differential head. The test site was flooded out during this test for one month. The burlap sacks
continued to deteriorate but the polypropylene sacks showed no such signs. There was no
evidence of leaching from the polypropylene sacks and seepage through the structure was only
minor. The differential head was raised to 3.0 and 3.5 feet for 48 hours. Seepage through the
structure increased slightly but leaching was not evident.
The dynamic testing included a 1.0 foot depth and exposure for 19 hours over a total
time of 264 hours. Damage was sustained by the burlap sacks but not the polypropylene sacks
to any extent. The test was repeated at a 2.0 foot depth for 19 hours over a total time of 670
hours. Site flooding was experienced which extended the testing period. Considerable structural
damage was incurred but the structure held. The spun woven polypropylene sustained the
greatest damage due to sewing thread deterioration. Finally an exposure of 8 hours of 3.0 water
depth wave action over a total of 116 hours conducted with the structure remaining structurally
sound.
The report concludes that earth filled sacks are effective for structures of two and four
feet in height under the static conditions tested and mild wave action. It did note that the woven
polypropylenes slick texture made handling difficult and resulted in some leaching of fill material.
The burlap and spun woven polypropylene experienced deterioration of fabric and seams due
to exposure to weather and sunlight.
15


This report was read with interest and generally was helpful but it does not address the
thrust of the thesis objectives. The subject tests were not designed to test the factor of safety
of the structures with respect to sliding, overturning or slumping. The structural configuration,
while similar to that proposed by the Corps of Engineers, is wider per foot of height than that in
the before referenced Flood Fighting with Sandbags. These tests did not directly attempt to fail
the structure to determine these factors of safety.
The above review includes all references encountered that were felt to be meaningful
to the thesis effort. The small number of references uncovered was a surprise and a
disappointment.
2.2 Geotextiles for Sandbags
The reference search for geotextile properties was into a much more fruitful arena. The
recent explosion in the utilization of geosynthetics in general and geotextiles specifically provides
an unlimited opportunity for references for all but the most critical property to the thesis, the
resistance to sliding of one geotextile on itself. Economics are universally the most powerful of
the driving forces and this is true in the research associated with geosynthetics. The surge in
savings using geosynthetics has spurred the base of knowledge about geosynthetics in all
economically attractive areas. Sandbags would seem to have missed to proverbial boat so far
as they have not been viewed as economically attractive for research.
16


A "geotextile" is defined by Robert M. Koerner6 in his book as "Any permeable textile
used with foundation, soil, rock, earth, or any other geotechnical engineering-related material as
an integral part of a human-made project, structure, or system.". The ASTM definition per D
4439-92a is a permeable geosynthetic comprised solely of textiles.". The Living Webster
Encyclopedic Dictionary defines a "textile" as "Woven or capable of being woven; pertaining to
weaving. n. A fabric made by weaving; any material as yarn or thread, which may be used for
weaving.". This is an expansive definition which can include anything woven by humans and
utilized to build or construct. The material used in the manufacture of sandbags would surely fit
this definition.
Koerner points out in the overview of his book that geotextiles form the largest group of
geosynthetics. In fact, geotextiles have served humans as far back as the Egyptians, Romans,
and early Chinese dynasties. They were of natural fibers woven into textiles which were
incorporated into structures for enhanced structural value. The sandbag may have been one of
the first such geotextile enhancements. Throughout history as textiles have improved so has the
contribution made by the incorporation of these textiles into engineered structures.
2.2.1 Geotextile Materials
The modern geotextiles are most often, but not always, made of synthetic materials.
This has both increased their material properties and their life expectancy. The most common
6 Koerner, R.M. (1990), Designing with Geosynthetics. Second Edition, Prentice-Hall,
Englewood Cliffs, NJ
17


materials used for geotextiles are7:
0 polypropylene 65%,
0 polyester 32%,
0 polyamide (nylon) 2%, and,
0 polyethylene -1%
Structurally they are either woven or nonwoven textiles of monofilament, multifiliment,
staple, staple yarn, split film, or split-film yarn. The woven geotextiles can be plain weave, basket
weave, twill weave, or satin weave. The nonwoven products can be needle-punched, melt-
bonded, or resin-bonded. The variety of options in material properties is nearly unlimited. The
geotextile can truly be designed for the application. The applications are conventionally divided
into five functions:
0 separation,
0 reinforcement,
0 filtration,
0 drainage, and,
0 moisture barrier.
The application of geotextiles for sandbags is both a separation and a reinforcement
function. The most common materials utilized are woven polypropylene or natural jute. These
are readily available and inexpensive geotextiles. These are important qualities when over
7 Koerner, R.M. (1990)
18


100,000,000 sandbags are required.
The polypropylene sandbags are usually a plain weave of split film. The Mirafi 500X
used in the thesis is a typical geotextile for sandbags. Its material properties are presented in
Appendix A and discussed in the next section. This is a black polypropylene with added carbon
for better ultraviolet, UV, light resistance. A white woven polypropylene is also commonly used
which is manufactured in the Philippines. The universal sandbag throughout history has been
the natural jute or hemp sack. These three geotextiles appear to make up the majority of
sandbags used today.
2.2.2 Material Properties
The material properties are conventionally divided into:
0 physical properties,
0 mechanical properties,
0 hydraulic properties,
0 endurance properties, and;
0 degradation properties.
The physical properties of specific gravity, mass per unit area, thickness, and stiffness
are not of great significance in sandbag design. These are index properties and not design
properties. The mechanical properties are more pertinent and indicate the geotextiles resistance
to the mechanical stresses it encounters. The primary mechanical properties are frictional
behavior, tensile strength and seam strength. To a lesser extent fatigue strength, tear strength,
19


impact resistance, and puncture resistance are important. Burst strength, compressibility and
pullout resistance are a small concern.
The most important geotextile property to this thesis is the frictional behavior, interfacial
shear resistance, of the geotextile with itself. It is, as stated above, the least defined property.
The definition of this property through direct shear testing is detailed in Chapter 3. Section 3.2.
This utilized ASTM Designation: D 5321 92 which will be reviewed in Section 3.2. The classical
interpretation of this property for soil or soil in combination with geosynthetics is the Mohr-
Coulomb Failure Criteria:
t = shear strength
on = normal stress
T=onTan($)+c $ = internal friction angle
c = cohesion factor
The frictional behavior for geotextile to geotextile is often assumed to follow the same
Mohr-Coulomb relationship and be linear. This may not always be the case as will be
demonstrated in the later chapters. Whatever the relationship, this is the property that defines
the sandbag structures ability to resist the lateral forces of the flood waters and is, therefore, the
single most important property of the geotextile in sandbag usage.
Tensile strength of the geotextile must meet a minimum value to be functional for
sandbags. This is a three part function: first the sandbags ability to be used to transport sand
from the loading point to the structure location, second the sandbags resistance to the shear
forces of the water, and thirdly the sandbags ability to resist the active earth pressures developed
20


within the sandbag structure. Each loaded sandbag weighs approximately 30 to 50 pounds and
the geotextile must accommodate this weight during construction. This does not seem to be a
limitation of any of the conventional sandbag geotextiles.
The resistance to shear forces is critical for the high end of the normal stress range
tested with the natural jute and Mirafi 500X. The shear resistance is sufficient for each of these
geotextiles that the tensile strength of the basic material is exceeded before the maximum shear
resistance is reached. The individual split film or yarn fibers fail in tension and the weave breaks
apart allowing the sandbag to fail. This could be a problem in the taller structures especially for
the natural jute. The Mirafi 500X has a tensile strength sufficient for any reasonable structure
height.
The active earth pressure developed in a sandbag structure is a function of the
structure's height, the net sand pack density, the saturation within the sandbags, and the
properties of the fill sand. The maximum value of active earth pressure will be at the bottom of
the structure below the peak of the structure. This maximum value can be calculated using the
Rankin analysis method. The failure of a single sandbag in this position should not result in
structural failure. However, multiple failures in the central foundation area could result in
structural slumping and reduced structural integrity which could result in structural failure.
Seam strength should exceed the geotextile tensile strength. A sandbag should never
fail in the seams. This property includes the total mechanical properties of the seam thread or
bonding material and the disturbed geotextile area of the seam. Modern seaming procedures
21


allow this requirement to be met routinely.
The hydraulic properties of porosity, permittivity (cross plane), and transmissivity (in
plane) are not of major importance in sandbag design. The object of the sandbag structure is
to abate flood waters, however, the fill materials porosity and permeability far overshadow the
sandbag geotextiles ability to transmit water in most instances. Corps of Engineers full scale
testing, reviewed in Section 2.1, found little seepage of water through the structure with
conventional sandbag geotextiles and either clayey silt or gravel.
The important hydraulic properties are the soil retention properties of apparent opening
size (AOS), percent open area (POA), and soil retention. These measure the geotextile's open
area as a function of total fabric area. This is most important on the water side when there is
wave action due to currents and/or wind. The leaching of fill material can result in localized
slumping of the structure and partial or total failure. Since most fill material is sand in the range
of sieve size 30 to 40, 0.590 to 0.420 mm, this is a practical minimum AOS for sandbags. The
use of finer fill material would require a larger AOS, ie. a more expensive geotextile.
Endurance property tests are designed for geotextiles used in more permanent
geotechnical applications than sandbag structures. However, the abrasion is important when
the sandbags are transported, rearranged, or otherwise moved. The geotextile must be resistant
to individual sandbag failure in these circumstances. Also to a minor extent creep resistance can
be important if the geotextile elongates easily.
22


The degradation properties with respect to temperature, sunlight, chemical exposure and
biological processes are all important depending on the circumstances. Only rarely are
sandbags exposed to excessively high temperatures. They are often exposed to relatively cold
temperatures and embrittlement due to cold can be of concern. Sunlight or UV degradation is
a common cause of sandbag structural problems. The use of carbon additives to increase
resistance to UV degradation is routine in sandbag geotextiles. The Mirafi 500X is an example
of this. The white woven Philippine polypropylene sandbags are reported to be highly
susceptible to UV degradation as are the natural fiber sandbags.
The exposure to chemicals is uncommon in flood abatement actions but can occur in
specialized situations so it must be considered in such circumstances. Toxic spill control is a
common use for sandbags. These situations require specialized geotextiles which will not be
considered in this thesis.
Biological processes are common in flood abatement efforts. These can range from the
rotting of natural fiber sandbags to exposure to wastewater and/or wastewater treatment
chemicals. This occurred in the 1993 Mid-Continent flooding. Another natural hazard is rodent
damage in storage or when in place. These problems are more common with natural fibers than
synthetic geotextiles.
The sandbag structure is by design a temporary structure. The desired useful life is the
duration of the flood event. It must then be removed and disposed of in a proper manner.
Utilizing too substantial a geotextile for the sandbag can result in increased cost of removal. This
23


must be considered in the selection of the geotextile and the fill material. The routine removal
procedure is to use front end loaders to break up the sandbags and load the material in earth
transport trucks for disposal in landfills. Seldom is it economical to salvage the sandbags or fill
material. This can be a plus for the natural fiber sandbags over the more resistant polypropylene
sandbags.
2.2.3 Designing with Geotextiles
The Industrial Fabric Association International8 defines three most common design
methods as:
0 Design-by-experience, which is how contemporary sandbag structures are most often
designed and built. The experience of centuries continues to be applied. It is not
broken so it does not need to be fixed.
0 Design-by-specification, practiced by most government agencies. Specifications are
set to meet specific agency needs that are routine and/or repetitive.
0 Design-by-function, becoming a preferred method for geotechnical applications of
geosynthetics. This method is the basis for the rational analysis procedures utilized
in the later chapters of this thesis.
The design-by-function can be divided into the following sequential steps:
0 determine the primary function of the geosynthetic,
0 determine the required property values,
8 A Design Primer: Geotextiles and Related Materials, First Edition, 1992
24


0 obtain property values of the candidate geosynthetic,
0 calculate the factor of safety as:
AllowedjTest) Value
Required{Design) Value
0 evaluate if the calculated factor of safety is sufficient for the application.
Throughout this thesis the functions of concern will be separation and reinforcement.
The required properties of the geotextile will be developed in Chapter 3, Section 3.1 and
discussed in detail there. The candidate geotextile properties are also discussed in Chapter 3,
Section 3.1 and 3.2. They were either obtained from the manufacturer or obtained through
testing.
The primary factor of safety values of interest to this thesis are the sliding, overturning,
and slumping values. Their derivation and calculation procedures are the subject matter of
Chapter 5, Stability Analysis. The sliding factor of safety is calculated as:
p-. _ ResistingForce
stid DrivingForce
The resisting force is the summation of vertical forces times the tangent of the interfacial friction
angle, 4>, plus the adhesion factor, ca, times the length over which it is effective. The vertical
forces include the weight of the structure, the weight of the water vertically on the structure, and
the effect of pore pressure, u. The driving force is the summation of the horizontal forces normal
to the sandbag structure. This includes the water pressure and the velocity related force normal
25


to the structure. The Corps of Engineers uses a factor of safety of 1.33 to 1.5 for sliding in
construction calculations for permanent structures9. They have no value requirement for
sandbag structures. The 1.5 value will be used in this thesis.
The Corps of Engineers criteria for overturning is a function of the minimum base area
in compression and is stated as a percentage of the total soil or rock foundation. Typical values
are 75 to 100% for soil and 50 to 75% for rock. Typical factor of safety values used in
geotechnical design are 1.5 for cohesionless soils and 2.0 for cohesive soils. The thesis uses
a value of 1.5 as the thesis deals with sand as the primary fill material.
The slumping potential is a modification of the rotational failure of soil slopes. The
encasement of the fill material in the sandbag adds the geotextile tensile strength to the resisting
force and negates this potential problem to a large degree. However, there still remains a
potential for the sandbags to slump individually or in a mass. This is evaluated using a modified
approach similar to the sliding potential with an angular, non-horizontal, driving orientation. The
resisting force is a function of the resistance to sliding of the geotextile, determined through
testing, over the area it is effective, and the weight normal to the area of the sliding surface. This
will be developed in detail in Chapter 5. Section 5.1.3.
The typical factor of safety used in geotechnical design for slope stability is 1.25 for an
embankment or 1.3 if significant damage is a potential. This thesis uses the 1.3 value for
9 U.S. Army Corps of Engineers, Retaining and Flood Walls. EM 1110-2-2502, 29 Sept
1989
26


evaluations of slumping potential.
Bearing capacity is not within the scope of this thesis. However, the thesis includes the
required minimum bearing capacities for all structural configurations developed. The
spreadsheet based rational analysis procedures incorporate calculation of minimum bearing
capacity based on the typical geotechnical value of 3.0 for factor of safety.
27


3. Properties of Geotextiles as Sandbag Materials
Four geotextiles were selected for inclusion in this thesis, three are current sandbag
materials and one is a potential alternative geotextile. The three current sandbag geotextiles are
woven black and white polypropylene and a woven natural jute. The fourth geotextile is a
nonwoven needle-punched polypropylene. The material properties of only the black woven
polypropylene and the nonwoven needle-punched polypropylene were available from the
manufacturer. These property sheets are available in Appendix A. The natural jute and the
white woven polypropylene are less expensive products not having a specification sheet.
The application functions required of these geotextiles for use as sandbags are
"separation" and reinforcement". The primary material properties are the "mechanical
properties" of "frictional behavior" or fabric interfacial shear resistance, Section 3.2; "tensile
strength", Section 3.1.1; and "seam strength", Section 3.1.1. The fundamental "hydraulic"
property is "soil retention", Section 3.1.2. The major degradation property is the "UV resistance",
Section 3.1.3. Other material properties can be important in specific circumstances but were not
considered as meaningful to the thesis development.
28


3.1 Material Properties
3.1.1 Mechanical Properties
The mechanical properties of interest are the frictional behavior, tensile strength, and the
seam strength. Other mechanical properties as fatigue strength, tear strength, impact resistance,
puncture strength, burst strength, compressibility, and pullout resistance are of lessor concern
to the thesis goals but were met by the selected geotextiles. The frictional behavior of the
selected geotextiles was not available from any source and specialized testing was required to
obtain this data. This is the subject matter of Section 3.2 and will not be addressed in this
section.
The tensile strength required of a sandbag geotextile is a function of the weight of sand,
fill material, to be transported in the sandbag, the shear force developed from water
impoundment, and the active earth pressure at the central foundation level of the structure. The
size of the sand bag and the effective density of the sand pack define the weight to be carried.
This equates to a volume of 0.3935 cubic feet and a maximum expected weight of 50 pounds
for the conventional sandbag. The required tensile strength, ar, can then be estimated as:
ar= Required Tensile Strength
a =^2-
r WSB Fsb = Filled Sandbag Weight = 50 [lbs]
WSB = Geotextile Supporting Width = 28 [in]
This calculates to be 1.79 [Ibs/in] a value far below the wide width tensile strength of the woven
black polypropylene value of 130 [Ibs/in] and the nonwoven needle-punched value of 225 [Ibs/in].
No tensile strength values were available for the other two geotextiles, however, it was
reasonable to assume they also surpassed this requirement.
29


The shear force developed from the water impoundment is a function of the water depth
and the current velocity normal to the structure. A maximum height of sandbag structures was
not encountered in the literature search and no limit was provided from contacts with Corps of
Engineers personnel. The testing reviewed in Chapter 2. Section 2.1 indicated a maximum
height of four feet for earth fill sack structures; however, this would seem too low for the
objectives of this thesis. A maximum height value was set for the thesis of 15 feet as it was felt
this would cover the reasonable range to be encountered in the practical world.
The water pressure at 15 feet is 6.5 psi or 936 psf. This acting on the bottom layer of
sandbags would equate to 316 Ibf per foot of structure. Using the equation on the previous page
for required tensile strength, ar, and the length of the sandbag, 12 inches, this equates to 26
[lb/in]. This is again well below the tensile strength of the selected geotextiles at static
conditions.
The velocity component can be estimated using the full flow velocity of the Mississippi
River, 5 to 10 fps, an assumed angle of incidence of 45 and the Bernoulli equation for steady
state flow of incompressible fluids. The high end velocity of 10 fps flow rate was calculated from
information reported in the World Book Encyclopedia for the Mississippi River. The incidence
angle of 45 is on the conservative side as typical values are expected to be 20 or less. The
Bernoulli equation velocity component becomes:
V = Velocity [fps]
Pressure=
Yw = Density of water = 62.4 [pcf]
g = gravity = 32.2 [fps2]
2 *g
30


This can be converted to force by multiplying by the effective area, 48 square inches.
This results in a force of 32 Ibf. Adding this to the static value of 316 Ibf and dividing by the
sandbag length, 12 inches, results in a required tensile strength of 29 [lb/in]. This is still well
below the reported tensile strength.
The active earth pressure at the foundation level and center of the structure is the
expected maximum for the structure. This was estimated using the Rankin method and
assuming a level back fill as a conservative case. The required tensile strength developed by
the active earth pressure is calculated as:
___ a, = Required tensile Strength
AREA = 48 in2
' WIDTH WIDTH =12 in
and:
oH=a'z* Tan(45 y) + u
= (ysb Yw) (Z)
u = (YW) (Z)
oH = 1234 PSF = 8.6 PSI
a, = 34.4 [lb/in]
This remains well below the reported tensile strength of the two geotextiles on which
material properties are available and can reasonably be expected to be below the other two
geotextiles. The forces generated within and on the sandbag structure are not anticipated to tax
the tensile strength of the geotextile used for sandbag construction.
Z = Maximum height = 15 feet
Yse = 127 PCF
Yw = 62.4 PCF
= 40 for a Dense Packed Sand
V = 0.8 <|> = 32
31


It was determined during direct shear testing that the limiting material property at normal
stresses above 7 psi for the natural jute and Mirafi 500X was the tensile strength of the
geotextile. At these normal stresses the geotextile failed in tension prior to shearing. This is
presented in more detail in Section 3.2.
Seam strength is the final critical mechanical property for sandbags. This property must
meet at a minimum and preferably exceed the tensile strength of the geotextile. The Corps of
Engineers pays much attention to this property in the standards they set for sandbag
procurement. The seams may be either sewn with a thread of the similar material as the
geotextile or bonded chemically or thermally.
The Mirafi 500X sandbags are bonded and stitched to meet this requirement. The white
polypropylene sandbags from the Philippines are chained stitched with a thread that is stronger
than the fabric. The natural jute sandbags are also chain stitched with a stronger thread, nylon,
using a quarter inch spacing for stitches. The Nicolon S800 is not a current sandbag geotextile
and no seam specifications are available.
Of the lessor important mechanical properties the tear strength, impact resistance, and
puncture strength need to be sufficient to resist damage caused by debris in the moving water.
This could be an area of research for another thesis. It has been assumed that these are not
critical properties for the objectives of this thesis. The testing reviewed in Chapter 2
demonstrated that these properties can be important but that conventional sandbag geotextiles
perform acceptably in most circumstances.
32


Burst strength would become critical if the structure was intended to support traffic as
trucks or construction equipment. This is rarely the case and then the structural configuration
must be radically modified. Compressibility becomes significant only where the geotextile is used
to transmit fluids in the plane of the geotextile. This would be contrary to the objectives of a
sandbag structure. The sandbag encapsulates the fill material as opposed to supporting the fill
through anchorage so pullout resistance is not a factor for this situation.
3.1.2 Hydraulic Properties
Most of these properties pertain to the movement of fluids in the plane or cross the plane
of the geotextile. As stated previously, this is contrary to the function of the sandbag utilization
and is not critical to the material property specifications. The flood waters will inundate the
structure and the sandbags and fill material will be saturated. The soil retention properties are
the only critical hydraulic properties to the integrity of the sandbag structure. The review of the
Corps of Engineers testing in Chapter 2. demonstrates the significance of fill leaching to
structural integrity.
The specific hydraulic properties of POA, AOS, and soil retention are related to the open
area of the fabric as compared to the total fabric area. The most critical of these is the AOS,
apparent opening size. This is defined as the U.S. standard sieve number that has openings
closest in size to the openings in the fabric10. This becomes a useful and convenient design
property as it can be directly related to the fill material utilized.
10 CW-02215
33


For this thesis a washed and graded silica sand of sieve size 30 was used. The common
range of fill material was not found in the literature search. Corps of Engineers personnel
reported that in emergency situations any available fill material will be considered. This can
range from clay to gravel. Ideally they would prefer 20 to 40 sieve size silica sand. Construction
grade sand is in this same range and is most usually available in bulk quantities throughout the
country.
The Mirafi 500X has an AOS of 40 and the Nicolon S800 an AOS of 70. The natural jute
AOS has not been tested but is, by inspection, much smaller in numerical value, ie. much larger
in opening size. These openings are, on average, .0625 inches or an equivalent of sieve size
number of 14. The white woven polypropylene openings appear to be approximately the same
as the Mirafi 500X. Only the natural jute's AOS exceeds what would seem desirable and it held
the 30 mesh sand very well when filled and jolted. Based on the above, it would appear that any
geotextile that effectively retains 30 to 60 sieve size sand will perform adequately for sandbag
utilization.
3.1.3 Degradation Properties
How long will the geotextile last? This is the basis for degradation properties. In the
context of the sandbag structure it only needs to last for the duration of the flood event, days or
weeks. Once the flood event has past the structure is destroyed and removed to a landfill. This
past summer's flooding was unusual and lasted several months. This defines the upper limit of
required geotextile life for conventional sandbag usage.
34


The degradation properties include resistance to temperature, sunlight, chemical
exposure, and biological processes. Sunlight is the most common concern as photons of light
break the polymers chemical bonds. There is a wavelength threshold for each polymer above
which bonds will not be broken but below which bonds are broken. This makes the shorter
wavelengths more critical. Sunlight is composed of:
0 infrared wavelengths longer than 760 mm
0 visible wavelengths from 400 to 760 mm
0 ultraviolet (UV) wavelengths shorter than 400 mm
The UV range can be divided as:
0 UV-A 400 to 315 mm results in some polymer damage
0 UV-B 315 to 280 mm causes most polymer damage
0 UV-C 280 to 100 mm only in outer space
The UV damage varies from season and with weather conditions as clouds, moisture,
temperature, wind, etc. The shorter wavelengths are more pronounced in the summer and least
in the winter.
The polypropylene geosynthetics are very susceptible to the UV degradation and may
lose material, strength, properties quickly if exposed to direct sunlightfor even short timeframes.
The addition of carbon to the geotextile enhances the resistance to this problem. The Mirafi
500X has a UV resistance rated at 70% of strength after 500 hours exposure. The Nicolon S800
geotextile is also rated at 70%. Both were tested per the ASTM D-4355 test procedure.
The natural jute and the white polypropylene are not rated. The natural jute was
35


reported by the manufacturer and Corps of Engineers personnel to retain strength properties for
the duration of normal flood events. The white polypropylene sandbags were used in Operation
Desert Storm where direct sunlight was excessive yet the sandbags provided the protection
desired. From the above it can be concluded that a wide range of geotextiles are UV resistant
enough to function as sandbags for flood events.
Temperature is seldom a problem leading to degradation of sandbags. It would most
likely be a problem in cold weather situations where the sandbags could become stiff and/or
brittle. The normal temperature range of flood events would not result in such problems.
Chemical and/or biological process exposure would not be common in the average flood
event and would not make these properties a concern to design of the average sandbag. The
1993 Mid-Continent flood event did have both chemical and biological exposures to flood
abatement structures. These were due to the flood overrunning wastewater treatment plants and
industrial facilities. Still this did not reach a critical stage or require specialized geotextiles. If the
sandbags were used for chemical spill control then specialized geotextiles would be appropriate.
However, this is well beyond the scope of this thesis.
Finally, there are degradation problems experienced uniquely with the natural fiber
sandbags. These are rotting, dry or wet, and rodent destruction. The jute or hemp material
deteriorates quickly in moist climates and more slowly in dry climates. Because of this natural
fiber sandbags are rarely warehoused for long time periods. This is a natural fiber not unlike
silage and is a potential rodent food source. This can be a problem in storage and in the
36


structure. Again this was beyond the scope of the thesis and was not pursued.
3.2 Direct Shear Test Program and Results
3.2.1 Test Procedure
The "Standard Test Method for Determining the Coefficient of Soil and Geosynthetic or
Geosynthetic and Geosynthetic Friction by the Direct Shear Method"11 was the procedure
followed for this material testing. This test method determines the interfacial shear resistance
between two geosynthetics and is intended as an indication of field conditions. ASTM indicates
in the procedure that the procedure is applicable for all geosynthetics.
Terminology used in the procedure is referenced back to a document Terminology D
443912, also an ASTM document. The procedure defines the "coefficient of friction" as "a
constant proportionality factor, relating normal stress and the corresponding critical shear stress,
at which sliding starts between two surfaces". "Direct shear friction test for geosynthetics" is
defined as "a procedure in which the interface between a geosynthetic and any other surface,
under a range of normal stress specified by the user, is stressed to failure by the horizontal
movement of one surface against the other". This procedure is similar to and consistent with
direct shear test procedures for soils.
The coefficient of friction between any two geosynthetic materials, including the same
11 ASTM Designation: D 5321 92 Approved October 15, 1992. Published December
1992.
12 D 4439 "Terminology for Geosynthetics" Annual Book of ASTM Standards, Vol. 04.08
37


materials, is determined by placing the two materials in a direct shear container, applying a
constant normal compressive stress, and applying a tangential shear force so that one section
of the container moves with respect to the other section. A record is made of shear force as a
function of horizontal displacement of the shear container. A minimum of three tests under three
unique normal stresses are tested. Peak, or residual, shear stresses are plotted versus the
corresponding applied normal stresses to define the desired material properties.
Generally the results are fitted to a linear relation and the slope, coefficient of friction,
and intercept, adhesion, are defined. The method is intended as a" performance test to provide
the user with a set of design values for the test conditions". The test measures the total
resistance to sliding. This may be a combination of several individual mechanisms as sliding,
rolling, or interlocking. The test cannot distinguish between the shear strength attributable to
each individual mechanism. The individual mechanisms can be important and an effort should
be made to determine their relative significance.
A rigid device is required to hold the geosynthetic material securely to ensure that a
uniform force without torque can be applied to the test material. The direct shear testing device
consists of a stationary and a moving container sufficiently rigid to prevent distortion during
shearing. This procedure requires the traveling container to be supported on firm bearings and
a rack to ensure movement only parallel to the applied shear force.
Square or rectangular shear containers are recommended. A minimum dimension of
300 mm is defined in the ASTM procedure. It is also recommended that the minimum depth be
38


50 mm or six times the maximum particle size when soil is tested with the geosynthetic.
Normal stress is applied by a device that can maintain a uniform normal stress ( 2%).
This can include weights, pneumatic or hydraulic bellows, or piston-applied stresses. Shear
force loading is applied by a device that is capable of maintaining a constant rate of displacement
( 10%) in a direction parallel to the direction of travel of the moving container. The procedure
specifies a constant readout of applied force. Conventionally, a load cell or proving ring
arrangement is utilized. Additionally, the shear loading connection should be in the plane of the
shearing interface.
A continuous readout of the horizontal shear displacement and if desired, the vertical
displacement are recommended. The indicators should have a minimum of 75 mm (3 inches)
for the horizontal and 25 mm (1 inch) for the vertical displacement.
The clamping devices to affix the geosynthetic to the containers should not interfere with
the shearing surfaces and must retain the geosynthetic flat during testing. Flat jawed devices
are recommended. Gluing is also an option if it better models the field situation.
The shear testing device must be calibrated to determine any internal resistance inherent
to the device. This is required later to adjust the measured results and to insure that device is
in calibration.
The testing of geosynthetic on geosynthetic does not require humidity control as tests
39


with soils would. The geosynthetic is placed flat on the container, covering the entire length of
the container and extending over the edges of the container. It is clamped to the sides as
recommended above. The upper container is placed in position above the lower container and
fixed as required by the device. The upper geosynthetic material is attached to a rigid media and
placed in the upper container so the geosynthetic materials are flat and in direct contact. The
upper material must protrude below the upper container within the test area.
A rigid stratum is placed over the upper material to ensure that uniform stress is applied
over the entire test area. The loading plate can then be placed and the desired normal
compressive load applied. The displacement indicators are then located and zeroed with respect
to the traveling container. The shear loading ram is attached and zeroed. The direct shear
testing is then ready to begin.
The shear force is applied using a constant rate of displacement of up to 5 mm/minute
(0.2 inches/minute). The shear force is recorded as a function of displacement horizontally.
Continuous readings are recommended or a minimum of 20 data points should be obtained per
test. The test is run until the shear force remains constant with increasing displacement. The
total displacement should be in the range of 25 to 75 mm (1 to 3 inches).
The normal stress is removed at the end of the test and device dissembled. The failure
surfaces should be carefully inspected to identify the failure mechanisms involved. Special note
should be made of the clamp areas.
40


The direct shear test is then rerun at at least two additional normal stresses on new
specimens of the geosynthetic. The results are then plotted as applied shear force versus
container displacement on linear scales. This allows identification of the peak and/or residual
shear force(s) and their respective horizontal displacements. These values are then corrected
for the internal shear correction defined above.
The apparent shear stress can then be determined for each test normal compressive
stress as:
T = shear stress
F, = shear force
Ae = corrected area (If the effective contact area decreases with
displacement)
Tests, where the area of the material contact does not decrease with displacement, do not
require area correction.
A plot is then made of each shear stress value versus the respective applied normal
compressive stress with the same scale for each axes. Conventionally, a linear fit is made of the
plotted data to determine the intercept, adhesion, and the slope, coefficient of friction. The
plotted data can optionally be fitted to non-linear functions as hyperbolic or parabolic curves.
41


3.2.2 Test Apparatus
The facilities of the Bureau of Reclamation at the Federal Center in Lakewood, Colorado,
were utilized for the direct shear testing. The basic components of the testing equipment are:
0 A Wykeham Farrance of Slough, England, 100K stepless shear box machine
(Figure 3.1) This was modified by the Bureau of Reclamation for testing
geosynthetics with soil or other geosynthetics. It has a traveling lower container and
a stationary upper container. Both with dimensions of 300 mm or more.
0 The load cell is a BLH Electric Type U3G1 (serial number 925227) with a capacity
of 10,000 pounds (Figure 3.2). This is coupled with a computer based recorder with
capacity for 15 second interval continuous readings and internal interpretation of
results.
0 The normal compressive stress is applied hydraulically by the Wykeham Farrance
shear box as an integral part of the machine. It utilizes a vertical yoke frame (Figure
3.3) to apply the normal force perpendicular to the material in the upper container
which is stationary. It has an upper load limit of 10,000 pounds. The normal stress
is monitored by a Schaevite SCM-700 strain gage (Figure 3.4) coupled with a
Hewlett-Packard 3468A Multimeter. The readings are recorded by the computer
contemporaneously with the shearforce and horizontal and vertical displacements.
0 The vertical and horizontal displacements are monitored and recorded on the
computer continuously. The displacement gages are of laboratory quality but were
42


43


Figure 3.3 Vertical Yoke Frame
44


not identified by make or type. They have a displacement range of up to 2 inches.
The equipment functioned well for the desired testing. The limitation of the 2 inch of
horizontal displacement did not prove to be a real detriment as all tests failed prior to 1.8 inches
of horizontal displacement. The equipment met all requirements of the ASTM testing procedure
except the desired horizontal displacement.
The traveling lower container (Figure 3.5) has dimensions of 16 inches wide by 16 inches
long, direction of travel. The geosynthetic is attached to the forward end of the lower container
by a four bolt jawed clamp (Figure 3.6). The aft end of the lower container utilizes two dual bolt
clamps on either side of the horizontal displacement ram fixture (Figure 3.7). These proved to
provide the desired positioning of the geosynthetic on the lower traveling container. The
container is made of 2 inch thick and 4 inch deep stainless steel frame with a solid plexiglass
filled center. The material placement surface is flat.
The upper stationary container is of the same dimensions as the lower container and
made of stainless steel (Figure 3.8). The core area, 12 inches square, is open. The upper
geosynthetic material is wrapped around a .75 inch thick 12 inch by 12 inch plywood square and
affixed to the back by 5/8s inch staples (Figure 3.9). It is then placed in the upper container and
an additional 3.125 inch thick 12 inch square plywood block added above it. The normal load
platen (Figure 3.10) was placed inside the core of the upper container and on the plywood
blocks. This held the material flat and in place as desired.
45


46


Figure 3.7 Lower Container Two Bolt End
Figure 3.8 Stationary Upper Container
47


The normal compressive stress was monitored continuously and adjusted as required
to maintain the desired normal load. The 2% limitation was not a practical problem but did
require monitoring to achieve. The horizontal displacement rate of 5 mm/minute was set into the
Wykeham Farrance stepless shear box instrument panel and was maintained within the 10% of
displacement rate limitation. The mechanical connection of the horizontal ram is parallel to the
plane of the shearing interface and at the top of the lower container.
As stated above all readings are by computer and were taken at 15 second intervals
through the peak shear force range. This included the shear force, normal compressive stress,
and all displacements. The data sets all exceeded 100 data points. These data are provided
in Appendix C.
The direct shear test equipment was calibrated by staff of the Bureau of Reclamation for
stresses and displacement prior to this testing and the resulting calibration curves are in the
Appendix B. The device is not zeroed for shear force prior to each test. It maintains a low level
reading which is recorded and subtracted from test results. This is a function of the load cell and
this is a standard practice by the Bureau staff.
3.2.3 Geosynthetics Tested
Tests were conducted on three geotextiles currently utilized as sandbag material:
0 woven split-film polypropylene (black);
0 woven natural jute, and;
0 woven split-film polypropylene (white).
49


The black woven split-film polypropylene was supplied by Nicolon and was tested in both
the machine13, long bag14, direction and the cross machine15, cross bag16, direction. The
woven natural jute geotextile was acquired from Langston Bags of Memphis, Tennessee.
Langston Bags is a routine supplier of sandbags to the Corps of Engineers. The jute geotextile
was obtained as sandbags and was tested in the cross and long bag directions. The white
woven split-film polypropylene geotextile was obtained from the Corps of Engineers' supply in
Vicksburg, Mississippi. It is manufactured in the Philippines. This too was obtained as sandbags
and was tested in the cross direction.
The fourth geotextile tested is not currently used as a material for sandbags:
0 nonwoven needle-punched (8 oz/yd).
It was supplied by Nicolon and is designated as S800 Nonwoven. This is a needle-
punched geotextile that was felt would have superior coefficient of friction and adhesion
properties. This material was tested in both the machine and cross machine directions.
The direct shear tests were conducted at normal compressive stresses of 2, 7, and 13
psi for most geotextiles for convenience of comparison. This was varied for the natural jute in
13 Machine direction is the direction the geotextile is manufactured in, ie. the long direction
of the roll.
14 Long bag direction is the long dimension from open end to bag bottom.
15 Cross machine direction is the direction at 90 to the machine direction.
16 Cross bag direction is the direction at 90 to the long bag direction.
50


the long bag direction because the results in the cross bag direction indicated excessive
structural damage at normal stresses above 7 psi. It was also varied for the white woven
polypropylene geotextile due to limited available material.
The testing was designed to determine the peak shear stress for each geotextile tested
as opposed to the residual shear values. The peak is felt to best represent the field situation for
sandbag structures. Values were determined for residual shear stress for the black woven
polypropylene, the nonwoven needle-punched, and the natural jute geotextiles tested. These
were obtained for comparison value. The results of the testing and analysis are presented
individually in the following sections.
3.2.4 Black Split-Film Woven Polypropylene
The physical property values for this geotextile are presented in Appendix A and were
discussed in section 3.1 above. This geotextile is utilized for sandbags supplied for flood
abatement. Samples of these sandbags were obtained and utilized in testing as well as samples
of the bulk geotextile. The geotextile was tested in both the machine direction, corresponding
to the long sandbag direction, and in the cross machine direction using new material and for
residual values using the material failed at 2 or 13 psi normal compressive stress.
Prior to testing a general observation was made that there is a significant difference in
the properties of interest between the long bag, machine, and cross bag, cross machine,
directions. The machine direction having the superior properties. The apparent adhesion in this
direction even in the absence of normal stress is significant. The cross machine or cross bag
51


direction displayed no such adhesion without applied normal stress. This is felt to be important
as the critical direction is the cross bag direction for sandbag structure stability. The testing
included the following conditions:
Test Group Test Type Peak/Residual Testing Direction Normal Stresses PSI
I Peak Machine 2, 7, & 13
II Residual (2 PSI) Machine 2, 7, & 13
III Peak Cross Machine 2, 7, & 13
IV Residual (13 PSI) Cross Machine 13
Table 3.1 Test Conditions
The recorded raw data is presented in Appendix C for all tests. This was processed and
plotted as normalized horizontal displacement versus shear stress to define the peak shear value
for each normal compressive stress. Normalized horizontal displacement is the percent of
horizontal displacement, ie. the horizontal displacement divided by the geotextile contact length
times 100. The results for the initial test group are presented on the following page, Figure 3.11.
These plots defined peak shear stresses of:
Normal Compressive Peak Shear
Stress Stress
PSI PSI
2 3.6
7 8.1
13 11.2
Table 3.2 Peak Shear Stress
These values were then plotted on linear scales to define the coefficient of friction and
52


Figure 3.11 Direct Shear Test Results
cn
(a)
DIRECT SHEAR TEST RESULTS TEST GROUP I NEW BLACK POLYPROPYLENE
0%

13.00- MACHINE DIRECTION g



's.





0U
8.00 N.
/
ft /
Cm / S' \.
/ \
a 5.50- / / / /
/ / "N
s ) /
S s
/' /
1 / s
A'" x
X
1.50- ~T~

! i )
0.0 3% 1.0 3% 2.0 3% 3.0 3% 4.0 )% 5.0 3% 6.0 NORM/ 3% 7.0 lLIZED 3% 8.0 HORIZC 3% 9.0 3NTALI 3% 10.( 3ISPLAC <3% ll.( ^MENI 0% 12.( 0% 13.( 0% 14.( 0% 15.( <3% 16.(
2 PS1 7 PSI 13 PSI



adhesion values based on a linear regression fit of the data points. This analysis is presented
in Appendix D. The analysis defined an adhesion value of 2.6 psi, a A, internal friction angle,
of 34.4, and a coefficient of friction of 0.6852.
The data was then analyzed using hyperbolic and parabolic curve fitting procedures,
spreadsheets in Appendix D. The comparative analysis is presented as Table 3.1, page 55.
This indicated the best fit was obtained with the non-linear curves. Both the hyperbolic and
parabolic fits are reasonable fits. The hyperbolic fit was selected based on the overall results of
all Test Groups.
This defines a hyperbolic equation for shear strength as a function of normal
compressive stress as:
_ t = shear strength
_________n________
(0.445 +0.055 oj an = normal compressive stress
T=
This equation is utilized later in Chapter 5., Stability Analysis, to analyze current sandbag
structures and design alternative structural configurations. The above analysis was conducted
for Test Groups II and III. The comparative analysis spreadsheet and graphical presentations
are presented on pages 56 to 60. As stated previously, the best fit of the data was with the
hyperbolic equation and the equations for Test Groups II and III are as follows:
Test Group II
Tn =
(0.635 +0.079 *o)
54


COMPARATIVE FIT ANALYSIS
NEW BLACK WOVEN POLYPROPYLENE
MACHINE DIRECTION_____________
NORMAL CALCULATED SHEAR STRESS RAW DATA
STRESS LINEAR HYPERBOLIC PARABOLIC SHEAR
PSI PSI PSI PSI PSI
0.000 2.6088 0.0000 0.2009
1.000 3.2940 1.9978 2.1780
2.000 3.9791 3.6000 3.5393 3.600
3.000 4.6643 4.9134 4.6458
4.000 5.3495 6.0098 5.6026
5.000 6.0346 6.9387 6.4578
6.000 6.7198 7.7358 7.2381
7.000 7.4049 8.4274 7.9604 8.100
8.000 8.0901 9.0330 8.6360
9.000 8.7753 9.5678 9.2728
10.000 9.4604 10.0435 9.8770
11.000 10.1456 10.4694 10.4529
12.000 10.8308 10.8529 11.0044
13.000 11.5159 11.2000 11.5341 11.200
14.000 12.2011 11.5157 12.0446
DIRECT SHEAR TEST RESULTS
15
CO
CL
CO
LU
cc
j
CO
cc
^ 5
LU
X
CO
0
0 5 10 15 20 25
NORMAL STRESS (PSI)
a LINEAR --HYPER. A PARB. RAW DATA
Table 3.3 Comparative Spreadsheet
NORMAL VS. SHEAR STRESS
55


Figure 3.12 Direct Shear Test Results
Ol
O)
DIRECT SHEAR TEST RESULTS TEST GROUP II BLACK POLYPROPYLENE
10%
8.60-
8.40= MA CHINE DIRECTl ON ^
s!oo-
7.80=
S'
s \,

/
u. ou_ ./ **...


s(>
u>uu_
r~
!5 J.uu fc 5.40=
5.20= \
55 j.uui X

} /
,*' /
*
/

i
t/i s
/
! V- i
ou /
f f
/
/
: i /
1. ou_ y
1 iUU i
/
s
/
/
[
)
J
0.0 )% 1.0 )% 2.0 )% 3.0 )% 4.0 )% 5.0 m 6.0 NORM/ )% 7.0 lLIZED )% 8.0 HORIZC )% 9.0 )NTALI )% 10.( 3ISPLAC 0% ll.t :ement 0% 12.( 0% 13.( 0% 14.( 0% 15.( 0% 16.(
2 PSI NEW 7PSIRES 13 PSI RES



COMPARATIVE FIT ANALYSIS
BLACK WOVEN POLYPROPYLENE
RESIDUAL 2 PSI MACHINE DIRECTION
NORMAL CALCULATED SHEAR STRESS RAW DATA
STRESS LINEAR HYPERBOLIC PARABOLIC SHEAR
PSI PSI PSI 1 PSI PSI
0.000 1.8888 0.0000 0.3245
1.000 2.3640 1.4000 1.7572
2.000 2.8391 2.5200 2.6675 2.520
3.000 3.3143 3.4364 3.3930
4.000 3.7895 4.2000 4.0146
5.000 4.2646 4.8462 4.5673
6.000 4.7398 5.4000 5.0698
7.000 5.2149 5.8800 5.5338 5.800
8.000 5.6901 6.3000 5.9669
9.000 6.1653 6.6706 6.3747
10.000 6.6404 7.0000 6.7611
11.000 7.1156 7.2947 7.1291
12.000 7.5908 7.5600 7.4812
13.000 8.0659 7.8000 7.8192 7.800
14.000 8.5411 8.0182 8.1447
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
a LINEAR HYPER. A PARB. RAW DATA
Table 3.4 Comparative Spreadsheet
57


Figure 3.13 Direct Shear Test Results
Ol
00
DIRECT SHEAR TEST RESULTS TEST GROUP III BLACK POLYPROPYLENE
0%

.1. ou
3.60- CROSS MACHINE DIRECTION ||
3.50- -/ / *
\ ; r.
/
/ v v /' \ t /
/ V ' \
/ *


/



t



o£ 1.80-
< 1.70- ; "7
J :
(/) 1.MJ-
:
t
l

i
t
j
i

0.50- Wj




0.00-| 0.0 Wo 1.0 Wo 2.0 Wo 3.0 Wo 4.0 Wo 5.0 NC Wo 6.0 IRMALIi Wo 7.0 IED HO Wo 8.0 RIZONTi Wo 9.0 \L DISPI Wo 10.1 JVCEME 0% ll.C NT 0% 12.1 0% 13.1 10% 14.( 0% 15.1
2 PSI 7 PSI 13PSI



COMPARATIVE FIT ANALYSIS
NEW BLACK WOVEN POLYPROPYLENE
CROSS MACHINE DIRECTION
NORMAL CALCULATED SHEAR STRESS RAW DATA
STRESS LINEAR HYPERBOLIC) PARABOLIC SHEAR
PSI PSI PSI I PSI PSI
0.000 -0.0965 0.0000 ERR
1.000 0.1976 0.2577 ERR
2.000 0.4916 0.5200 0.3858 0.520
3.000 0.7857 0.7871 1.0342
4.000 1.0798 1.0590 1.4806
5.000 1.3738 1.3359 1.8435
6.000 1.6679 1.6181 2.1573
7.000 1.9620 1.9055 2.4377 1.910
8.000 2.2560 2.1983 2.6937
9.000 2.5501 2.4968 2.9306
10.000 2.8442 2.8010 3.1521
11.000 3.1382 3.1112 3.3610
12.000 3.4323 3.4274 3.5591
13.000 3.7264 3.7500 3.7480 3.750
14.000 4.0204 4.0791 3.9288
co
CL
CO
CO
ID
QC
\
CO
cc
<
ID
I
CO
DIRECT SHEAR TEST RESULTS
0 5 10 15
NORMAL STRESS (PSI)
h LINEAR HYPER. A PARB. RAW DATA
Table 3.5 Comparative Spreadsheet
59


Figure 3.14 Direct Shear Test Results
O)
o
DIRECT SHEAR TEST RESULTS TEST GROUP IV BLACK POLYPROPYLENE
0%

cDnccMArunjc nmcmnN L
3.60-
3.50- \
\
/ \ A
\ / V / \ r-
s * \ - V' ' \ / \^\ -/ \




2.60





K A.UU











u.ou

u.uu





0.00% 1.0 Wo 2.0 Wo 3.0 Wo 4.0 Wo 5.0 NC Wo 6.0 1RMALU Wo 7.0 LED HO] Wo 8.0 UZONTi Wo 9.0 \L DISPI Wo 10.C .ACEME X)% ll.( NT 0% 12.C 0% 13.( 0% 14.C 0% 15.(
13 PSI



Test Group III
T" (3.915-0.034 *an)
Test Group IV is for a single normal compressive stress and no analysis was conducted.
It was run for comparison with the 13 psi normal stress of Test Group III.
Test Group II tested the residual shear stress of material failed at 2 psi normal stress.
The data plot very similar but below those of Test Group I. However, there was noticeable
structural damage to the fabric after failing at 2 psi normal stress. The apparent adhesion of the
fabric structure is overcome during failing by breaking or distorting individual elements of the
split-film weave. Additionally the weakened fabric parted outside the area of the normal stress
on the lower traveling container. This parting began at the edges and progressed to the center.
It was more pronounced at higher normal stress levels. Test Group I did not encounter the same
amount of structural damage. The average residual to peak shear ratio was determined to be
70%.
Both Test Group I and II did evidence measurable elongation, 4 to 8%, as depicted on
the normalized horizontal displacement versus shear stress plots. This was expected based on
the physical properties of the geotextile. This geotextile elongates prior to shear failure but did
not fail in tension. This relates back to the very high 8 angle per the linear analysis of 34.4. The
ft angle for the 2 to 7 psi normal stress range is 41. This indicates that the machine direction
would be most suited for the cross sandbag direction. The opposite of the manufacturers current
61


orientation.
Test Group III is the cross machine direction tests. Initial observations were confirmed
by the direct shear testing. The effective A angle for this direction was 16.4 or less than half that
of the machine direction. The testing proceeded as if the fabric were lubricated when compared
to the machine direction. There was no fabric structural distortion. Test Group IV, the residual
at 13 psi normal stress of the specimen failed at 13 psi normal stress, tested a residual shear
stress value of 96% of the peak value.
In summary this geotextile appears to have superior material properties for utilization as
sandbag fabric if the sandbags take advantage of the higher shear resistance inherent in the
machine direction. The peak shear values at the three normal stresses are compared below:
Normal Peak Shear Stress (PSI)
Stress Test Group Number
PSI I II III
2 3.60 2.52 0.52
7 8.10 5.80 1.91
13 11.20 7.80 3.75
6 34.4 25.4 16.4
ca (PSI) 2.61 1.89 -0.096
Table 3.6 Peak Shear Stress & Linear Comparison
62


3.2.5 Woven Natural Jute
The woven natural jute geotextile was obtained from Langston Bags, a regular supplier
of sandbags to the Corps of Engineers. No physical property sheet was available for this
geotextile. The testing included both the cross bag and long bag directions. The cross bag
being the critical direction.
Test Group Test Type Peak/Residual Testing Direction Normal Stresses PSI
V Peak Cross Bag 2, 7, & 13
VI Peak Long Bag 2 & 4
Table 3.7 Test Conditions
The analysis was the same as that of Test Group I of Section 3.2.4. The raw data is presented
in Appendix C. The results summarize as:
Normal Peak Shear Stress (PSI)
Stress Test Group
PSI V VI
2 1.86 1.45
4 2.2
7 3.70 _
13 5.10 -
6 16.3 20.56
c. (PSI) 1.41 0.70
Table 3.8 Peak Shear Stress & Linear Comparison
The direct shear test results plot and comparative spreadsheet follow.
63


Figure 3.15 Direct Shear Test Results
o>


COMPARATIVE FIT ANALYSIS
NEW NATURAL JUTE MATERIAL
CROSS BAG DIRECTION
NORMAL CALCULAT ED SHEAR STRESS RAW DATA
STRESS LINEAR HYPERBOUCI PARABOLIC SHEAR
PSI PSI PSI 1 PSI PSI
0.000 1.4081 0.0000 0.1555
1.000 1.7007 1.0624 1.2569
2.000 1.9932 1.8600 1.8588 1.860
3.000 2.2857 2.4809 2.3272
4.000 2.5782 2.9779 2.7244
5.000 2.8708 3.3848 3.0754
6.000 3.1633 3.7240 3.3934
7.000 3.4558 4.0111 3.6862 3.700
8.000 3.7484 4.2573 3.9591
9.000 4.0409 4.4707 4.2155
10.000 4.3334 4.6575 4.4582
11.000 4.6259 4.8223 4.6892
12.000 4.9185 4.9689 4.9100
13.000 5.2110 5.1000 5.1218 5.100
14.000 5.5035 5.2180 5.3257
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
a LINEAR HYPER.- PARB. x RAW DATA
Table 3.9 Comparative Spreadsheet
65


Figure 3.16 Direct Shear Results
o>
O
DIRECT SHEAR TEST RESULTS NATURAL JUTE SANDBAG MATERIAL
0%

2.70- LONG B AG DIRECTION
2.60-
Z

A i
s. \ \ /
**' \
il V v
,v* *
i ml
yj J
A f* A
eC V' V
iti 7n_ t'
:

1 .LW ! s'

U.OU
1 r


!


>
0.01 Wo 1.01 Wo 2.0 )% 3.0i )% 4.0 )% 5.0 NC )% 6.0 IRMALi; )% 7.0 LED HOI Wo 8.0 fUZONT/ Wo 9.0 \L DISPI Wo 10.C .ACEME 0% ll.C NT 0% 12.( 0% 13.( 10% 14.C 0% 15.(
2 PSI 4 PSI



COMPARATIVE FIT ANALYSIS
NEW NATURAL JUTE MATERIAL
LONG BAG DIRECTION
NORMAL CALCULAT ED SHEAR STRESS RAW DATA
STRESS LINEAR HYPERBOLIC I PARABOLIC SHEAR
PSI PSI PSI I PSI PSI
0.000 0.7000 0.0000 0.0862
1.000 1.0750 0.8622 0.9808
2.000 1.4500 1.4500 1.4697 1.450
3.000 1.8250 1.8765 1.8502
4.000 2.2000 2.2000 2.1728 2.200
5.000 2.5750 2.4538 2.4579
6.000 2.9500 2.6583 2.7162
7.000 3.3250 2.8266 2.9540
8.000 3.7000 2.9674 3.1756
9.000 4.0750 3.0871 3.3839
10.000 4.4500 3.1900 3.5811
11.000 4.8250 3.2794 3.7687
12.000 5.2000 3.3579 3.9480
13.000 5.5750 3.4273 4.1200
14.000 5.9500 3.4891 4.2857
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
s LINEAR HYPER. PARB. RAW DATA
Table 3.10 Comparative Spreadsheet
67


This geotextile fits best with the parabolic curve fit analysis in both bag directions. The
A angle from the linear regression analysis was 16.3 in the cross bag direction and 20.5 in the
long bag direction. However, it should be noted that the long bag direction had a normal stress
range of 2 to 4 psi and the cross bag direction a range of 2 to 13 psi. The 2 to 7 psi A angle for
the cross bag direction is 20.2. The parabolic equation for the cross bag direction is:
(tn + 0.320)2 = 4 0.565 (on + 0.100)
and the equation for the long bag direction is:
(t + 0.300)2 = 4 0.373 (on + 0.100)
Inspection of the tested fabric subsequent to the 7 psi and 13 psi normal compressive
stress tests indicated extensive fabric failure on the lower traveling container outside of the test
area. This is interpreted as exceeding the tensile strength of the natural jute fiber prior to the
failure by shear. In a practical sense this indicates a real limit on this fabric's ability to withstand
shearing forces in the field. This became evident during testing by the popping sound of the
individual fibers failing. This is depicted on the normalized horizontal displacement versus shear
stress plots in the erratic shape of the curve even after it was smoothed using an averaging
technique. The analysis spreadsheets are presented in Appendix D.
Because of this structural destruction in the cross bag direction, the long bag direction
tests were limited to 2 and 4 psi normal stresses. This is not a fatal limit on the use of this fabric
for sandbags as most sandbag structures do not exceed the 4 to 7 psi normal stress range.
68


3.2.6 White Woven Split-Film Polypropylene
This geotextile was obtained in sandbag form from the Corps of Engineers office at
Vicksburg, Mississippi. It was manufactured in the Philippines and is much thinner and lighter
than the black woven geotextile discussed in Section 3.2.4. Due to limited fabric, this was tested
only in the cross bag direction and only at normal stresses of 2 and 7 psi. The analysis was the
same as in Section 3.2.4. The plot of normalized horizontal displacement versus shear stress,
Figure 3.17, depicts elongation of 2 to 3%. The comparative spreadsheet is presented as Table
3.6. Individual analysis spreadsheets are in Appendix D.
Having only two points a reasonable fit can be attained by all three analysis methods.
However, considering the experience with the black woven polypropylene, the hyperbolic fit was
selected for use in Chapter 5. Stability Analysis. The hyperbolic equation becomes:
T.----------.
" (1.675+0.088*0^
The linear analysis defined a A angle of 21.5 and an adhesion of 0.292 psi. This is well
below the machine direction value for Section 3.2.4 of 41 for the same normal stress range. The
fabric showed no significant distortion after testing. There was the same level of elongation,2
to 4%, experienced in the prior tests of other geotextiles.
The Corps of Engineers indicated this is a preferred material, in many situations, for the
sandbags because of the price as compared to other fabrics. They maintain a large supply of
these sandbags which do not rot or become rodent food as the natural jute sandbags often do.
69


Figure 3.17 Direct Shear Test Results
~>l
o
DIRECT SHEAR TEST RESULTS WHITE PP MATERIAL FROM CORPS of ENG.
0%





y .

^.IaJ *
Om Z.4U-

N*.
< /
a t 60~ (A ;
i./
j

/
uuu r"


0.0 Wo 1.0 Wo 2.0 Wo 3.0 )% 4.0 Wo 5.0 NC )% 6.0 IRMALU Wo 7.0 £ED HO Wo 8.0 RIZONTy Wo 9.0 \L DISP1 Wo 10.( .ACEME 0% ll.C NT 0% 12.C 0% 13.C 0% 14.C 0% 15.(
2 PSI 7 PSI



COMPARATIVE FIT ANALYSIS
NEW WHITE POLYPROPYLENE MATERIAL
CROSS BAG DIRECTION
NORMAL CALCULATED SHEAR STRESS RAW DATA
STRESS DREAR HYPERBOLIC PARABOLIC SHEAR
PSI PSI PSI PSI PSI
0.000 0.2920 0.0000 -1.2000
1.000 0.6860 0.5671 0.4093
2.000 1.0800 1.0800 1.0759 1.080
3.000 1.4740 1.5460 1.5874
4.000 1.8680 1.9713 2.0186
5.000 2.2620 2.3610 2.3985
6.000 2.6560 2.7193 2.7419
7.000 3.0500 3.0500 3.0578 3.050
8.000 3.4440 3.3561 3.3517
9.000 3.8380 3.6402 3.6278
10.000 4.2320 3.9047 3.8890
11.000 4.6260 4.1515 4.1374
12.000 5.0200 4.3823 4.3747
13.000 5.4140 4.5986 4.6023
14.000 5.8080 4.8017 4.8214
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
a LINEAR ---HYPER. A PARB. x RAW DATA
Table 3.11 Comparative Spreadsheet
71


3.2.7 Nonwoven Needle-Punched Geotextile
This geotextile was selected for testing based on two properties as a potential alternative
sandbag material. First, it was anticipated by the supplier to have a high coefficient of friction
and second its AOS, 0.212 mm, would allow cement to be pumped into and retained in a
sandbag made of this material. This is not a current sandbag geotextile. The geotextile was
supplied by Nicolon of Norcross, Georgia, and is identified as S800 Nonwoven. It is a nonwoven
polypropylene needle-punched fabric. It is stable within a pH range of 2 to 13. It is non-
biodegradable and resistant to commonly found soil chemicals. The physical properties are
presented in Appendix A and were discussed in Section 3.1.
This geotextile was tested in both the machine and cross machine directions and for
peak and residual shear values. The test groups were:
TEST GROUP TEST TYPE PEAK/RESIDUAL TESTING DIRECTION NORMAL STRESS (PSI)
A PEAK MACHINE 2, 7, & 13
B RESIDUAL 2 PSI MACHINE 2, 7, & 13
C RESIDUAL 13 PSI MACHINE 2, 7, & 13
D PEAK CROSS MACHINE 4& 13
Table 3.12 Test Conditions
The analysis of the recorded data was conducted identically to the sections above. The
test group plots of normalized horizontal displacement versus shear stress, and the comparative
spreadsheets are on the following pages. Numerical analysis spreadsheets are in Appendix D.
72


Figure 3.18 Direct Shear Test Results
-N|
Ca)
DIRECT SHEAR TEST RESULTS NON-WOVEN NEEDLE PUNCHED (8oz/yd)
1 0%
MACHINE DIRECTION L
' *'

*>,N
s '''' *
O' y y
SI =
vL \ /

yj Z.5U S 1 *

i i i
1 / i S
i } ^ "V ?
/
o.o< Wo 1.0 Wo 2.0 Wo 3.0 )% 4.0 Wo 5.0 Wo 6.0 NORM/ Wo 7.0 iLIZED )% 8.0 HORIZC 3% 9.0 3NTALI )% 10.C 3ISPLAC 0% U.( CEMENT 0% 12.C 0% 13.C 0% 14.C 0% 15.( 0% 16.C
2 PS1 7 PSI 13PSI



COMPARATIVE FIT ANALYSIS
NEW NONWOVEN NEEDLE-PUNCHED
____ MACHINE DIRECTION
NORMAL CALCULAT ED SHEAR STR ESS RAW DATA
STRESS UNEAR HYPERBOLIC PARABOLIC SHEAR
PSI PSI PSI PSI PSI
0.000 0.2215 0.0000 -0.8791
1.000 0.5977 0.5577 0.0813
2.000 0.9738 1.0800 0.7936 1.080
3.000 1.3500 1.5700 1.3864
4.000 1.7262 2.0308 1.9051
5.000 2.1023 2.4647 2.3720
6.000 2.4785 2.8742 2.8002
7.000 2.8546 3.2612 3.1979 2.660
8.000 3.2308 3.6275 3.5707
9.000 3.6069 3.9747 3.9230
10.000 3.9831 4.3043 4.2577
11.000 4.3592 4.6177 4.5772
12.000 4.7354 4.9159 4.8834
13.000 5.1115 5.2000 5.1779 5.200
14.000 5.4877 5.4711 5.4619
cn
CL
CO
CO
LU
oc
\
CO
cc
<
LU
X
CO
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
.INEAR HYPER. A PARB. RAW DATA
Table 3.13 Comparative Spreadsheet
74


Figure 3.19 Direct Shear Test Results
cn
DIRECT SHEAR TEST RESULTS NON-WOVEN NEEDLE PUNCHED (8oz/yd)
...... | 1 1 1 2 PSI RESIDUAL IN MACHINE DIRECTION 1 0%
/' / / s z"'

* /
So I w
$ 25

< w X (A
/ r l i i

/
/
0.0 3% 1.0 3% 2.0 3% 3.0 3% 4.0 3% 5.0 NC 3% 6.0 3RMALL 2PSI 3% 7.0 ZED HOI - 7 F 3% 8.0 RIZONTj si 3% 9.0 \L DISPI 13PSI 3% 10.( -ACEME 0% ll.C NT 0% tZ( 0% 13.C 0% 14.C 0% 15.C


COMPARATIVE FIT ANALYSIS
NEW NONWOVEN NEEDLE-PUNCHED
RESIDUAL 2 PSI MACHINE DIRECTION
NORMAL STRESS PSI CALCULAT ED SHEAR STR ESS RAW DATA SHEAR PSI
LINEAR PSI HYPERBOLIC PSI PARABOLIC PSI

0.000 0.2846 0.0000 0.0131
1.000 0.5931 0.4768 0.4556
2.000 0.9015 0.9200 0.8636 0.920
3.000 1.2100 1.3330 1.2441
4.000 1.5185 1.7189 1.6021
5.000 1.8269 2.0802 1.9410
6.000 2.1354 2.4191 2.2638
7.000 2.4438 2.7378 2.5724 2.410
8.000 2.7523 3.0379 2.8686
9.000 3.0608 3.3211 3.1537
10.000 3.3692 3.5887 3.4290
11.000 3.6777 3.8420 3.6954
12.000 3.9862 4.0821 3.9537
13.000 4.2946 4.3100 4.2046 4.310
14.000 4.6031 4.5266 4.4486
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
JNEAR HYPER. A PARB. RAW DATA
Table 3.14 Comparative Spreadsheet
76


Figure 3.20 Direct Shear Test Results
-n|
DIRECT SHEAR TEST RESULTS NON-WOVEN NEEDLE PUNCHED (8oz/yd)
13 PSI RESIDUAL IN MAfHINP niDFmON 1 0%
/ ....

J.jU / f'
55 I
K z- _
< w X i : t

j
I
0.0 Wo 1.0 )% 2.0 Wo 3.0 )% 4.0 )% 5.0 NC Wo 6.0 1RMALL 2PSI Wo 7.0 ZED HO 7 p Wo 8.0 RIZONT; SI Wo 9.0 VL DISPI 13PSI Wo 10.< .ACEME 10% ll.C NT 0% 12.C 0% 13.C 0% 14.( 0% 15.(


COMPARATIVE FIT ANALYSIS
NONWOVEN NEEDLE-PUNCHED
[RESIDUAL 13 PSI MACHINE DIRECTION
NORMAL CALCULATED SHEAR STRESS RAW DATA
STRESS LINEAR I HYPERBOLIC PARABOLIC SHEAR
PSI PSI I PSI PSI PSI
0.000 0.2813 0.0000 -0.0251
1.000 0.5816 0.4667 0.4138
2.000 0.8819 0.9000 0.8186 0.900
3.000 1.1821 1.3034 1.1961
4.000 1.4824 1.6800 1.5512
5.000 1.7827 2.0323 1.8874
6.000 2.0830 2.3625 2.2076
7.000 2.3832 2.6727 2.5137 2.350
8.000 2.6835 2.9647 2.8076
9.000 2.9838 3.2400 3.0905
10.000 3.2841 3.5000 3.3636
11.000 3.5843 3.7459 3.6278
12.000 3.8846 3.9789 3.8841
13.000 4.1849 4.2000 4.1329 4.200
14.000 4.4852 4.4100 4.3751
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
UNEAR HYPER. A PARB. x RAW DATA
Table 3.15 Comparative Spreadsheet
78


Figure 3.21 Direct Shear Test Results
^1
CO
DIRECT SHEAR TEST RESULTS NON-WOVEN NEEDLE PUNCHED (8 oz/yd)
CR OSS MA CHINE I DIRECT] [ON J 0%
.... *


55 ^ - y'''
1 3': fe = t i
CC Z.5U- < : u s /
-
1.50-


0.0 Wo 1.0 Wo 2.0 Wo 3.0 Wo 4.0 Wo 5.0 Wo 6.0 NORM/ Wo 7.0 iLIZED 4 P Wo 8.0 HORIZC SI - Wo 9.0 INTALI 13PSI Wo 10.( JISPLAC 0% ll.C :ement 0% 12.C 0% 13.C 0% 14.C 0% 15.C 0% 16.(


COMPARATIVE FIT ANALYSIS
NEW NONWOVEN NEEDLE-PUNCHED
CROSS MACHINE DIRECTION
NORMAL STRESS PSI CALCULAT ED SHEAR STR ESS RAW DATA SHEAR PSI
LINEAR PSI HYPERBOLIC PSI PARABOLIC PSI

0.000 0.3722 0.0000 -0.9503
1.000 0.7167 0.4642 -0.0254
2.000 1.0611 0.9099 0.6605
3.000 1.4056 1.3382 1.2314
4.000 1.7500 1.7500 1.7309 1.750
5.000 2.0944 2.1463 2.1806
6.000 2.4389 2.5280 2.5929
7.000 2.7833 2.8958 2.9758
8.000 3.1278 3.2505 3.3349
9.000 3.4722 3.5928 3.6741
10.000 3.8167 3.9233 3.9964
11.000 4.1611 4.2427 4.3040
12.000 4.5056 4.5514 4.5989
13.000 4.8500 4.8500 4.8825 4.850
14.000 5.1944 5.1390 5.1560
CO
£L
CO
CO
LD
CC
t
CO
QC
<
LL1
X
CO
DIRECT SHEAR TEST RESULTS
NORMAL VS. SHEAR STRESS
NORMAL STRESS (PSI)
JNEAR HYPER. A PARB. RAW DATA
Table 3.16 Comparative Spreadsheet
80


The analysis indicated that the best fit was accomplished with the linear analysis. The
defined 8 angles and adhesion values are:
TEST GROUP PEAK/ RESIDUAL TEST DIRECTION 6 ANGLE ADHESION (PSI) RESIDUAL TO PEAK RATIO
A PEAK MACHINE 20.6 0.221 N/A
B RESIDUAL MACHINE 17.1 0.285 0.862
C RESIDUAL MACHINE 16.7 0.281 0.841
D PEAK CROSS 19.0 0.372 N/A
Table 3.17 Test Summary
As anticipated the material properties did not vary significantly between the machine and
cross machine directions. The exception was the elongation in the cross machine direction
which exceeded that in the machine direction by about 4% as depicted on the respective plots
of normalized horizontal displacement versus shear stress.
The equation for this geotextile to be used in Section 5. Stability Analysis is:
t = 0.221 + on Tan (20.6)
This is the Mohr-Coulomb Criteria which define the failure envelope conventionally for soils.
Neither the hyperbolic nor parabolic analysis fit the measured data as demonstrated on the
respective spreadsheets in Appendix D.
This geotextile is an apparent alternative to the conventional sandbag geotextiles. It
81


does have two possible drawbacks. First, it is much bulkier in the 8 ounce/yard variety. This
could be offset somewhat by utilizing a lighter option of 4 or 6 ounce/yard. The second is price
which will be discussed in Section 6.4.
The major expected benefit was in the A angle and potentially the adhesion coefficient.
This did not materialize to the extent expected. The black woven split-film polypropylene, in the
machine direction, exceeds this geotextile by 67% with respect to the A angle. Neither has a
significant adhesion coefficient.
3.2.8 Summary
All four of the geotextiles tested have proved worthy to be included in the design and
analysis of Chapter 5 and Chapter 6 to follow. It is noted that the A angles, though not to be
used in all analyses, are convenient to compare the geotextiles one to another and will be utilized
as such for the remainder of this thesis. A comparison of linear parameters in the most effective
fabric direction for the four geotextiles is as follows:
Geotextile Direction A Angle c. (PSI)
Black Woven Polypropylene Machine 34.4 2.61
Natural Jute Cross Bag 16.3 1.41
White Woven Polypropylene Cross Bag 21.5 0.292
Nonwoven Needle-Punched Machine 20.6 0.221
Table 3.18 Linear Parameter Comparison
82


individual sandbags are filled will impact the structure weight to a greater extent.
The gradation, distribution of sand size, used as fill material will vary with the bulk supply
and can not be preset except by general guidelines. The impact of using a poorly or well sorted
sand should not affect the objectives of this thesis as long as the geotextile used for the sandbag
is capable of containing the fill material, ie. keeping leaching to a minimum. The test sand used
was well sorted containing only a small percentage fines.
4.2 Pore Pressure Distribution
Sandbag structures are, in flood abatement, designed to be nearly submerged on the
waterside. Corps of Engineers manuals recommend that a minimum freeboard of approximately
six inches be designed. The actual designs are based on predicted storm events and anticipated
runoff amounts, ie. no factual data is available. The most conservative case is for the water to
reach the total height of the sandbag structure. This is the case utilized in the remainder of the
thesis.
Assuming the water is contained level with the sandbag structure, how much of the
sandbag cross-sectional area is saturated? This depends on the fill material and its permeability
and the geotextiles ability to transmit water. Using sand as described in Section 4.1 as fill
material and the geotextiles tested in this thesis, it was determined that most of the cross-
sectional area would be saturated. This was determined using two papers written by Professor
88


Yang H. Huang1819. The first paper is based on original work by Schaffernak and Iterson and
described by Casagrande20 "as applicable to an outslope not over 30". Mr. Huang expanded
this work to include inclined ledges. The paper assumes that the seepage begins at the pool
level at a distance of "0.3 A from the dam, in which A = the horizontal distance shown in Fig. 1."
Mr. Huangs Fig. 1 is presented below depicting the parameters required to predict the phreatic
surface within the dam.
FIG. 1.Earth Oam on Inclinad Ledge
Figure 4.1
Earth Dam on Inclined Ledge
(Huang 1981)
Mr. Huangs procedure identifies three points on the phreatic surface through the dam
using a graphical solution and the basic dam dimensions. Permeability is not a parameter in this
18 Huang, Y. H. (May 1981), "Lines of Seepage in Earth Dams on Inclined Ledge, Journal
of Geotechnical Engineering.Vol. 107, No. GT5
19 Huang, Y. H. (January 1986), "Unsteady State Phreatic Surface in Earth Dams", Journal
of Geotechnical Engineering. Vol. 112, No. 1,
20 Prof., Department of Civil Engineering, University of Kentucky, College of Engineering,
Lexington, Ky.
89


solution method. The estimated phreatic surface for a sandbag structure of base-to-height ratio
2.5:1 and an a angle of zero was determined and is presented below.
Figure 4.2
Phreatic Surface of Sandbag Structure
The predicted phreatic surface indicates that over 98% of the cross-sectional area is
saturated, ie. belowthe phreatic surface. This would be consistent with the findings of the Corps
of Engineers discussed in Section 2.1 of this thesis.
The second paper of Mr. Huang is an extension of Cedergrens21 work using transient
flownets to define the phreatic surface in an earth dam. The phreatic surface's upper end is fixed
to the pool elevation at the upstream end and declines to the impervious base toward the
downstream end of the dam. This is depicted below with an example from Mr. Huang's paper.
This solution uses dimensionless time and distance in a graphical procedure to define the
21 Prof., Department of Civil Engineering, University of Kentucky, College of Engineering,
Lexington, Ky.
90


phreatic surface with respect to time.
(Huang 1986)
The permeability of the sand, 3.3x1 O'1 to 3.3x1 O'5 [ft/sec], results in the phreatic surfaces
lower end location being over 400 feet from the toe of the structure, an unrealistic situation. Even
though the procedure did not provide descriptive results, it did indicate that the total cross-section
would be saturated. This confirmed the first approximation and saturated conditions will be used
in the analyses of Chapter 5.
The assumption of a submerged structure requires the definition of the pore pressure
distribution along the base of the structure and along the interface between layers of sandbags.
This was needed for the development of the models in Chapter 5. The definition was made
using two approaches. First a flownet was constructed for a dimensionless cross-section with
a base-to-height ratio of 2.5 to 1. This is presented as Figure 4.4. From this flownet the pore
distribution along any horizontal surface was defined as a function of the structure height
91


assuming the water depth is equal to structure height.
This was plotted as normalized, dimensionless, values and is presented as Figure 4.5.
These values were then utilized to calculate the weighted average normalized values for the
water side half, the land side half, and the total structure. This is presented as Table 4.3. The
total structure value of 0.668 [ft/ft] is the pore pressure in feet divided by the height of the
structure in feet. This is referred to throughout the thesis as the "pore pressure factor". The
normalized horizontal distance is the distance from the water side heal of the structure in feet
divided by the base width of the structure in feet. This approach allows the pore pressure to be
defined for any structural height and any interface between sandbag layers.
The above approach is valid only for a base-to-height ratio of 2.5:1.0. The goal of this
thesis is to define alternative configurations which better utilize the resources available. This
required repeating the approach for other base-to-height ratios to define a relationship between
92


PORE PRESSURE ANALYSIS
BASE OF SANDBAG STRUCTURE
noAmAliZE) NORMALIZED
HORIZONTAL PORE
DISTANCE PRESSURE
[FT/FTl [FT/FTl
0.000 1.000
0.147 0.969
0.333 0.899
0.433 0.809
0.525 0.719
0.607 0.629
0.680 0.539
0.742 0.449
0.793 0.360
0.850 0.270
0.907 0.160
0.957 0.090
1.000 0.000
I
S"
O
* 0.CD
NORMALIZED PORE PRESSURE
AT SANDBAG STRUCTURE BASE
NORUALIZEO HORIZONTAL DISTANCE [FT/FTJ
Figure 4.5
Pore Pressure Analysis

CALCULATION OF WEIGHTED AVERAGE
NORMALIZED PORE PRESSURE |

NORMALIZED NORMALIZED WEIGHTED AVERAGE
HORIZONTAL PORE NORMALIZED
DISTANCE PRESSURE PORE PRESSURE
[FT/FTl [FT/FTl [FT/FT]
0.00 1.00 0.100
0.10 0.99 0.098
0.20 0.97 0.095
0.30 0.92 0.069 UPSTREAM SIDE ll
0.40 0.65 0.060 0.921
0.50 0.75 0.070
0.60 0.65 0.056
0.70 0.51 0.044
0.80 0.36 0.027 I DOWNSTREAM SIDE ll
0.90 0.16 0.009 0.415 I
1.00 0.00 [ TOTAL
0.668 I 0.668
Table 4.3 Weighted Average Pore Pressure
93


the "pore pressure factor" and the base-to-height ratio. This was done for base-to-height ratios
of 2.0:1.0 and 1.5:1.0, which were expected to cover the reasonable range of alternative
structural configurations. The flownet for the 2.0:1.0 ratio is presented as Figure 4.6 and the
weighted average pore pressure factor calculation is Table 4.4. The flownet for the 1.5:1.0 ratio
is Figure 4.7 and the pore pressure factor is calculated in Table 4.5.
The pore pressure values compare closely with that of the 2.5:1.0 base-to-height ratio
as do the normalized pore pressure versus normalized horizontal distance plots, Figure4.8. The
numerical relation between pore pressure factor and base-to-height ratio was determined
graphically and found to best fit a semi-logarithmic relationship. The linear plot revealed a near
fit but had a minor curve to the line. The plot of pore pressure factor versus Log(base-to-height)
defined a straight line as depicted in Figure 4.9. This was resolved into a linear equation for
convenience and later use in Chapter 5. The relation became:
PPF = pore pressure factor
PPF = 0.226 LOG(BTHF) + 0.578 BTHR = base-to-height ratio
This relationship is used in Chapter 5. to calculate the pore pressure force for all interfaces
between sandbags and that of the base of the structure.
The second approach for defining the pore pressure distribution was using SEEP22, a
22 "SEEP:A Computer Program for Seepage Analysis of Saturated Free Surface or
Confined Steady State Flow", Department of Civil Engineering, Virginia Tech,
Blacksburg, Va.
94


Figure 4.6 Flownet for 2.0:1.0 Structure
CALCULATION OF WEIGHTED AVERAGE
NORMALIZED PORE PRESSURE
NORMALIZED NORMAUZED
HORIZONTAL PORE
DISTANCE PRESSURE
[FT/FTI fFT/FTI
WEIGHTED AVERAGE
NORMALIZED
PORE PRESSURE
fFT/FTI
0.00 1.00
0.10 0.97
0.20 0.94
0.30 0.89
0.40 0.B1
0.50 0.71
0.60 0.61
0.70 0.50
o.eo 0.36
0.90 0.19
1.00 0.00
0.099
0.096
0.092
0.065 UPSTREAM SIDE l
0.076 0.8931|
0.066
0.055
0.043
0.027 DOWNSTREAM 8IDE I
0.010 0.403 II
TOTAL
0.648 I 0.6481
Table 4.4 Weighted Average Pore Pressure (2.0:1.0)
95


Figure 4.7 Flownet for 1.5:1.0 Structure

CALCULATION OF WEIGHTED AVERAGE 1 NORMALIZED PORE PRESSURE |

NORMALIZED HORIZONTAL DISTANCE [FT/FT] NORMALIZED PORE PRESSURE fFT/FTl WEIGHTED AVERAGE NORMAUZED PORE PRESSURE [FT/FTI

0.00 1.00 0.098
0.10 0.96 0.094
0.20 0.92 0.090
0.30 0.87 0.083 UPSTREAM SIDE I
0.40 0.78 0.073 0.874
0.50 0.68 0.062
0.60 0.56 0.050
0.70 0.44 0.037
0.80 0.30 0.023 DOWNSTREAM SIDE ll
0.90 0.17 0.009 0.362
1.00 0.00 TOTAL
0.618 0.618
Table 4.5 Weighted Average Pore Pressure (1.5:1.0)
96


PORE PRESSURE FACTOR COMPARISON
NORMALIZED VALUES
NORMALIZED HORIZONTAL DISTANCE [FT/FT]
2.5:1.0 2.0:1.0 -*-1.5:1.0
Figure 4.8 Pore Pressure Comparison
NORMALIZED PORE PRESSURE vs. LOG(BTHR)
ij 50
gj 0-OJ OS £ as o- aj
as -60_ O On D as as N
as O 2
0-5 0. oo' ' 'o. 05 0. 10 0. 15 0. 20 0. LOG(l 25 0. 3THR) 30 0. 35 0. 40 0. 45 0.
Figure 4.9 Normalized Pore Pressure Factor
97


commercial finite element analysis program. This program solves for "steady state problems of
free surface or confined flow of groundwater, in two dimensional or axisymmetric porous
regions. The two dimensional free surface flow solution was utilized to estimate the pore
pressure distribution. This program was used for the base-to-height ratio of 2.5:1.0 as a
confirmation of the manually constructed flownets. The layout for the finite element analysis is
depicted in Figure 4.10. A total of 26 nodes and 20 elements were utilized. The results are
presented in Table 4.6.
The results of the SEEP analysis is compared to the flownet results for the 2.5:1.0 base-
to-height ratio in Figure 4.11. As is depicted, the results compare most favorably. This was
taken as confirmation of the flownet analysis. The flownet relations are used in the remainder
of the thesis.
/
17 18
Figure 4.10 SEEP Layout (2.5:1.0)
98


Nodal Point Normalized Horizontal Distance [FT/FT] Normalized Pore Pressure SEEP Analysis [FT/FT] Normalized Pore Pressure Flownet Analysis [FT/FTI
1 0.0 1.000 1.00
6 0.2 0.971 0.97
11 0.4 0.859 0.85
16 0.6 0.654 0.65
21 0.8 0.360 0.36
26 1.0 0.000 0.00
Table 4.6 SEEP Analysis Results
NORMALIZED PORE PRESSURE SEEP vs. FLOWNET
DO

FT 0.909

2
sn 0.70 a 0.65 2 0.60 0.55 y 0.50


s.
s

O '4S v
Q 0.35 S3 0.30 N 0.25 J 0.20 2 0.15
V

s,
'k
S
5 0.10! o.osi 2 o.ool z 0
1 \
00 0. D5 0. 10 0. 15 0. N 20 0. IOR 25' 0. MAI 30 0. _1ZE 35 0. .D H *0 0. OR 45 0. IZO 50 0. NT A 1 1 1 1 55 0. ID 60 0. I ST/ 65 0. KSC 1111 70 0. E [F 75 0. T/Fl B0 0. l 85 0. 90 0. 951.
SEEP FLOWNET

Figure 4.11 Comparison SEEP vs. Flownet Results (2.5:1.0)
99


5. Stability Analysis
Sandbag structural stability analyses were conducted using spreadsheet based models
for potential of "sliding", "overturning", and "slumping". Additionally the minimum required
bearing capacity of the foundation material was determined also utilizing a spreadsheet based
model. The next sections present the development of and the basis for the models.
The critical condition or boundary for sandbag structures is when the water height equals
the structure height and the structure is saturated. This is not an atypical situation per the
references found. In fact, the structures are designed with very limited freeboard, ie. about six
inches. The models were developed to accommodate variable water heights for any given
structural height, but the analyses were run at the two being equal.
The foundation material for any structure is as variable as the number of surfaces in a
given area and is not usually a discretionary item. The structure's location is a function of the
location of the flood water and what is to be protected. What ever the foundation material there
is the foundation material that will be built on and the structure can not be moved to better
foundation material. This variability of foundation material and lack of alternative placement
potential placed the bearing capacity factor of safety outside the scope of the thesis. The
100


minimum required bearing capacity was determined and is presented for guidance.
5.1 Development of Analysis Models
The forces acting on the sandbag structure, regardless of height or base-to-height ratio,
are depicted below in Figure 5.1.
models. They are:
W = effective weight of the sandbag structure = (ymI yj Vol-ndbig8
Ww1 = weight of water in sandbag structure = yw Volniriitir.on
Ww2 = weight of water on water side of structure = Vi yw H B
Fw = static water force = Vi yw H2
Fvel = force due to water velocity = V2 y H / (2 g)
U = pore pressure at base (see Section 4.2) = (Pore Pressure Factor) H B yw
101


F = foundation resisting force = W + Ww2 + Ww1 U
T = resisting shear force = /{(frictional behavior), (Base Area)}
It can be noted from the above that there are few properties of the structure that control
the stability of the structure. This simplifies the model development and data procurement
greatly. Including variable water height only adds one more variable and adds only minor
changes to the models.
5.1.1 Factor of Safety Sliding
The factor of safety for a sandbag structure's potential to slide horizontally is defined by
the equation:
_ ResistingForce
SUD DrivingForce
Resisting Force = T Area
Driving Force = Fw + Fv#l
The resisting force is a function of the sandbag structure effective weight, W', the weight of the
water within the structure, Ww1, the weight of the water acting on the water side of the structure,
Ww2, the pore pressure, U, and the frictional behavior of the geotextile to itself or the foundation
material. These properties are calculated using the sandbag structure dimensions, including the
dimensions of the unit sandbag, the sand index properties, the water height, and the frictional
behavior relations, as determined in Chapter 3., and the fixed parameter of water density.
The sandbag dimensions of height and base width are the only parameters required for
the calculations and they can be related as "base-to-height ratio". The standard base-to-height
102