Citation
The performance of a curvilinear versus a rectangular basement foundation design in expansive clay soils

Material Information

Title:
The performance of a curvilinear versus a rectangular basement foundation design in expansive clay soils
Creator:
Gardiner, Michael James
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic resource : ;

Subjects

Subjects / Keywords:
Foundations -- Design and construction ( lcsh )
Basements -- Design and construction ( lcsh )
Clay soils ( lcsh )
Basements -- Design and construction ( fast )
Clay soils ( fast )
Foundations -- Design and construction ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Shallow foundation design in expansive soils has generally been approached in the industry using a typical pier/beam or spread footing/foundation wall and reinforcement design. The addition of supporting piers anchored in stable soils and excavated expansion areas under beam elements have modified a traditional foundation design for expansive soils. This traditional approach to foundation design uses designs that mitigate around the swelling effects, rather than designing to take advantage of, or resist the imposed forces. This research defines a new shallow foundation design that uses a curvilinear structure to take advantage of the forces exerted on the foundation by the expansive forces of the soil. In addition, the design allows for a cast-in-place or precast implementation. The curvilinear foundation design was modeled using LS DYNA Finite Element analysis and compared to a traditional rectangular foundation design using 3D models. In addition, the soil and concrete models were evaluated by comparing the LS DYNA model results (deflection, shear, tensile/compression) for both foundation designs.
Thesis:
Thesis (M.S.)--University of Colorado Denver. Civil engineering
Bibliography:
Includes bibliographical references.
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Michael James Gardiner.

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:
862977767 ( OCLC )
ocn862977767

Downloads

This item has the following downloads:


Full Text
THE PERFORMANCE OF A CURVILINEAR VERSUS A RECTANGULAR
BASEMENT FOUNDATION DESIGN IN EXPANSIVE CLAY SOILS
by
Michael James Gardiner
B.A., The Colorado College, 1985
B.S., University of Colorado, Denver, 2001
A thesis submitted to the
Faculty of the Graduate School of the
University of Colorado in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
2013


This thesis for the Master of Science degree by
Michael James Gardiner
has been approved for the
Civil Engineering Program
by
Nien-Yin Chang, Chair
Brian T. Brady
Yail Jimmy Kim
April 10, 2013


Gardiner, Michael James (Master of Science, Civil Engineering)
The Performance of a Curvilinear Versus a Rectangular Basement Foundation
Design in Expansive Clay Soils
Thesis directed by Professor Nien-Yin Chang
ABSTRACT
Shallow foundation design in expansive soils has generally been
approached in the industry using a typical pier/beam or spread footing/foundation
wall and reinforcement design. The addition of supporting piers anchored in
stable soils and excavated expansion areas under beam elements have modified
a traditional foundation design for expansive soils. This traditional approach to
foundation design uses designs that mitigate around the swelling effects, rather
than designing to take advantage of, or resist the imposed forces.
This research defines a new shallow foundation design that uses a
curvilinear structure to take advantage of the forces exerted on the foundation by
the expansive forces of the soil. In addition, the design allows for a cast-in-place
or precast implementation. The curvilinear foundation design was modeled using
LS DYNA Finite Element analysis and compared to a traditional rectangular
foundation design using 3D models. In addition, the soil and concrete models
were evaluated by comparing the LS DYNA model results (deflection, shear,
tensile/compression) for both foundation designs.


Also included in the research is the analysis of a discontinuity (window) in
the structural design and the effects of the discontinuity on the structure.
The form and content of this abstract are approved. I recommend its publication.
Approved: Nien-Yin Chang
IV


DEDICATION
I dedicate this thesis to my wife Jerre and express my gratitude and
appreciation for her support, encouragement and sacrifice throughout the extent
of my studies. She has been the foundation on which I have built my life and I
am truly blessed to have such a wonderful partner. I would also like to dedicate
this thesis to my father and mother, John and Viola Gardiner, for instilling in me
perseverance to complete what I have started and for their support. I would also
like to include a special dedication to Jean Durham for her continuous
encouragement throughout the many years of research and study.
v


ACKNOWLEDGEMENT
I would like to express my sincere appreciation and heartfelt thanks to my
advisor Professor Nien-Yin Chang for his continuous guidance, support and
unwavering encouragement throughout my studies and the completion of this
research. The catalyst for this research began with Dr. Chang during coursework
in Intermediate Foundation Engineering in the fall of 2001. Dr. Changs
inexhaustible patience and personal commitment allowed me the freedom and
time to manage multiple priorities during this course of study. In the end I have
the privilege of calling Dr. Chang my friend.
I would like to express my gratitude to Dr. Brian Brady and Dr. Jimmy Kim
for serving on my defense committee and providing me with their valuable inputs
and comments to improve the content of this thesis.
VI


TABLE OF CONTENTS
Chapter
1. Introduction 1
1.1 Purpose of the Study 2
1.2 Scope of the Study 2
1.3 Organizational Outline 3
2. Properties of Expansive Clay Soils 6
2.1 Introduction Expansive Soils 6
2.2 Soil Classification 7
2.2.1 Gradation of Soils 9
2.2.2 Atterberg Limits 13
2.2.3 Activity 15
2.3 Clay Minerals 18
2.3.1 Kaolinite 22
2.3.2 Halloysite 23
2.3.3 Montmorillonite/Smectite 23
2.3.4 lllite 24
2.3.5 Chlorite 25
2.4 Swelling Potential of Clay Soils 27
2.4.1 Consolidometer Swell Test 28
3. Lateral Forces on a Foundation 34
3.1 Lateral Earth Pressure 34
3.2 Ground Water and the Lateral Earth Pressure Coefficient 41
vii


4. Rigid Wall Foundation Design 45
4.1 International Building Code 45
4.2 International Residential Code 47
4.3 American Concrete Institute 49
5. New Foundation Design Approach 51
6. Finite Element Analysis Foundation 54
6.1 LS Dyna Model 54
6.2 Model Configuration 55
6.2.1 Element Types 57
6.2.2 Loading and Boundary Conditions 57
6.2.3 Contact Type 61
6.2.4 Material Properties 62
7. Results 65
7.1 General Behavior of the Foundation Structures 65
7.2 Wall Displacements of the Foundation Structures 67
7.3 Vertical Stress State of the Foundation Structures 70
7.4 Shear Stress State of the Foundation Structures 72
8. Discussion of Analyses Results 76
8.1 Displacements of Foundations 76
8.2 Vertical Stress of Foundations 79
8.3 Shear Stress of Foundation 85
9. Conclusions and Recommendations for Further Research 91
VIII
References
96


LIST OF TABLES
Table
2.1 Clay Activity (after McCarthy, 1998) 17
3.1 Typical values of the coefficient of lateral earth pressure K0. 41
6.1 LS DYNA model material properties 62
8.1 Lateral Displacement (in) of Rectangular and Curvilinear Wall Designs 77
8.2 Vertical Stress of Rectangular and Curvilinear Wall Designs 79
8.3 Stress Concentration Factors around Window - Rectangular 84
8.4 Stress Concentration Factors around Window - Curvilinear 84
8.5 Shear Stress of Rectangular and Curvilinear Wall Designs 85
8.6 Stress Concentration Factors around Window - Rectangular 90
8.7 Stress Concentration Factors around Window - Curvilinear 90
IX


LIST OF FIGURES
Figure
2.1 Grain-size Classification System After U.S. Army Waterways
Experiment Station (1960) and Howard (1977) 9
2.2 Grain Size Distribution (after ASTM International, D2487-11) 11
2.3 Plasticity Index versus Liquid Limit (ASTM D2487 -11, Standard
Practice for Classification of Soils for Engineering Purposes) 15
2.4 Probable Clay Expansion as Estimated from Classification Test
Data (after Holtz, 1959) 16
2.5 Characteristics of Common Clay Minerals (after Mitchell, 1976) 16
2.6 Silica Tetrahedron and Silica Tetrahedral Molecules (after Grim,
1968) 20
2.7 Alumina Octahedron and Alumina Octahedral Molecules (after
Grim, 1968) 21
2.8 Schematic Diagram of Kaolinite (after Lambe, 1953) 22
2.9 Schematic Diagram of Montmorillonite (after Lambe, 1953) 24
2.10 Schematic Diagram of lllite (after Lambe, 1953) 25
2.11 Schematic Diagram of Chlorite (after Mitchell, 1976) 26
2.12 Free Swell Oedometer Test Results 31
2.13 Correction for Sample Disturbance (Fredlund, et.al., 1980) 33
3.1 At-rest Earth Pressure 35
3.2 Wall Movement for Active Earth Pressure 36
3.3 Wall Movement for Passive Earth Pressure 37
3.4 Relationship between Vertical and Horizontal Soil Stress 38
3.5 Subsurface Stresses Soil in At-rest Condition 43
x


3.6 Subsurface Stresses Soil in At-rest Condition With Vertical
Surcharge and Influenced by the Water Table 44
5.1 Rectangular Structure and Soil Backfill 51
5.2 Curvilinear structure and soil backfill 52
6.1 Rectangular Structure Finite Element Model 55
6.2 Curvilinear Structure Finite Element Model 56
6.3 Structural Loading of Rectangular and Curvilinear Foundations 58
6.4 Z Body Soil Loading 60
7.1 Deformed Rectangular Basement Structure (5X Displacement
Factor) 65
7.2 Deformed Curvilinear Basement Structure (5x Displacement Factor) 66
7.3 Lateral Deflection of Rectangular Basement Structure 67
7.4 Lateral Deflection of Rectangular Basement Structure Fringe Plot 68
7.5 Lateral Deflection of Curvilinear Basement Structure 68
7.6 Lateral Deflection of Curvilinear Basement Structure - Fringe Plot 69
7.7 Z-Stress (Vertical) Plot of Window Area of Rectangular Foundation 70
7.8 Z-Stress (Vertical) of Window Area of Rectangular Foundation -
Fringe Plot 71
7.9 Z-Stress (Vertical) Plot of Window Area of Curvilinear Foundation 71
7.10 Z-Stress (Vertical) of Window Area of Curvilinear Foundation -
Fringe Plot 72
7.11 Shear Stress Plot along Window Area of Rectangular Foundation 73
7.12 Shear Stress along Window Area of Rectangular Foundation -
Fringe Plot 74
7.13 Shear Stress Plot along Window Area of Curvilinear Foundation 75
XI


7.14 Shear Stress along Window Area of Curvilinear Foundation -
Fringe Plot 75
8.1 Lateral Deflection of Rectangular Basement Structure Fringe Plot 78
8.2 Lateral Deflection of Curvilinear Basement Structure Fringe Plot 78
8.3 Z-Stress (Vertical) of Window Area of Rectangular Foundation -
Fringe Plot 80
8.4 Z-Stress (Vertical) of Window Area of Curvilinear Foundation -
Fringe Plot 81
8.5 Vertical Stress around Window Rectangular Structure 82
8.6 Vertical Stress around Window Curvilinear Structure 83
8.7 Shear Stress along Window Area of Rectangular Foundation -
Fringe Plot 86
8.8 Shear Stress along Window Area of Curvilinear Foundation Fringe
Plot 87
8.9 Shear Stress around Window Rectangular Structure 88
8.10 Shear Stress around Window Curvilinear Structure 89
xii


1. Introduction
Expansive (swelling) soils are extremely common in the Front Range area
of Colorado and can be found on almost every continent across the globe. The
destructive effects caused by expansive soils have been reported in numerous
countries such as the United States, Canada, Australia, China, Israel, South
Africa and India (Nelson and Miller, 1992; Steinberg, 1998). It has been widely
reported that losses due to expansive soils have been measured in several
billions of dollars yearly (Nelson and Miller, 1992). The cost of repairing damage
caused by swelling soils amounts to more than the cost for all other natural
hazards combined. This is especially true of light structures, pavements and
service piping.
Expansive soils are capable of mobilizing huge vertical and lateral
pressures which, in turn, become a hazard primarily to structures and pavements
built on top of the expansive soil or within the volume of expansive soils that are
subject to moisture changes. The damage may not manifest itself immediately,
depending on the soil composition, moisture history, future moisture/desiccation
cycling and type of foundation construction. Up until now, mitigation in the
industry has followed two primary paths: mitigation of existing structures by
adjusting drainage and/or underpinning; and mitigation of the design by using
pier and beam foundations with drainage implemented at the foundation to
prevent soil expansion. To a lesser extent, changing the soil properties by
chemical mixing, removal of the offending soils or mixture of the soils with more
suitable soil has been implemented in the industry.
1


Although the current industry standards for mitigation of light structures
constructed in expansive soils, as described above, offer piece of mind against
catastrophic damage, long term stability is not guaranteed. Over time, severe
drought and flooding cycles can directly influence the foundation performance
and exceed the design mitigations implemented. In addition, changes in soil
chemistry can occur during periods of high moisture exposure negating the
benefits of soil treatment methodologies. Over time, changes in soil drying can
occur due to rises in the ambient temperatures and/or the growth of vegetation
within the soil mass. With unprecedented weather events taking place all over
the globe, the design of structures must rely less on mitigation of moisture
intrusion and more on foundation designs that take advantage of the potential
forces mobilized by the soil.
1.1 Purpose of the Study
The purpose of this study is to introduce a foundation design that accounts
for and reacts to the pressures generated by expansive soils. It allows for
changes in the moisture content of the expansive soil and is designed to
accommodate the resulting forces. This approach is consistent with the intent of
the International Building Code and the American Concrete Institutes 318-05:
Building Code Requirements for Structural Concrete.
1.2 Scope of Study
The primary scope of this study is to examine the current design and
construction practices for light foundation designs relative to expansive clay soil
pressures with changing soil properties. Included in the study is a curvilinear
2


foundation design approach/solution that works with the in situ soil conditions to
resist the changing lateral pressures and soil heave. The study includes a Finite
Element Model and analysis of a new foundation design compared to a
traditional rectangular foundation design used in industry today.
1.3 Organizational Outline
A brief description of each chapter in this study is presented below.
Chapter 1 introduces the issue of expansive soils as they relate to
foundation design and damage.
Chapter 2 presents a review of clay soil properties and
experimental techniques used in determining swell potential and lateral
swell pressure. The risk of foundation movement relates to the amount
of vertical and horizontal heave/swell that will occur. Heave depends
on more than just the percent swell of the soil. Calculations of
predicted heave must also take into account the stress or surcharge
applied to the soil when the soil is inundated with water. Various
methods are widely used in the industry to classify swell potential and
determine soil properties related to unsaturated clays. One method
commonly used to determine the expansion potential of a soil is based
on the index properties (Holtz and Gibbs 1956; Holtz and Kovacs
1981). This requires knowledge of the clay content and the plasticity
index. These properties can be determined by performing a gradation
test including the Atterburg limits. Another method widely used for
estimating the expansion potential of a clay soil uses soil classification
3


information. Seed et al. (1962) developed a classification chart
method (activity) based on the amount and type of clay particles in the
soil. In addition, experimental methods are also used to determine
swell induced strains, swell potential and swell pressure. These are
typically accomplished by means of a consolidation-swell type test
such as ASTM D4546, One-Dimensional Swell or Collapse of
Cohesive Soils.
Chapter 3 addresses the application of lateral pressures resulting
from the soil mass, surcharge and the water table. It also describes
the methods that are used to define the lateral pressure profile on a
structure.
Chapter 4 reviews the current design practices that are used for
foundation wall design as described in the International Building Code
(International Code Council, 2005), the International Residential Code
for One- and Two-Family Dwellings (International Code Council, 2005),
and the American Concrete Institute, 318-05: Building Code
Requirements for Structural Concrete.
Chapter 5 presents a new, curvilinear foundation design approach
to effectively use the properties of an expansive soil to achieve long-
term survivability and serviceability of the structure. This includes, as
an assumption, changing soil conditions that prove to be problematic to
traditionally designed foundations including moisture and climatic
changes, soil chemistry changes and changes in drainage.
4


Chapter 6 presents the Finite Element Analysis approach for the
new foundation design using LS-DYNA software (Livermore Software
Technology Corporation). The analysis parameters for structural
properties of the foundation and loading due to the soil are presented.
Also discussed is a traditional rectangular foundation design for
comparison.
Chapter 7 presents the findings of the analysis. It compares the
new curvilinear foundation design to the traditional rectangular design,
evaluating wall stress versus applied loading.
Chapter 8 presents the discussion of the results and a comparison
of performance between the traditional rectangular foundation design
and the curvilinear design.
Chapter 9 presents the summary and conclusions of the
research/analysis and recommendations for future, related research.
5


2. Properties of Cohesive Soils
2.1 Introduction Expansive Soils
Clay soils are often described as cohesive, fine-grained soils having
plasticity and containing clay minerals such as kaolinite, halloysite,
montmorillonite, illite, chlorite and vermiculite (Holtz & Kovacs, 1981). However,
not all fine-grained soils are cohesive and/or clay. Silts, for example, are
classified as fine-grained and granular but are not cohesive and are not plastic.
For clay soils, grain size distribution has little influence on the properties of the
clay whereas for granular soils the grain size distribution and the grain shape can
have marked effects on the properties of the soil. Additionally, water content is
relatively unimportant (with a few exceptions) for granular soils but has a definite
influence on clay soils. Silts are fine-grained and granular but are not plastic and
are non-cohesive. Their strengths, like sands, are essentially independent of
water content.
Clay minerals owe their unique properties and behavior to some very
distinct characteristics. Clay minerals are extremely small particles (< 1 pm
diameter) that are electrochemically active. They are affected by the quantity
and type of clay minerals present, the moisture content, the type and chemistry
of the soil water surrounding the clay particles, the arrangement, soil density and
specific surface area of the clay particles. In a mixed clay and soil mass, as the
clay content increases, the behavior of the soil mass is increasingly governed by
the clay fraction properties. As the clay content approaches and exceeds
approximately 50%, the sand and silt grains in the mixed clay/soil mass are
6


floating within the clay matrix which dominates the soil mass behavior (Holtz &
Kovacs, 1981).
2.2 Soil Classification
The purpose of soil classification is to provide for a common means of
determining or predicting the behavior of soils and/or evaluating soils for
engineering purposes. There are numerous soil classification systems in use. In
the United States, the Unified Soil Classification System (USCS) is the most
widely used soil classification system for structural considerations (Howard,
1977) while the American Association of State Highway and Transportation
Officials (AASHTO) classification system is typically used for pavement design.
The Unified Soil Classification System (USCS) was initially developed by
Casagrande in1948 and later modified by Casagrande in 1952.
Within the USCS (Figure 2.1), soil materials are classified into three main
groups: Coarse-grained, Fine-grained and Peat (highly organic soils) depending
on the predominant particle sizes and make-up within the soil matrix. Soils are
identified within the three major groups primarily on the basis of particle sizes
and changes to the soil properties and volume when interacting with water.
Coarse-grained soils (sands and gravels) contain particles that are visible to the
naked eye (larger than about 0.003 in. [0.075 mm]) and are generally described
as cohesionless, with engineering behavior primarily influenced by the
composition of particle sizes, particle shape, and relative density. Coarse-grained
soils are further defined within the USCS as greater than 50% (by dry mass)
retained on the number 200 Standard Sieve with a mesh opening of 0.075mm.
7


Subdivisions within this classification system are largely based on particle size:
gravels (75mm to 4.75mm) and sands (4.75mm to 0.075mm). Both sands and
gravels are further subdivided into four secondary groups (GW, GP, GM, GC;
SW, SP, SM, SC). The four secondary classifications are based on whether the
soils are well graded, poorly graded, contain silt-size particles or contain clay-
size particles.
Fine-grained soils include silts and clays containing particles that are not
visible to the naked eye. Fine-grained soils are those composed primarily of silt
and clay-sized particles smaller than 0.075 mm. Fine-grained soils are defined
as having 50 percent or more (by dry mass) of soil particles passing through the
number 200 Standard Sieve. Silts and clays are largely distinguished based on
the plasticity properties of the soil, as measured by the soils' Atterberg Limits.
Both silts and clays are further subdivided into three secondary groups (ML, CL,
OL; MH, CH, OH). The three secondary classifications are based on the
inorganic and organic nature of the soil and on its plasticity. Silts may be either
cohesive or cohesionless and are granular materials with sizes falling between
sands and clays. Silts may occur as a soil or as suspended sediment. Clays, on
the other hand, are cohesive soils, with engineering behavior primarily influenced
by plasticity and cohesion. Soils containing high natural organic content
comprise the third major group. Peats (organic soils) can be of extremely low
strength and high compressibility, depending on organic content and
composition, and geologic history.
8


Unified Soil Classification System (USCS)
Coarse- grained soils (More than 50% retained on No. 200 (0.075mm) sieve) Gravels (More than 50% of coarse fraction retained on no. 4 (4.75 mm) sieve) GW Well-graded gravels or gravel-sand mixtures, little or no fines Less than 5% finest*] Cu >4 and 1 < Cc < 3
GP Poorly graded gravels or gravelly sands, little or no fines Less than 5% finest*] Does not meet Cu and/or Cc criteria listed above
GM Silty gravels, gravelsand-silt mixtures More than 12% finest*] Minus no. 40 soil plots be low the Aline
GC Clayey gravels, grave Isand-clay mixtures More than 12% finest*] Minus no. 40 soil plots on or above the Aline
Sands (50% or more of coarse fraction passes no. 4 (4.75 mm) sieve) sw Well-graded sands or gravelly sands, little or no fines Less than 5% fines[*] Cu £ 6 and 1 SP Poorly graded sands or grave lly sands, little or no fines Less than 5% fines[*] Does not meet Cu and/or Cc criteria listed above
SM Silty sands, sand-silt mixtures More than 12% finest*] Minus no. 40 soil plots below the A Line
SC Clayey sands, sandclay mixtures More than 12% finest*] Minus no. 40 soil plots on or above A line
Fine-grained soils (50% or more passes no. 200 (0.075mm) sieve) Silts and clays (Liquid limit less than 50) ML Inorganic silts, rock flour, silts of low plasticity Inorganic soil PI < 4 or plots be low A line
CL Inorganic clays of low plasticity, gravelly clays, sandy clays, etc. Inorganic soil PI > 7 and plots on or above Aline [f]
OL Organic silts and organic clays of low plasticity Organic soil LL (ovendried)/ LL (not dried) < 0.75
Silts and clays (Liquid limit 50 or more) MH Inorganic silts, micaceous silts, silts of high plasticity Inorganic soil Plots be low A line
CH Inorganic highly plastic clays, fat clays, silty clays, etc. Inorganic soil Plots on or above A line
OH Organic silts and organic clays of high plasticity Organic soil LL (oven dried) / LL (not dried) < 0.75
Peat Highly Organic PT Peat and other highly organic soils Primarily organic matter, dark in color, and organic odor
[*]Fines are those soil particles that pass the no. 200 sieve. For gravels with 5 to 12 percent fines, use of dual symbols required (i.e., GW-GM, GW-GC, GP-GM, or GP-
GC). For sands with 5 to 12 percent fines, use of dual symbols required (i.e., SW-SM, SW-SC, SP-SM, or SP-SC).
[t]lf 4 £ PI £ 7 and plots above A line, then dual symbol (i.e., CL-ML) is required.
Figure 2.1: Grain-size Classification System (after U.S. Army Waterways
Experiment Station (1960) and Howard (1977))
2.2.1 Gradation of Soils
Gradation tests are performed on a soil to determine the particle size
distribution which is used in the classification of a soil. The gradation of a soil
has a major effect on its mechanical and hydraulic properties and enables an
evaluation of engineering characteristics such as permeability, strength, swelling
potential, and susceptibility to frost action. The tests consist of two types: sieve
analysis for coarse-grained soils (sands, gravels) and hydrometer analysis for
fine-grained soils (silts, clays). Materials containing both types of soils
9


(sands/gravels and silts/clays) are tested by both methods and the results are
merged to create one particle size distribution result.
Gradation of coarse-grained soils consists of a mechanical grain size
analysis. The analysis consists of taking an oven-dried soil sample and
subjecting it to a series of standard sieves with progressively smaller openings
while mechanically shaking the sieves. Once complete, the amount of material
retained on each of the sieves is weighed. The total percentage passing each
sieve is determined and the data plotted on a semilogarithmic graph of grain size
versus percent finer by weight (Figure 2.2). Based on the results of the particle
size distribution testing, soils can be classified as poorly-graded (uniform), when
it contains a narrow distribution of particle sizes or well-graded, when the soil has
a wide range of particle sizes. The flatter the grain size curve the larger the
range of particle sizes found in the soil and the steeper the curve the fewer the
particle sizes. Generally speaking, a well-graded soil has a curve that is smooth
and contains particles over a relatively large range of sizes while a poorly-graded
soil has a curve where a high portion of the soil particles contain sizes within a
narrow band. If particles of large and small sizes are present with a low
proportion of particles in the intermediate sizes the soil is categorized as a gap-
graded soil (McCarthy, 1998).
10


Grain Size Distribution
Figure 2.2: Grain Size Distribution (after ASTM International, D2487-11)
A hydrometer analysis is performed on soils finer than the No. 200 sieve
(0.075 mm) since a sieve analysis is impractical for small diameter particles
(grains). The hydrometer analysis is a sedimentation process where the rate of
settlement of a soil in water is measured as an indication of particle size. The
test is based on Stokes law for falling spheres in a viscous fluid where the
terminal velocity of fall depends on the grain diameter and the densities of the
grains in suspension and of the fluid. The particle diameter can be determined
from knowledge of the distance of fall and the time. Stokes law does not apply to
particle sizes below 0.0002 mm as these particle sizes are influenced by
Brownian movement (U.S. Army Corps of Egineers, 1998).
Interpretation of the gradation analysis focuses on the range of particle
diameters found in the sample. This information can be readily determined from
11


the semi-logarithmic grain size distribution curve (Figure 2.2). The particle size
representing a given percentage smaller can be directly determined from
reading the particle size from the specific percentage finer number. Sizes
commonly used in calculating uniformity coefficients are the percentage smaller
than 10%, 30% and 60% and are denoted D-|0, D30 and D6o, respectively. As a
measure of the gradation of a soil, the coefficient of uniformity (Cu) is used to
describe a soils range of particle sizes. It is defined as the ratio of the D6o size of
the soil (the particle size in mm where 60% of the soil particles are finer than) to
the D-io size (the particle size in mm where 10% of the soil particles are finer
than). The uniformity coefficient (Cu) is calculated as the following ratio:
D60 2.1
mo
Where: D60 = soil particle diameter at which 60% of the mass of a soil sample is
finer and D10 = the diameter at which 10% of the mass of a soil sample is finer.
The D-io is often referred to as the effective particle size and is utilized in
many empirical methods to characterize the soil as a whole, particularly with
regard to hydraulic conductivity. Generally, the higher the value of the coefficient
of uniformity (Cu) the greater the range of particle sizes in the soil sample.
Another quantity that may be used to judge the gradation of a soil is the
coefficient of curvature, designated by the symbol Cc. The coefficient of
curvature is defined as the following:
12


2.2
D302
D60xD10
Where: D60 = soil particle diameter at which 60% of the mass of a soil sample is
finer, D10 = soil particle diameter at which 10% is finer and D30 = soil particle
diameter at which 30% of the mass of the soil is finer.
A well-graded soil is defined as having a good representation of all particle
sizes from the largest to the smallest and the shape of the grain size distribution
curve is considered "smooth." In the USCS, well-graded gravels must have a Cu
value > 4, and well-graded sands must have a Cu value > 6. For well-graded
sands and gravels, a Cc value from 1 to 3 is required. Sands and gravels not
meeting these conditions are considered poorly graded.
2.2.2 Atterberg Limits
Atterberg limits are limits of moisture content (mass of water in the soil to
the mass of the solid particles) used to define fine-grained soil behavior. In
engineering practice, three of the limits (the liquid, plastic and shrinkage limits)
are commonly used.
The Liquid Limit (LL) is the water content, in percent, that defines where
the soil changes from a viscous, fluid state to a plastic state. Above this point the
soil behaves as a liquid, while below this point the soil behaves as a plastic
material. The Liquid Limit can be measured using the (Casagrande) liquid limit
device.
13


The Plastic Limit (PL) is defined as the water content, in percent, where
the soil changes from a plastic state to a semi-solid state. Above this point the
soil behaves as a plastic material, while below this point the soil behaves as a
semi-solid. The Plastic Limit is also the moisture content at which a soil
crumbles when rolled into a thread of 1/8 inch in diameter (Das, 2002).
The Shrinkage Limit (SL) is defined as the moisture content where the soil
volume will not reduce further if the moisture content is reduced. Above this
point the soil behaves as a semi-solid, while below this point the soil behaves as
a solid.
Plasticity Index (PI) is defined as the difference between the moisture
content at the Liquid and Plastic Limits. This represents the range of water
content where a material behaves plastically (Das, 2002).
PI = LL- PL
2.3
Since the PI is determined from Atterberg Limits testing on the fraction of
soil that passes the no. 40 sieve (0.425 mm), a correction factor is applied for
soils that contain a large fraction of particles coarser than the no. 40 sieve.
Fine-grained (cohesive) soils can be classified either as low or high
compressibility materials based on the results of the Atterberg Limits tests. By
plotting the Plasticity Index versus the Liquid Limit the classification can be
determined graphically (Figure 2.3).
14


Plasticity Chart for Classification of Fine-Grained Soils
Figure 2.3: Plasticity Index versus Liquid Limit (ASTM D2487 -11, Standard
Practice for Classification of Soils for Engineering Purposes)
The A-Line separates clay classifications and silt classifications, while the
U-Line represents an approximate upper limit of LL and PI combinations for
natural soils.
2.2.3 Activity
A variety of soil engineering properties have been correlated to the liquid
and plastic limits as well as being used to classify fine-grained soils according to
the Unified Soil Classification System. Knowledge of the Atterberg limits for a
cohesive soil and the natural moisture content can tell a good deal about its
geologic history and engineering performance (Figure 2.4).
15


Probable Expansion as Estimated from Classification Test Data
Degree of Expansion Probable Expansion as a % of the Total Volume Change (Dry to Saturated Condition)* Colloidal Content(%- lum) Plasticity Index, PI Shrinkage Limit, SL
Very High >30 >28 >35 <11
High 20-30 20-30 25-41 7-12
Medium 10-20 13-23 15-28 10-16
Low <10 <15 <18 >15
Figure 2.4: Probable Clay Expansion as Estimated from Classification Test Data
(after Holtz, 1959)
The presence of small amounts of certain types of clay minerals can have
significant impacts on the soils properties. The identification of the type and
amount of the clay minerals present can help in determining or predicting the
soils behavior or to determine how to minimize the effects of the clay minerals
present (McCarthy, 1998). Indirect methods are available to determine
information about the type and effects of clay minerals in a soil that are relatively
easy to perform and give qualitative, if not quantitative, results (Figure 2.5).
Characteristics of Common Clay Minerals
Atterberg Limits
Mineral Group Basal Spacing (A) Particle Features Interlayer Bonding Specific Surface (m2/g) Liquid Limit % (LL) Plastic Limit % (PL) Shrinkage Limit % (SL) Activity Ratio (PI/% clay)
Montmorillonite 9.6 Thin, filmy, flakes > 10A X1.0 to 10 pm Very weak van der Waals bonds 700 840 100-900 50-100 8.5-15 7.2
lllites (Mca-like) 10 Thin, stacked plates 0.003 to 0.1 X 1.0 to 10 pm Strong potassium bonds 65-100 60-120 35-60 15-17 0.9
Kaolinites 14.4 Thick, stiff 6-sided flakes 0.1 to 4 X 0.05 to 2 pm Strong hydrogen bonds 10-20 30-100 25-40 25-29 0.38
Figure 2.5: Characteristics of Common Clay Minerals (after Mitchell, 1976)
16


One such method is the clay Activity and is determined from a relationship
of the plasticity of the clay and the quantity of clay-sized particles. For a given
amount of clay mineral the plasticity resulting in a soil varies for different clay
minerals. The Activity, defined as
Plasticity Index 2.4
% by Weight Finer than 2gm
Where the percent by weight finer than 2 pm is equal to that portion of the
soil (by weight) consisting of particles <0.002 mm. This information can be
determined by the hydrometer analysis (ASTM International D422-63).
Three classes of clays according to activity suggested by Skempton
(1953) are inactive (A < 0.75), normal (0.75 < A < 1.25) and active (A > 1.25).
The clays with the highest activity have the most potential for expansion.
Table 2.1: Clay Activity (after McCarthy, 1998)
Activity Classification
<0.75 Inactive Clay
0.75-1.25 Normal Clay
>1.25 Active Clay
Typical values of activities for various clay minerals range from quartz
(activity = 0) to sodium montmorillonite (activity = 4 to 7) at the extremes while
17


Illite (activity = 0.9) and calcium montmorillonite (activity = 1.5) fall somewhere in
between.
2.3 Clay Minerals
The cohesive properties of natural soils are related to the presence of clay
minerals (e.g., kaolinite, halloysite, montmorillonite, illite, chlorite and
vermiculite). The most important of these clay minerals associated with
expansion are kaolinite, montmorillonite and illite. All are very small crystalline
substances of hydrous aluminosilicates (phyllosilicates or layered silicates). The
crystalline substances are a result of weathering (physical or chemical) of parent
rock materials, primarily igneous and sedimentary rocks (G.W. Donaldson,
1969).
Three of the most important clay mineral groups are the kaolinites
(generally non-expansive), the mica mineral group which includes the illites and
vermiculites (can be expansive), and the smectites which includes
montmorillonite (can be highly expansive). The swelling potential varies widely
with the type of clay mineral and is generally ranked in the following order from
most to least expansive: montmorillonite, illite, kaolinite (Lambe and Whitman,
1969).
Clay minerals are generally constructed of stacks of two types of sheets:
silica tetrahedral sheets and alumina octahedral sheets. Each sheet is only
angstroms thick (1 .OOOOe'10 m) and can be thousands of angstroms wide in each
of their lateral dimensions. The individual crystals can only be observed with an
electron microscope and their structure has been observed using x-ray
18


diffraction. Each family of clay mineral type consists of vertical stacks of these
elementary plates or sheets in differing arrangements. Each plate or sheet has a
repeating atomic structure consisting of the two primary crystal sheets.
The silica tetrahedral sheets are constructed of individual tetrahedron molecules
with the general composition of SiCU (in some instances contain substitutions of
aluminum ions for silica ions). The alumina octahedral sheets are constructed of
individual octahedron molecules with the general composition of AI(OH)6
(Gibbsite) with substitutions of either iron or magnesium (Brucite) for aluminum
ions. If all of the anions are hydroxyls and at least 2/3 of the cation positions are
filled with aluminum the mineral is labeled gibbsite. If the cation positions are
filled with magnesium instead of the aluminum then the mineral is labeled brucite.
The arrangement of the stacking of these sheets, with the various metal ion
substitutions throughout the crystal lattice, makes up the different clay mineral
types (Holtz and Kovacs, 1981).
The basic tetrahedral sheet is a combination of silica tetrahedral
molecules. The tetrahedral molecule consists of four oxygen atoms at each of
the corners of the tetrahedron with a silicon atom at its center (tetrahedral sheet
Figure 2.6). The oxygen atoms at the base of a single tetrahedron molecule are
combined to form continuous sheets with the base oxygen atoms arranged in a
single plane.
19


o Oxygen # Silicon
Figure 2.6: Silica Tetrahedron and Silica Tetrahedral Molecules (after Grim,
1968)
The basic octahedral sheet is a combination of octahedral molecules. The
octahedral molecule consists of six oxygen atoms or hydroxyl molecules
(Oxygen-Hydrogen molecule) with aluminum, magnesium, iron, or other atom at
its center (single octahedral molecule Figure 2.7). The oxygen atoms or the
hydroxyl molecules are positioned such that two planes are formed within the
sheet. Substitutions of the cations within the octahedral sheet are common and
produce different clay minerals.
20


o Hydroxyls or Oxygen Aluminum, Magnesium, etc.
Figure 2.7: Alumina Octahedron and Alumina Octahedral Molecules (after Grim,
1968)
All clay minerals contain the two basic tetrahedral and octahedral sheets
and are stacked together in various combinations with differing substitutions of
cations within the molecules to form the minerals. The various types of clay
minerals result from the stacking of the specific tetrahedral and octahedral sheets
and the type of chemical bonding between each sheet. The variety of
mechanical and physical properties of different clays is a result of the unbalanced
electrical charges that are a result of certain cationic substitutions and the overall
net electronegative polarity resulting from the sheet stacking arrangements. This
includes differences in cohesion, water absorption and expansion found in the
different clay minerals.
The more common clay minerals found in clay soils within the Unites
States are described in the following sections.
21


2.3.1 Kaolinite
Kaolinite minerals are formed of repeating layers of one tetrahedral (silica)
sheet and one octahedral (alumina or gibbsite) sheet and are classified as 1:1
minerals. The two layers are held together by hydrogen bonding (Hydroxyl ions
on the octahedral sheet to oxygen atoms on the tetrahedral sheet) forming a
single layer (Figure 2.8). The strong hydrogen bonding resists swelling stresses
when water is present and also allows the construction of large crystal structures.
These mineral layers are approximately 0.72 nm (7.2 A) in thickness and extend
laterally in both directions indefinitely. Kaolinite crystals are constructed of
repeating layers of the 0.72 nm mineral layer. It is not uncommon to have
kaolinite crystals 70 to 100 layers thick (Holtz and Kovacs, 1981). The cation
exchange capacity of Kaolinite is very low (3 to 15 meq/100 gm) owing to little
substitution within the mineral sheets.
ft
Repeating structure
Gibbsite -typ.
Silica sheet typ.
Repeating structure
ft
Figure 2.8: Schematic Diagram of Kaolinite (after Lambe, 1953)
22


2.3.2 Halloysite
Halloysite is another 1:1 clay mineral but differs from Kaolinite due to
water entrained between the mineral layers. This hydration results in
deformations of the crystal structure and random stacking of the mineral sheets
giving it a tubular structure. The water can be removed from the mineral layers
by heating or air drying but is irreversible. The Halloysite mineral sheet will not
rehydrate upon addition of water to the clay mineral (Holtz and Kovacs, 1981).
When hydrated, the clay exhibits a 1 nm spacing of the layers and when
dehydrated the mineral layers are 0.7 nm in spacing. Halloysite naturally occurs
as small cylinders which average 30 nm in diameter with lengths between 0.5
and 10 micrometers. The cation exchange capacity depends on the amount of
hydration, as 2H20 has 5-10meq/100g, while 4H20 has 40-50meq/100g.
2.3.3 Montmorillonite/Smectite
Montmorillonite, or Smectite, is a 2:1 mineral containing two silica
tetrahedron sheets and one alumina (gibbsite) octahedron sheet. The
octahedron sheet is located between the two silica tetrahedron sheets forming a
single layer (Figure 2.9). The tips of the silica tetrahedrons form a bond with the
hydroxyls of the alumina octahedrons by Van der Waals forces. The bonding
forces are weak and allow water and exchangeable ions can enter the layers.
Typical thickness of a 2:1 mineral layer is approximately 0.96 nm (9.6 A) and
extends laterally in both directions indefinitely (Holtz and Kovacs, 1981). Due to
the very small size and the affinity for water, Montmorillite is highly expansive
23


depending on the initial and final moisture contents. The cation exchange
capacity of smectite is in the range of 80 to 150 meq/100 gm.
If
Repeating structure
Repeating structure
Figure 2.9: Schematic Diagram of Montmorillonite (after Lambe, 1953)
2.3.4 lllite
lllite is another 2:1 mineral similar to Montmorillite with the interlayers
bonded with a potassium atom. Illites consist of one octahedral sheet
sandwiched between two silica tetrahedral sheets. The potassium atom fits into
the hexagonal space created in the silica tetrahedron sheet and bonds the
mineral layers together (Figure 2.10). The resulting charge is somewhat
balanced by the potassium atoms in the hexagonal space between the layers.
This makes the potassium essentially non-exchangeable since bonding is very
strong. Ionic substitutions do occur in lllite, usually in the silica tetrahedral layers.
24


This strong bonding reduces the potential for expansion by preventing water
intrusion between the layers allowing the layers to remain essentially constant
(Mitchell, 1993). The cation exchange capacity of lllite is in the range of 10 to 40
meq/100 gm.
ft
Repeating structure
Potassium ion
Gibbsite -typ.
Silica Sheet typ.
Repeating structure
ft
Figure 2.10: Schematic Diagram of lllite (after Lambe, 1953)
2.3.5 Chlorite
Chlorite is a 2:1:1 mineral consisting of a silica tetrahedron sheet, an
alumina octahedral sheet, a silica tetrahedron sheet followed by either a gibbsite
or brucite sheet (Figure 2.11). Unlike other 2:1 clay minerals, chlorite's interlayer
space is made up of Mg2+ or Fe3+, more commonly referred to as the brucite-
like layer. Chlorite can be missing an occasional brucite or gibbsite layer and also
have considerable isomorphous substitution. This may lead to a higher
25


susceptibility to swelling due to water entering between the sheets. Generally, it
is significantly less active than montmorillonite (Holtz and Kovacs, 1981). Cation
exchange capacity for silt size chlorites varies from 4 to 32 meq/100 gm and for
-2 p chlorite particles from 30 to 47 meq/100 gm. Cation exchange capacities for
-2 p and -1 p chlorites are essentially the same.
ft
Repeating structure
Gibbsite or Brucite
Silica Sheet -typ.
7-

/
Brucite
\ Si /
Al
-/ Si \
Brucite
\ /

/ \
Tetrahedral sheet typ.
Octahedral sheet typ.
14.0 A
Repeating structure
ft
Figure 2.11: Schematic Diagram of Chlorite (after Mitchell, 1976)
26


2.4 Swelling Potential of Clay Soils
One of the most important considerations in determining risks to structures
due to expansive soils is identifying the vertical and lateral swell potential of the
soil and the resulting swell pressures. One-dimensional tests are by far the most
widely used method to estimate expansive soil swelling potential and pressures.
This is due largely to the simplicity of the procedures and the availability of the
testing equipment. Standard test methods for evaluating the potential for one-
dimensional heave / settlement and swell pressure of cohesive soils is described
in ASTM D 4546 (One-Dimensional Swell or Collapse of Cohesive Soils). Three
oedometer/consolidometer tests useful for measuring potential swell / settlement
and swelling pressure can be described as follows:
1) Free-swell test A seating pressure (e.g., 0.01 tsf) is applied to the test
sample in an oedometer/consolidometer, the sample is inundated with
water and allowed to swell vertically until primary swell is complete. The
sample is loaded following primary swell until its initial void ratio / height is
obtained. The total pressure required to reduce the test sample height to
the original void ratio / height prior to inundation is defined as the swell
pressure.
2) Overburden-swell test A vertical pressure exceeding the seating
pressure is applied to the test sample in an oedometer/consolidometer
and the sample is inundated with water. The test sample may swell, swell
then contract, contract, or contract then swell. The vertical pressure is
27


typically chosen to replicate the in situ overburden pressure and may
include structural loads depending on the purpose of the test.
3) Constant-volume swell test -. A seating pressure and additional vertical
pressure (typically equivalent to the in situ overburden pressure) is applied
to the test sample in an oedometer/consolidometer and the sample is
inundated with water. Additional vertical pressure is applied as needed or
removed to maintain a constant void ratio / height of the test sample. A
consolidation test is subsequently performed and the total pressure
required to maintain a constant void ratio / height of the test sample is the
measured swell pressure. This measured swell pressure is corrected to
compensate for sample disturbance by using the results of the subsequent
consolidation test. A suitable correction procedure is similar to that for
estimating the maximum past pressure.
The procedures outlined typically use a rigid soil confining ring in an
oedometer/consolidometer apparatus to measure the vertical stress and strain
components of swell.
2.4.1 Consolidometer Swell Test
ASTM Method D4546, Standard Test Methods for One-Dimensional
Swell or Collapse of Cohesive Soils, describes three laboratory methods for
measuring free swell, swell pressure and the magnitude of one-dimensional swell
or collapse of compacted or intact cohesive soils. The test methods can be used
to measure the magnitude of one-dimensional wetting-induced swell or collapse
(hydrocompression) under different vertical loading, as well as the magnitude of
28


vertical swell pressure and the magnitude of vertical free swell. It can also be
used to obtain data for stress-induced compression following wetting-induced
swell or collapse. All of the methods involve the use of a one-dimensional
oedometer/consolidometer apparatus to laterally restrain the soil sample and
allow for access to free water. Three alternative methods are described to
determine the swell behavior and measure the swell parameters of the soil. The
three testing procedures for determining the swelling pressure of a soil can be
described as:
Method A wetting-after-loading testing of multiple samples differing
surcharge loading performed on compacted or natural soil samples
followed by inundation with free water.
Method B single point wetting-after-loading testing of a single sample -
a single surcharge load on a single intact specimen of natural soil, or a
single intact specimen of compacted soil obtained from an existing fill or
embankment followed by inundation with free water.
Method C loading-after-wetting test after completion of the Method A
or B testing increments of additional vertical loads are applied to the
sample and the load-induced deformations are determined. The results
would apply to situations where new fill and/or additional structural loads
are applied to the ground that has previously gone through wetting-
induced heave or settlement.
Typically, two classes of testing are performed, free swell and either a free
swell test followed by consolidation or a continuous consolidation upon
29


inundation of water to keep a constant volume. In the free-swell test, a sample is
subjected to an applied load and allowed to swell freely. The resulting final
volume changes are plotted against the corresponding applied loads or stresses.
The stress corresponding to zero volume change (reloading the sample to the
initial void ratio / height) is termed the swelling pressure (Hardy et al, 1962). The
swelling pressure may be further defined as the pressure that prevents either a
positive or negative volume change.
In the free-swell followed by consolidation or continuous consolidation upon
inundation with water, the soil sample is inundated with water and allowed to
swell freely with a given load applied. The soil is gradually consolidated back to
its original void ratio / height in the conventional manner of a consolidation test
procedure. The swelling pressure is defined as the stress necessary to
consolidate the specimen back to its original volume or the minimum stress
required to prevent swelling (ASTM D4546-08). The swell pressure (Ps) is the
applied load required to prevent swell strain (L) divided by the cross-sectional
area of the specimen (A):
Ps = L/A2 2.5
Stability is assumed to occur when no further applied load is required to restrict
vertical strain.
Swell strain is determined by measuring the resultant heave after reaching
stability (no further increase in sample height with time) upon inundation of water
30


to the sample. The heave is defined as the change in height of the sample and
the swell strain in the change in height divided by the initial height of the sample:
The test results are commonly plotted as shown in Figure 2.12, vertical height or
void ratio versus the log of the vertical load (or stress) applied.
Free Swell Oedometer Test
Pressure (psf)
Figure 2.12: Free Swell Oedometer Test Results
31


The actual stress paths followed during the test can be traced beginning
with water inundation initiating the swell of the soil followed by loading of the
sample to reach the original height/void ratio. At this point the pressure required
(Ps) to resist the swelling can be determined.
To empirically account for sampling disturbance, Fredlund et al. (1980)
defined a correction procedure which could be applied to the data to give a
corrected swelling pressure. The correction procedure is a modification of the
Casagrande type of geometrical construction as shown in Figure 2.13. Using the
constant volume oedometer test plot (void ratio versus logarithm of total
pressure) a correction can be established by identifying the point of maximum
curvature of the curve. The point is found immediately past the uncorrected
swelling pressure. From that point, horizontal and tangential lines can be drawn
and the resulting angle bisected. The intersection of a line parallel to the
rebound portion of the curve and the bisector indicates the corrected swelling
pressure.
32


Figure 2.13: Correction for Sample Disturbance (Fredlund, et.al., 1980)
The potential swelling pressure and volume change of a soil sample can
be determined from ASTM Method D4546, Standard Test Methods for One-
Dimensional Swell or Collapse of Cohesive Soils, laboratory tests. The results
can be applied to actual foundations with reasonable accuracy when the stress
distribution of the foundation soil is understood and the effects of seasonal
variations and the movement of moisture beneath the foundation is known. The
oedometer tests, in which a soil sample is subjected to estimated surcharge
loads, will help understand the in-situ potential expansion and pressure.
33


3. Lateral Forces on a Foundation
Lateral forces applied to a structural foundation from the surrounding soil
are directed so as to apply a load perpendicular to the plane of the wall. This
loads the wall in flexure (bending) and it must act as a one-way slab, beam, or
two-way slab, depending on the design. Determination of the magnitude and
orientation versus the depth of these forces are crucial to the development of a
safe and economic design. For a proper structural design, these lateral forces
are generated from three sources: the soil used for backfill, the water table depth
in the backfill and the surcharge loads (if any) at the top of the backfill in
proximity to the wall.
3.1 Lateral Earth Pressure
The magnitude of the lateral soil pressure that can develop in a soil mass is a
function of several factors: the type of soil, the strength of the soil, the stress-
strain properties of the soil, the unit weight of the soil, the drainage conditions of
the soil, the water table depth, and the amount and direction of wall movement
when subjected to the lateral soil pressure. In defining the stress-state of the
wall three conditions may exist:
(1) The wall is restrained from moving either toward the soil mass or away
from the soil mass and the lateral earth pressure on the wall is defined as
the at-rest earth pressure. No deformations or displacements are
occurring in this stress-state. This is described as an at-rest condition.
34


(2) The wall may tilt or translate away from the soil mass where a triangular
soil wedge, behind and adjacent to the wall, may fail. This lateral earth
pressure in this condition is defined as an active earth pressure.
(3) The wall may tilt or translate into the soil mass where, with sufficient
movement, a soil wedge may fail. This lateral earth pressure in this
condition is defined as a passive earth pressure.
The at-rest earth pressure condition can be explained by means of Figure 3.1,
where length A-B is a structural retaining wall that supports a retained soil mass.
The backfill is horizontal, typical of a structural basement wall. If the structural
wall does not move (rotate) or yield either toward or away from the retained soil,
the horizontal lateral earth pressure at any depth to which the wall will be
subjected is called the lateral earth pressure at-rest. The total force per unit
length of the wall is equal to o0.
Figure 3.1: At-rest Earth Pressure
35


If the structural wall tends to move (rotate) or yield away from the retained
soil mass, creating a plastic deformation in the soil mass, the lateral earth
pressure at any depth to which it will be subjected is called the active earth
pressure. This condition is described in Figure 3.2. This active condition can be
a result of rotation of the wall about its bottom or top, or by translation of the wall
away from the retained soil mass. The resultant forces (oa) magnitude,
direction, and location per unit length of the wall depends on several factors
including soil type, shear strength, backfill incline and stiffness of the structural
wall and foundation.
Figure 3.2: Wall Movement for Active Earth Pressure
If the structural wall tends to move (rotate) or yield into the retained soil
mass, creating a plastic deformation in the soil mass, the lateral earth pressure at
any depth to which it will be subjected is called the passive earth pressure. This
condition is described in Figure 3.3. This passive condition can be a result of
36


rotation of the wall about its bottom or top, or by translation of the wall into the
retained soil mass. The resultant forces (op) magnitude, direction, and location
per unit length of the wall also depend on several factors including soil type,
shear strength, backfill incline and stiffness of the structural wall and foundation.
Figure 3.3: Wall Movement for Passive Earth Pressure
In this research the foundation wall is modeled as a two-way slab with
sufficient support at the base and top of the foundation wall to prevent either
active or passive earth pressures and is therefore modeled using an at-rest earth
pressure.
For an at-rest condition, vertical stresses existing in a soil mass at a given
depth Z below the ground surface where the water table exists below the depth Z
of the soil element is the weight of the overburden and can be written as
&v ~ YsoilZ
37


where Ysoii is the unit weight of the soil mass for a homogenous, isotropic mass of
infinite extent (McCarthy, 1980).
Figure 3.4: Relationship between Vertical and Horizontal Soil Stress
The relationship between the vertical stress and the horizontal stress
(Figure 3.4) is determined by calculation of the coefficient of lateral earth
pressure at-rest Ko and is the ratio of horizontal stress oh and the vertical stress
av.
<*h v 3 2
K0 = = ------- = 1 sin in
av 1 v
Where v is Poissons ratio and cp is the angle of internal friction of the soil. The
above equation (Jaky, 1944) is an empirical approximation and is defined as the
38


at-rest condition for normally consolidated sands designated Ko which can be
written:
(Th
3.3
For normally consolidated clays, the coefficient of lateral earth pressure Ko
in the at-rest condition is approximated by Brooker and Ireland (1965) as
K0 =0.95 sin (p 3.4
where cp is the drained peak friction angle of the soil. In addition, Brooker and
Ireland experimented with normally consolidated clay soils and determined the
value for Ko may be approximated with the Plasticity Index (PI) as
K0 = 0.4 + 0.007PI
3.5
This for a Plasticity Index between 0 and 40 and
K0 = 0.64 + 0.001P7
3.6
for a Plasticity Index between 40 and 80.
39


Where overconsolidated clays are concerned Ko can be defined as
follows:
Ko (over consolidated) ~ ^Ofnormany consolidated.)* OCR
Where OCR is defined as the overconsolidation ratio. A soil is described as
normally consolidated when the preconsolidation pressure equals the existing
vertical overburden pressure (the soil has not experienced anything greater than
the current overburden stress). If the preconsolidation pressure is greater than
the existing vertical overburden pressure the soil is described as being
overconsolidated (the soil had experienced a pressure that is larger than the
current overburden condition). The overconsolidation ratio can be determined by
the ratio of the preconsolidation stress divided by the existing vertical effective
overburden stress
OCR =
3.8
where op is the preconsolidation pressure and ovo is the effective vertical
overburden pressure (Holtz & Kovacs, 1981). As can be seen from the OCR
equation, soils that have an OCR = 1 are normally consolidated (op=ovo), soils
that have an OCR > 1 are overconsolidated (ap>avo) and finally a soil that has
an OCR < 1 is considered underconsolidated (op Typical values have been tabulated for Ko and are shown in Table 3.1.
40


Table 3.1: Typical values of the coefficient of lateral earth pressure, Kq.
Soil Type K
Granular, Loose LO o I LO o
Granular, Dense 0.3-0.5
Clay, Soft 0.9- 1.1 (undrained)
Clay, Hard 0.8- 0.9 (undrained)
3.2 Ground Water and the Lateral Earth Pressure Coefficient
The discussion of lateral earth pressure from a natural soil perspective is
applicable only if the soil remains in the current state and water is not introduced
to the system. Since the presence of groundwater and specific groundwater
levels can fluctuate over time, the lateral earth pressure coefficient Ko is not a
constant for a soil deposit or for a given time period. One way to address this
variability is to express the lateral earth pressure coefficient in terms of effective
stress. The equation now becomes
a'h = K0a'v 3.9
This expresses the in situ soil stress state in terms of effective stresses to
allow for independence from the presence and level of the groundwater table. If
the water table level changes the lateral earth pressure coefficient K0 remains
constant as long as we remain in the current soil layer with a constant density
41


(Holtz & Kovacs, 1981). The coefficient of lateral earth pressure at rest, Ko, is
sensitive to many things including the density of the overlying soil layer(s) and
the geologic stress history of the soil itself. Where the soil is completely
submerged, soil is below the groundwater table, the intergranular (effective)
stress between soil particles is reduced by the magnitude of the water pressure
at that specific depth as follows:
where u is equal to the water pressure at the same depth. The presence of
water acts to reduce the amount of load the soil intergranular particles accept by
taking on some of the load itself. For a normally-consolidated soil that is located
above the groundwater table (water below depth H, Figure 3.5) the lateral at-rest
pressure acting against the wall would increase with depth uniformly. The
resulting distribution (Figure 3.5) would be represented by a triangle with the
maximum pressure existing at the base of the triangle and equal to
0" v YsoilZ H
3.10
&h ~ = KoYsoilZ
3.11
and the resultant lateral force per unit of wall length is
1
Pq ~ YsoilH
2
3.12
42


Structural
Basement Wall
Figure 3.5: Subsurface Stresses Soil in At-rest Condition
In the case where the soil is below the groundwater table, the
intergrannular or effective stress between the soil particles is reduced by the
hydrostatic pressure of the water taken at the same depth
(Tv YsoilZ U (Tv U 3.13
Where u is the magnitude of the water pressure at depth Z. The net effect of the
soil being within the water table (submerged) is that compared to a non-
submerged soil given the same conditions, the lateral soil pressure is less.
43


However, since the soil is located below the water table (submerged) you have to
account for the hydrostatic pressure of the water on the wall. The total effect on
the wall lateral pressure due to the hydrostatic pressure and the lateral soil
pressure imposes a larger total lateral force than a non-submerged soil (Figure
3.6).
q = Vertical Surcharge
' , ^ ^ j j | |
A J
h 1 'V UnitWeight(y) < \ \ > z f
> f V
> < \ Saturated Unit
"A Weight (Ysat)
1-
2
f < \ <
Lateral Earth Pressure
Due to Soil Above and
Below Water Table
Lateral Earth
Pressure Due
to Surcharge (q)
Lateral Earth
Pressure Due
to Water Table
Figure 3.6: Subsurface Stresses Soil in At-rest Condition with Vertical
Surcharge and Influenced by the Water Table
44


4. Rigid Wall Foundation Design
The purpose of a reinforced concrete basement wall is multifold. It
retains or holds in place the wall backfill, which tends to push inward on the
basement of the structure, and supports the structural loads which typically react
vertically. Design information for basement foundations is scattered across
multiple sources leaving design engineers to navigate and assimilate the
appropriate information for a complete design. Most of the sources used today
use prescriptive design practices and are related back to the Unified Soil
Classification system for backfill loads. Typically, the design of basement walls
uses a combination of the International Building Code (IBC), the International
Residential Code (IRC) and the American Concrete Institutes (ACI) 318, Building
Code Requirements for Structural Concrete.
4.1 International Building Code
The International Building Code (IBC), within multiple sections, outlines a
prescriptive approach to basement wall design for walls that are supported at the
top (floor diaphragm) and at the bottom (Keyed or doweled into a footing).
Section 1610, Table 1610.1 provides soil lateral loads based on the Unified Soil
Classification (USC) system. If the top of the wall is restrained from horizontal
movement the design uses an at-rest pressure. Surcharge loading is to be
included in the analysis and if expansive soils are a concern a note indicates that
the lateral pressure may increase. Section 1802 outlines foundation and soils
investigations with some discussion on expansive soils.
45


Section 1802.3.2 outlines the provisions for classifying an expansive soil
as follows:
1) Plasticity Index (PI) greater than or equal to 15 in accordance with
ASTM D4318
2) Greater than 10% of the soil particles passing a #200 sieve (75uM) in
accordance with ASTM D 422
3) Greater than 10% of the soil particles less than 5uM in size in
accordance with ASTM D4829
4) The Expansive Index greater than 20 in accordance with ASTM D4829
One caveat is included in the criteria that if #4 is true, the provisions described in
#1, #2 and #3 are not required. Section 1804 outlines the load bearing
pressure, lateral pressure and lateral sliding resistance capacity for soils (Table
1804.2) using the UCS system. Section 1805.2 outlines footing options and
depths and section 1805.5 continues with the foundation wall design for concrete
and masonry wall designs. Table 1805.5 (1) provides the minimum wall
thickness for plain concrete and plain masonry walls. Tables 1805.5 (2,3,4)
provide the vertical reinforcement requirements for 8-inch, 10-inch, and 12-inch
wall thicknesses based on an unbalanced backfill and three groupings of soil
types based on the UCS system. Included in the table are reinforcement
location, grade of reinforcement, minimum concrete strength and alternate sizes
of vertical reinforcement available for use. The provision of the code gives the
required amount of vertical reinforcement to resist lateral forces or pressures
46


given a general soil type. The design approach is based on a one-way behavior
of the basement wall. The wall acts as a vertical beam pinned at the top and
bottom with the beam length equal to the height of the wall and the beam width
equal to the length of the foundation wall. No explicit benefit is given to support
at corners or bump-outs. The vertical pinned-pinned beam design assumes a
triangular soil loading profile, exerting pressure on the exterior surface of the
basement wall. The resulting beam reacts in tension on the interior, unsupported
wall and compression on the exterior of the beam. Reinforcing steel is
prescribed on the interior side of the beam due to the tensile forces present at
that location. The footing is assumed to provide no rotational resistance and is
designed primarily to resist the gravity loading of the structure and pinned by
way of the floor. As part of the prescriptive approach, reinforcement
requirements are given for various unbalanced backfill heights reflecting the
amount of backfill being retained. Section 1805.8 addresses foundations in
expansive soils but is limited to the design of slabs-on-grade and mitigation of the
soils as opposed to the actual design of a structure to resist the forces applied.
Section 1907 describes the details of the reinforcement to include hooks, the
placement of reinforcement, protection, etc. and primarily references ACI 318 for
the actual design details.
4.2 International Residential Code
The International Residential Code, Chapter 4, addresses foundation
design for one and two family dwellings. Section R401.4.1 references Table
R401.4.1 where presumptive load-bearing values for foundation materials are
47


given. Values in the table are based on the UCS system. Section R401.4
includes a note to direct soils testing for expansive, compressible, shifting or
other unknown soil conditions without any distinct criteria specified. Section
402.2 outlines the minimum specified compressive strength of concrete for the
type or location of the concrete construction against the weathering potential for
the application. Section R403 continues with the sizing of the footings, described
in section R403.1.1, and includes the minimum width of concrete or masonry
footings in Table R403.1(1). Expansive soils, as related to the footing, are
addressed in section R403.1.8 by referencing the International Building Code
(IBC), section 1805.8. It gives the same provisions for classifying an expansive
soil as defined in section 1802.3.2 of the IBC, namely testing for the Plasticity
Index, soil fines and the Expansion Index. Section R404 outlines prescriptive
solutions for the top reactions for a foundation wall based on UCS system for
varying soil unbalanced backfill heights (Table R404.1(1)) and the maximum
plate anchor-bolt spacing based on unbalanced backfill height using the UCS
system (Table R404.1(2)). The section also specifies the maximum aspect ratio
(l/W) for unbalanced backfill height based on the UCS system (Table R404.1(3)).
Concrete foundation wall minimum vertical reinforcement size and spacing is
addressed in Table R404.1(5) and is based on the unbalanced backfill height and
the soil classification from the UCS system. Some of the vertical reinforcement
design is referenced to the ACI 318 spec. The design approach outlined in the
IRC is also based on a one-way behavior of the basement. The wall acts as a
vertical beam pinned at the top and bottom with the beam length equal to the
48


height of the wall and the beam width equal to the length of the foundation wall.
Similar to the IBC, no explicit benefit is given to additional support at corners or
bump-outs. The vertical pinned-pinned beam design assumes a triangular soil
loading profile, exerting pressure on the exterior surface of the basement wall.
4.3 American Concrete Institute
The American Concrete Institute ACI 318 (2005), Building Code
Requirements for Structural Concrete, does not specifically address the design of
basement foundation walls as a single design section. It must be developed
through the use of multiple sections. The design process typically begins with
the definition of the structural loads outlined in Chapter 9 for various load
combinations. Lateral soil loads are not addressed and must be determined
using other geotechnical sources. Chapter 14 of the code specifically addresses
the design and analysis of concrete walls. The design must satisfy sections 14.2
(general requirements) and 14.3 (minimum reinforcement requirements) plus
14.4, 14.5 or 14.8, which provide design methods (only one of which is used in a
given design problem). The shear design of the wall must satisfy requirements in
section 11.10 (special provisions for walls) requirements. Vertical and horizontal
reinforcement is designed in accordance with Chapter 14, section 14.4 for walls
designed as compression members,
All of the design literature requires that the top and bottom of the wall be
restrained (pinned) and to satisfy the prescriptive design aids it must act as a
pinned-pinned connection. This is accomplished at the base by means of a
concrete slab (floor) opposing movement of the base of the wall. In addition, the
49


top is restrained by the floor diaphragm by means of the connection of the sill
plate to the concrete wall, the connection of the sill plate to the floor framing and
the framing and stiffness of the floor diaphragm itself.
50


5. New Foundation Design Approach
Two structural designs were used in this analysis. The first was a
rectangular structure (Figure 5.1) that represents a typical two-story, single family
home. The structure measures 30 feet on its side by 50 feet in length. The wall
maximum height was chosen as 9 feet with a maximum unbalanced backfill
height of 8 feet. The thickness of the structural wall was chosen to be 12 inches
which meets the International Residential Codes recommendation given backfill
conditions, wall height and the soil pressures exerted on the structure. The
backfill soil was assumed to be CL at 60 psf, per foot of depth. No allowance
was given for granular fill against the structural basement wall or drainage for the
soil against the exterior wall.
Figure 5.1: Rectangular Structure and Soil Backfill
51


The second structure was chosen to be a curvilinear structure in the
shape of an ellipsoid (Figure 5.2). The minor axis was chosen to be a 15 foot
radius and the major axis was chosen to be a 25 foot radius. This was to model
the curvilinear structure as close to the same dimensions as the rectangular
structure. The same wall thickness of 12 inches was used and the height of the
structural wall was also 9 foot high. The soil backfill type, height and drainage
was also chosen to be the same as the rectangular structure at CL at 60 psf per
foot of depth, 8 foot backfill height and no drainage, respectively.
Figure 5.2: Curvilinear structure and soil backfill
The following assumptions were made in the design of the structures:
Each structure has a full basement floor.
3 inches thick and tight against the bottom of the foundation with
adequate stiffness to model a pinned joint.
52


Joists are used at the top of the wall and connected in such a
manner to provide for full-span support of the top of the wall the
top ssurface acts as diaphragm and is the pinned top support for
wall.
The lateral soil loading is identical on all sides of the structure (no
unbalanced loading).
Structure resides in a low seismic region and therefore seismic
forces do not control the design.
The backfill against the structure on all walls is horizontal no
sloped backfill.
The total service-level vertical load on wall = 1.5 kips/ft.
53


6. Finite Element Analysis Foundation
6.1 LS DYNA Model
LS-DYNA is a general-purpose finite element program with the ability to
simulate highly nonlinear, transient dynamic finite element analysis using explicit
or implicit time integration for complex real-world problems. The software was
developed by the Livermore Software Technology Corporation (LSTC),
Livermore California. The LS-DYNA code easily handles highly nonlinear,
transient, dynamic finite element analysis using explicit time integration. Being a
nonlinear code it handles changing boundary conditions, large deformations
and nonlinear materials that do not exhibit ideally elastic behavior. In addition,
the code handles transient dynamic" problems analyzing high speed, short
duration events where inertial forces are important. The software code is
especially applicable to soil-structural interaction problems involving nonlinear
soil materials with potentially large deformations.
LS-DYNA (version 971, Revision 7600.1224) contains numerous
concrete material constitutive models that can be used for this analysis. When
little is known about the concrete material properties, options to generate
material constants given the unconfined compressive strength as an input can be
very useful (e.g. MAT 159 CSCM). In this research, MAT Elastic 001 was
selected for both the concrete wall and soil due to the ability to define Youngs
modulus, the Bulk Density (Rho) and Poissons ratio. In addition, high strain rate
effects are not needed since loading of the concrete structure is achieved over
relatively long spans of time. Since great variability exists in the field when
54


constructing concrete foundations, defining concrete properties based on finely
detailed input parameters is not warranted.
6.2 Model Configuration
The foundation model was fashioned after a typical residential
concrete foundation designed in accordance with Chapter 4 of the 2006
International Residential Code. The foundation was designed with a wall
thickness of 12 inches and a height of 9 feet (108 inches). Two foundation types
were modeled, a typical rectangular structure (Figure 6.1) with a length of 50 feet
(600 inches) and a width of 30 feet (360 inches) and a curvilinear structure
(Figure 6.2) shaped in the form of an ellipse with a major diameter of 25 feet (300
inches) and a minor diameter of 15 feet (180 inches).
55


Figure 6.2: Curvilinear Structure Finite Element Model
The curvilinear structure was designed to fit within the same footprint of
the rectangular structure. The designs also included an escape window as
defined in the 2006 International Residential Code, with a minimum of 5.7 ft2
opening, 24 inch minimum height and a 20 inch minimum width. The window
was placed in the center of the longest wall to evaluate the stress state of the
elements around the window. The soil backfill was modeled after a clay soil
(Unified Soil Classification System of CH) with a backfill height of 8 feet (96
inches) around the structure with uniform depth of fill around the entire structure.
The unbalanced backfill height as defined in the 2006 International Residential
Code was calculated as 8.0 ft. The top of the concrete foundation was modeled
as a pinned joint, supported by a fixed diaphragm (framed floor) at the top of the
wall structure. Motion was restricted in the plane of the wall in the lateral
direction but allowed to move relative to the vertical. No rotational limitations
56


were imposed on the structures. The bottom of the wall was restrained by a
continuous 3 inch thick, cast-in-place concrete floor at the base of the foundation.
The wall is restrained on the exterior by the soil mass. It was modeled as a
pinned joint with rotational freedom at the upper and lower joints. Reinforcement
used in the models was based on minimal vertical reinforcement outlined in the
American Concrete Institute Building Code Requirements for Structural
Concrete), Chapter 14. This outlines a minimum spacing of 18 inches on-center
for #4 vertical reinforcement. This results in an area of steel of 0.093% for a #4
bar. In the models, the incorporation of reinforcement was accomplished by
using the Rule of Mixtures to define the composite properties rather than
incorporating discrete beam elements in the model to account for the
reinforcement. The models were limited to 10,000 total elements due to the
licensing of the LS DYNA software.
6.2.1 Element Types
All models and parts used constant-stress solid brick elements for the
concrete foundation and soil backfill material. The LS DYNA default solid
element uses a single integration point. In addition, the model used two discrete
parts for the concrete structure and the soil backfill.
6.2.2 Loading and Boundary Conditions
Two types of loading were used in the modeling of the concrete
foundations, a vertical structural load and an earth-pressure load on the
structure. The vertical structural load was applied to the nodes at the top of the
57


foundation over a time interval of zero to 0.1 seconds and continuing constant to
10 seconds while applying the full vertical load (Figure 6.3).
Time Seconds
Rectangular
Foundation
Node Loading
Curvilinear
Foundation
Node Loading
Figure 6.3: Structural Loading of Rectangular and Curvilinear Foundations
The specific load applied to the rectangular foundation nodes was 1500
Ib/lf and was determined for each node by the following equation:
Applied Load =
1500
/ lbs
312 nodes ~ 750 node
156 If
6.1
58


The specific load applied to the curvilinear foundation nodes was 1500
Ib/lf and was determined for each node by the following equation:
Applied Load =
332 nodes
129. 5 If
585.1
lbs
node
6.2
The soil loading was accomplished using two approaches.
1. The first was the application of a gravity load to the soil mass surrounding
the foundation utilizing the Load_Body_Z card. This imposes a body force
due to gravity on the soil mass. The acceleration factor used was 1g or
32.2 ft/s2 (386.4 in/s2) in the model.
2. The second soil loading was the application of a lateral load to the soil
mass in the X and Y directions to simulate the pressure applied by an
expansive soil. The soil mass was extended 10 feet from the foundation
walls on all sides of the foundation. This facilitated the placement of
lateral loads on the structure in the X and Y directions which was
important for the load application on the foundation walls. The applied
load in both the X and Y directions was 5,000 psf.
Both loading conditions were applied from zero load to max loading at 4.0
seconds and continuing constant to 10 seconds to ensure stability of the loading
(Figure 6.4).
59


The boundary conditions used for the models included constraining
translational movement (X, Y and Z directions) for all of the soil and concrete
foundation nodes located at the bottom of the models. This ensures that the
foundation and soil cannot move due to the applied loading. In addition, the top
of the concrete foundation was constrained in the X and Y direction to model the
Joists used at the top of the wall and connected in such a manner to provide for
full-span support of the top of the wall the joists and floor acts as diaphragm
and is the pinned top support for wall.
60


6.2.3 Contact Type
LS DYNA defines a contact by identifying (using parts, part sets, segment
sets, and/or node sets) what is to be checked for potential penetration of a slave
node through a master segment. To accomplish this, LS DYNA searches for
penetrations, using a number of different algorithms, at each time-step
throughout the analysis. When a penetration is found (penalty-based contact) a
force proportional to the penetration depth is applied to resist, and ultimately
eliminate, the penetration. In this analysis a two-way treatment of contact was
chosen between the concrete wall and soil. This type of contact calls the
subroutines twice which checks the slave nodes for penetration the first time and
checks the master nodes for penetration through the slave segments the second
time. The treatment is therefore symmetric and the definition of the slave surface
and master surface is arbitrary since both are called in the subroutine and the
results end up the same. This results in an increase of computation time due to
the extra subroutine calls.
The interface between the concrete wall and soil was modeled using the
Surface_To_Surface contact card. The objective of the contact definition is to
eliminate any penetration between the interfacing surfaces. The Slave Segment
Set and Master Segment Set were set up to use the Part I.D. with the Master
defined as the Soil. The Static Friction coefficient was defined as 0.4 and the
Dynamic Friction coefficient was defined as 0.2 (ETL 1110-3-446, 1992,
Department of the Army). The friction coefficients were based on a wet clay soil
against a troweled concrete surface. This is a reasonable assumption since
61


concrete forms are smooth and similar to a troweled surface. There is general
agreement on these values in the existing literature although the friction
coefficients are dependent on the type of clay, moisture content, density and the
relative surface condition of the concrete.
6.2.4 Material Properties
The concrete properties used for this analysis were based on a normal
concrete with a compressive strength of 3000 psi, density of 150 lb/ft3, Poissons
Ratio of 0.16 and a Modulus of Elasticity of 3.146 x 106psi. In addition, the
reinforcing steel used in the model had a minimum yield strength of 60,000 psi
and a Modulus of Elasticity of 29 x 106 psi.
The soil properties used for this analysis was based on a clay soil
(Unified Soil Classification System CH) with a Unit Weight of 100 lb/ft3, a
Density of 1.498 x 10'4 lb/in3, a Modulus of Elasticity of 2175 psi and a Poissons
Ratio of 0.3.
Table 6.1: LS DYNA model material properties
Material Name Rho Youngs Modulus Poissons Ratio
(lb/in3) (lb/in2)
Concrete/Steel 2.247E-4 3.146E+6 0.16
Composite
Soil 1.498E-4 2.175E+3 0.30
62


The concrete/steel properties were further combined to generate a
composite material using the Rule of Mixtures (Callister, 2001). The Rule of
Mixtures asserts that the properties of the combined material are a combination
of the individual components of the individual materials. Typical composites
have two phases consisting of a matrix (continuous) phase and a dispersed
(particulates, fibers) phase. The properties of the composite depends on the
specific properties of the phases, the geometry of dispersed phase (particle size,
distribution, orientation) and the amount of each phase in the composite.
Composites are typically classified as particle-reinforced (large-particle and
dispersion-strengthened) composites, fiber-reinforced (continuous (aligned) and
short fibers (aligned or random) composites and structural (laminates and
sandwich panels) composites. The rule of mixtures calculates an upper limit of
the elastic modulus of the composite in terms of the elastic moduli of the matrix
(Em) and the particulate (Ep) phases by the equation:
F = F V + F V
"c ljmv m Llpv p
6.3
where Vm and Vp are the volume fraction of the two phases. The calculated
composite elastic modulus for a 12 inch thick foundation wall with #4
reinforcement spaced at 18 inch on-center is:
63


Ecomposite ~ Econc
^cone ^
steel
total
T Esteel
V steel\
V total'
6.4
The resulting Modulus of Elasticity for the composite material used in the model
was 3,150 ksi.
64


7. Results
7.1 General Behavior of the Foundation Structures
A review of the structural behavior of the rectangular foundation
indicates the model is behaving as a pinned-pinned structural wall with combined
vertical and horizontal loading. As shown in Figure 7.1, the deformed shape
shows rigidity at the corners with a corresponding lack of displacement. It also
shows the greatest displacement at the center of the longest walls exhibiting
bowing to the interior due to the lateral pressure from the soil mass. The shape
of the deformation in Figure 7.1 is exaggerated (5X displacement factor) to
visually demonstrate the deformations.
Figure 7.1: Deformed Rectangular Basement Structure (5X Displacement Factor)
Comparing the deformations of the rectangular foundation wall with that
of the curvilinear foundation (Figure 7.2) under the same loading conditions, very
65


little deformation along the entire wall (5X displacement factor) is observed. The
load acts in a nearly radial direction and is distributed semi-uniformly around the
periphery of the structure. Truly radial forces acting on a circular cross-section
result in only compressive forces on the section. There are no bending forces in
a completely circular section unless there is a discontinuity such as a window or
door. The modeled curvilinear structure capitalizes on this loading advantage by
balancing constructability and space utilization of the structure with a shape that
takes advantage of compressive forces to increase the load capacity of the
structure.
LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Effective Stress (v-m)
min=0.306182, at elem# 6356
max=40.6953, at elem# 864
max displacement factor^S
z
Fringe Levels
4.070e+01
3.934e+01 ]
3.798e+01 1
3.663e+01 ]
3.527e+01 J
3.391e+01 j|
3.256e+01 _
3.120e+01 _
2.984e+01 _
2.849e+01 _
2.713e+01 _
2.577e+01 _
2.442e+01 _
2.306e+01 _
2.170e+01 _
2.035e+01 _..
1.899e+011
1.763e+01 _
1.628e+01 _
1.492e+01 _
1.357e+01 _
1.221e+01 _
1.085e+01 _
9.496e+00 _
8.139e+00
6.783e+00 ]
5.426e+00 ]
4.070e+00 J
2.713e+00l
1.357e+0oll
0.000e+O0li
Figure 7.2: Deformed Curvilinear Basement Structure (5x Displacement Factor)
The finite element model indicates that a nearly compressive shape can easily be
applied to basement foundation design.
66


7.2 Wall Displacements of the Foundation Structures
The results of the finite element analysis demonstrate a considerable
difference in the mid-wall, out-of-plane lateral displacements between the
rectangular basement wall and the curvilinear basement wall (Figure 7.3).
Node no.
A 2842
B 2844
C 1780
D 1782
Figure 7.3: Lateral Deflection of Rectangular Basement Structure
The maximum displacement experienced by the rectangular wall was
2.67 inches inward at the top of the window. Similarly, the corresponding node
on the opposite wall (without window) shows a deflection of 2.23 inches which
was the second greatest deflection in the model. The magnitudes of all four
elements analyzed at the mid-span of the longest wall were very similar as
shown in the fringe plot of the rectangular structure (Figure 7.4).
67


LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Y-dispiacement
min=-2.67023, at node# 1637
max2.22914, at node# 2771
a.
Fringe Levels
2.229e+00
2.066e+00]
1.903e+00 1
1.739e+00 T
1.576e+00 J
1.413e+00 Ji
1.249e+00 _
1.086e+00 _
9.226e-01 _
7.593e-01 _
6.960e-01 _
4.327e-01 _
2.694e-01 _
1.061e-01 _
-5.723e-02 _
-2.20Se-01 __
-3.839e-01 _*
-5.472e-01 _
7.105e-01 _
8.738e-01 _
-1.037e+00 _
-1.200e+00 _
-1.364e+00_
-1.527e+00 _
1.690e+00
1.854e+0o]
-2.017e+Ool
-2.180e+00 J
-2.344e+00ll
2.507e+00 II
2.670e+00 II
Figure 7.4: Lateral Deflection of Rectangular Basement Structure Fringe Plot
Comparing the results of the rectangular wall design to the curvilinear
wall design there are differences to note. The deflections are essentially zero
inches (Figure 7.5) for mid-span elements at the window and on the opposite wall
of the curvilinear structure. These locations are representative of the same
locations analyzed for the rectangular structure.
c LS-DYNA KEYWORD DECK BY LS-PREPOST
O-i------------------------------------

<*) 9 n_ \ A

W ^ c 0) E (0 ,w Q. CO TjJ >
-20-
0 5 10 15
Time
Node no.
A 1100
B 1098
C 2681
D 2683
Figure 7.5: Lateral Deflection of Curvilinear Basement Structure
68


All of the displacements within the curvilinear structure were less than
0.1 inches, an order of magnitude less than the rectangular structure (Figure
7.6).
LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Y-displacement
min=-0.0648193, at node# 1137
Z
**Ay
Fringe Levels
2.771e-02
2.463e-02 l
2.154e-02 J
1.846e-02l
1.537e-02 J
1.2296-02 Jl
9.204e-03 _
6.120e-03_
3.035e-03_
4.883e-05_
-3.133e-03_
-6.217e-03 _
-9.302e-03 _
-1.239e-02_
1.547e-02_
1.8S5e-02_.
-2.1646-02 _.
-2.472e-02 _
-2.781e-02_
-3.089e-02_
-3.398e-02 _
-3.706e-02_
-4.014e-02_
-4.323e-02_.
4.631e-02
4.940e-02 J
-5.248e-02]
-5.557e-02l
-5.8656-02 II
4.174e-02ll
-6.482e.02 H
Figure 7.6: Lateral Deflection of Curvilinear Basement Structure Fringe Plot
It was also noted that the maximum deflections were located along the
top of the window, similar to the rectangular structure, although all deflections
were below 0.1 inches.
69


7.3 Vertical Stress State of the Foundation Structures
In addition to the deflection of the walls, the stress state of each
foundation type was analyzed. The Z-direction stress (vertical plane of the wall)
in the area of the window was analyzed between the two wall designs. The
results of the rectangular wall structure indicate a compressive stress of -12 psi
and -8 psi in the two elements above the window and increasing larger stresses
from 0 psi at the window sill to 66 psi at the bottom of the foundation, all in
tension (Figure 7.7).
Midspan Window Elements Center of Window Vertically
F p
y X' /'
(A (A 0) on yS' D
N ^ / r IRC C
20 1 u £-a_ A ~ 1
,
0 5 10 15
Time
Element no.
A 858
B 857
C 853
D 852
_E_851
F 850
Figure 7.7: Z-Stress (Vertical) Plot of Window Area of Rectangular Foundation
The maximum stress in the rectangular structure is 412.9 psi in tension
(Figure 7.8). This occurs at the lower corners of the foundation.
70


Figure 7.8: Z-Stress (Vertical) of Window Area of Rectangular Foundation -
Fringe Plot
Comparing the results of the rectangular wall design to the curvilinear
wall design the wall stresses are smaller in magnitude with a maximum stress for
the Elements at the window opening ranging from -10.9 psi to 1.8 psi (Figure
7.9).
Element no.
_A_540
B 541
C 542
D 543
_E_547
F 548
Figure 7.9: Z-Stress (Vertical) Plot of Window Area of Curvilinear Foundation
71


In addition, the curvilinear structure has a maximum stress at -38.2 psi
located at the upper right side of the window cut out. Of greater interest is that all
of the vertical stresses in the curvilinear structure are compressive stresses
(Figure 7.10) with the exception of element #547 at the top of the window
opening. The greatest tensile stress is 1.9 psi which is a significant difference
from the rectangular wall design.
Fringe Levels
1.964e+00
6.248e-01
-7.145e-01
2.054e+00
-3.393e+00
-4.732e+00
-6.072e+00 _
7.411e+00_
8.750e+00 _
-1.009e+01 _
-1.1430+01 _
-1.277e+01 _
-1.411e+01_
-1.6460+01 _
-1.6790+01 _
-1.8130+01
-1.9470+01 _
-2.080e+01 _
-2.214e+01 _
2.348e+01 _
2.482e+01 _
-2.616e+01 _
-2.750e+01 _
-2.884e+01 _
-3.018e+O1
-3.162e+01 J
-3.286e+01 ]
-3.420e+011m
-3.554e+01 _1
-3.688e+01 J
-3.822e+01
Figure 7.10: Z-Stress (Vertical) of Window Area of Curvilinear Foundation -
Fringe Plot
7.4 Shear Stress State of the Foundation Structures
The shear stress for the rectangular structure was evaluated against the
curvilinear structure. The greatest shear stress in the rectangular structure was
determined to be 235.4 psi (Figure 7.12) located toward the upper corner of the
structure.
LS-DYNA KEYWORD DECK BY LS-PREFOST
Time = 19.99
Contours of Z-stress
min=-38.2157, at elem# 573
max=1.96417, at elem# 556
4
72


Element no.
A 894
B 893
C 892
D 891
E 890
F 889
G 888
H 887
I 886
Figure 7.11: Shear Stress Plot along Window Area of Rectangular Foundation
The highest shear stresses are located in the corner areas where the
walls transition from X to Y directions. The lowest shear stress in the rectangular
structure was 1.8 psi located on the upper side-wall of the rectangular
foundation.
73


LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Maximum Shear Stress
Fringe Levels
2.354e+02
2.276e+021
2.199e+02 1
2.121e+02l
2.043e+02 J
1.965e+02 ji
1.887e+02_
1.809e+02_
1.731e+02_
1.653e+02_
1.576e+02_
1.498e+02_
1.420e+02_
1.342e+02_
1.264e+02_
1.186e+02_
1.108e+02_
1.030e+02_
9.525e+01 _
8.746e+01 _
7.968e*01 _
7.189e+01 _
6.410e+01 _
5.631e+01 _
4.853e+01 ~
4.074e+011
3.295e+01 J
2.516e+01 _
1.738e+01 II
9.588e+O0ll
1.8006+00
Figure 7.12: Shear Stress along Window Area of Rectangular Foundation -
Fringe Plot
Comparing the rectangular foundation to the curvilinear foundation the
maximum shear stress for the curvilinear structure is 21.2 psi (Figure 7.14)
located along the bottom side of the foundation while the minimum shear stress
is 1.4 psi and is located above the window.
74


Element no.
A 522
B 523
C 524
D 525
E 526
F 527
G 528
H 529
i 530
Figure 7.13: Shear Stress Plot along Window Area of Curvilinear Foundation
The difference between the maximum shear stress in the rectangular
structure and the curvilinear structure is approximately a factor of ten with the
highest shear stresses exhibited in the rectangular structure.
LS-DYNA KEYWORD DECK BY LS-PREPOST
Time = 19.99
Contours of Maximum Shear Stress
min=1 .41687, at elem# 547
max=21.203, at elem# 864
Fringe Levels
2.120e+01
2.054e+01
1.988e+01
1.922*01
1.856e+01
1.791e+01
1.725e+01
1.6590+01
1.593e+01
1.527e+01
1.461e+01
1.395e+01
1.329e+01
1.263e+01
1.197e+01
1.131e+01
1.065e+01
9.991e+00
9.331e+00
8.672e+00
8.012e+00
7.353e+00
6.693e+00
6.034e+00
5.374e+00
4.715e+00
4.055e+00
3.395e+00
2.736e+00
2.076e+00
1.417e+00
I
j
Figure 7.14: Shear Stress along Window Area of Curvilinear Foundation Fringe
Plot
75


8. Discussion of Analyses Results
Results from the analyses of the rectangular and curvilinear structural
designs were analyzed and compared to determine the performance advantages
of a curvilinear foundation in environments with high lateral soil loading in
addition to the vertical structural loading. This included comparing the lateral
displacements of the walls, the vertical stress performance, the shear stress
performance and the stress concentration distribution around a discontinuity
(window) in each design with combined vertical and horizontal loading.
8.1 Displacements of Foundations
The first observation regarding the lateral deflection of each design was
the lateral displacements of the rectangular wall design were greater than 100X
of the displacements for the curvilinear design for the same locations. Both
foundations experienced the greatest lateral deflections at mid-wall of the long
span section of the foundation. Four locations were studied for each design, two
around the window and two opposite the window on the solid, continuous long
wall.
76


Table 8.1: Lateral displacement (in) of Rectangular and Curvilinear Wall Designs
YWall Displacement Results
Structure Node ID Displacement (in) Description
Rectangular Node 1780 -2.3 Lower window
Node 1782 -2.7 Upper window
Node 2842 2.1 Lower Adjacent Wall
Node 2844 2.2 Upper Adjacent Wall
Curvilinear Node 2681 -0.0009 Lower Adjacent Wall
Node 2683 -0.0002 Upper Adjacent Wall
Node 1098 -0.018 Lower window
Node 1100 -0.0023 Upper window
This is a significant performance difference between the two structural
designs and would be advantages from a concrete cracking performance
standpoint. This might also have the added advantage of using less reinforcing
steel in the design providing better economy of construction, although this would
have to be further modeled and analyzed. In addition, the maximum
displacement for the rectangular foundation occurred at the right, upper corner of
the window at node 1637 and measured 2.67 inches while the maximum
displacement for the curvilinear structure occurred at node 1137 and measured
0.06 inches at a location above the window (Figures 8.1 and 8.2).
77


LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Y-dispiacement
min=-2.67023, at node# 1637
maxn2.22914, at node# 2771
a.
Fringe Levels
2.229e+00
2.066e+00 J
1.903e+00l
1.739e+00 T
1.576e+0ol
1.413e+00 J
1.249e+00 _
1.086e+00 _
9.226e-01 _
7.593e-01 _
6.960e-01 _
4.327e-01 _
2.694e-01 _
1.061e-01_
-5.723e-02 _
-2.20Se-01
-3.839e-01 J
-5.472e-01 _
7.105e-01 _
-8.738e-01 _
-1.037e+00 _
-1.200e+00 _
-1.364e+00_
-1.527e+00 _
1.690e+00
-1.854e+00j
-2.017e+Ool
-2.180e+00 J
-2.344e+00ll
2.507e+00 II
2.670e+00 !
Figure 8.1: Lateral Deflection of Rectangular Basement Structure Fringe Plot
LS-DYNA KEYWORD DECK BY LS-PREPOST
Time= 19.99
Contours of Y-displacement
min-0.0648193, at node# 1137
Z
X-i^Y
Fringe Levels
2.771e-02
2.463e-02
2.154e-02
1.846e-02
1.537e-02
1.2296-02
9.204e-03
6.120e-03
3.035e-03
!
-3.133e-03_
-6.217e-03 _
-9.302e-03 _
1.239e-02_
-1.547e-02 _
-1.8556-02
-2.164e-02_
-2.472e-02 _
-2.781e-02_
-3.089e-02_
-3.398e-02_
-3.706e-02
-4.014e-02
-4.323e-02
-4.631e-02
-4.9406-02
-5.248e-02
-5.557e-02
-5.865e-02
-6.174e-02
-6.482e-02
I
Figure 8.2: Lateral Deflection of Curvilinear Basement Structure Fringe Plot
The results of the lateral displacement analysis demonstrates that the
curvilinear foundation design is a much better design regarding lateral deflections
78


under combined vertical and lateral loading. All of the deflections for the
curvilinear structure were much smaller than the rectangular design with worst-
case deflections of the rectangular structure 100 times the magnitude of the
curvilinear design.
8.2 Vertical Stress of Foundations
In addition, the vertical stress performance was analyzed in both the
rectangular and curvilinear models. A section of wall was analyzed through the
depth of the wall located in the middle of the window opening. The results of the
analysis demonstrate that the rectangular structure experiences both tensile and
compressive forces within the structure due to the combined vertical and lateral
loading (Table 8.2). The top of the foundation remains in compression above the
window opening and at the window sill, while the bottom of the foundation below
the window remains in tension.
Table 8.2: Vertical Stress of Rectangular and Curvilinear Wall Designs
Z Stress Results
Structure Element ID Stress (psi) Description
Rectangular Element 858 -12.2 Top of Foundation
Element 857 -7.2
Element 853 -0.2
Element 852 37.1
Element 851 66.2
Element 850 60.9 Bottom of Foundation
Maximum Z-stress in structure -412.8
Curvilinear Element 548 -3.5 Top of Foundation
Element 547 1.8
Element 543 -0.5
Element 542 -5
Element 541 -9.5
Element 540 -10.8 Bottom of Foundation
Maximum Z-stress in structure -38.2
79


Comparing the performance of the rectangular structure to the curvilinear
structure it is noted from the analysis that the curvilinear structure remains in
compression through the depth of the foundation at the window, with the
exception of the top of the window which is only slightly in tension at 1.8 psi. In
addition, the maximum vertical stress in the rectangular structure is located at the
left lower corner of the structure at element 1284 and measured 412.9 psi while
the maximum vertical stress of the curvilinear structure, located at element 573,
in the upper right corner of the window measured 38.2 psi. This represents a
difference of greater than 10X in the maximum vertical stress level in the
rectangular structure (Figures 8.3 and 8.4) as compared to the curvilinear
structure. The maximum vertical stress for both structures was a compressive
stress.
Fringe Levels
9.274e+01
7.589e+01
5.903e+01
4.218e+01
2.533e+01
8.472e+00
8.382e+00
-2.524e+01
4.2096+01
-5.894e+01
-7.580e+01
9.265e+01
-1.095e+02_
1.2648+02 _
-1.432e+02_
-1.6018+02 _
1.769e+02_
1.938e+02_
-2.1068+02 _
-2.275e+02 _
-2.443e+02 _
-2.612e+02_
-2.780e+02 _
-2.949e+02 _
-3.118e+02
-3.286e+02]
-3.456e+02 J
-3.623e+02 J
3.792e+02
-3.960e+02 J
4.129e+02^l
LS-DYNA KEYWORD DECK BY LS-PREPOST
Time = 19.99
Contours ofZ-stress
mln=412.876, at elem# 1284
max=92.7412, at elem# 800
JL
Figure 8.3: Z-Stress (Vertical) of Window Area of Rectangular Foundation -
Fringe Plot
80


LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Z-stress
min=-38.2157, at elem# 573
max-1.95417, at elem# 556
4,
Figure 8.4: Z-Stress (Vertical) of Window Area of Curvilinear Foundation Fringe
Plot
Focusing in on the discontinuous area created by the incorporation of the
window in both the rectangular structure and the curvilinear structure, the higher
vertical stresses peak along the side of the window for the rectangular structure
and at the upper corner of the window for the curvilinear structure (Figures 8.5
and 8.6).
81


Figure 8.5: Vertical Stress around Window Rectangular Structure
82


Figure 8.6: Vertical Stress around Window Curvilinear Structure
The maximum vertical stress in the area around the window for the
rectangular structure was 92.7 psi while the maximum vertical stress for the
curvilinear structure for the same area was 38.2 psi. The rectangular structure
around the window was primarily in tension while the curvilinear was in
compression. Evaluating the stress concentration factors in that same area
around the corner of the window (Table 8.3 and Table 8.4) for both the
rectangular and curvilinear designs the stress concentration factors for the
rectangular structure range from 0.88 to 190.18 while the curvilinear structure
83


ranged from 1.04 to 2.71. This represents a huge difference on structural
performance relative to the applied vertical stress around the discontinuity of the
window opening.
Table 8.3: Stress Concentration Factors around Window Rectangular
Rectangular Foundation Design Vertical Stress
Element ID Window Z-Stress (psi) Element ID No Window Z-Stress (psi) Stress Concentration Factor
875 23.5 1406 -7.9 3.97
857 -7.2 1388 -7.9 0.91
839 -6.85 1370 -7.8 0.88
821 23 1352 -7.8 3.95
803 21.7 1334 -7.7 3.82
802 17 1333 -5.5 4.09
801 17.2 1332 -3 6.73
800 92.7 1331 -0.49 190.18
Table 8.4: Stress Concentration Factors around Window Curvilinear
Curvilinear Foundation Design Vertical Stress
Element ID Window Z- Stress (psi) Element ID No Window Z- Stress (psi) Stress Concentration Factor
538 -13.4 1339 -12.9 1.04
547 1.8 1330 -13.1 1.14
556 1.9 1321 -13 1.15
565 -14.2 1312 -13.3 1.07
574 -27.3 1294 -12.6 2.17
573 -38.2 1293 -14.1 2.71
572 -25.9 1292 -14.9 1.74
571 -25.4 1291 -15.9 1.60
The performance improvement around the window in the curvilinear
foundation demonstrates the advantage of designing a foundation to remain in
84


compression throughout the structure and the improvement in performance of
crack initiation around a high stress concentration area.
8.3 Shear Stress of Foundations
When comparing the shear stress performance of the rectangular and
curvilinear structures, a vertical section of wall was selected along the side of the
window for each structural design. Overall, the rectangular structure exhibited
much greater shear stresses for the same locations in the structure, as much as
10X for some locations (Table 8.5).
Table 8.5: Shear Stress of Rectangular and Curvilinear Wall Designs
Shear Stress Results
Structure Element ID Stress (psi) Description
Rectangular Element 894 89.6 Top of Foundation
Element 893 149.6
Element 892 52.1
Element 891 34.5
Element 890 91.8
Element 889 145.4
Element 888 62.2
Element 887 32.7
Element 886 5.4 Bottom of Foundation
Maximum shear stress in structure 235.4
Curvilinear Element 530 6.8 Top of Foundation
Element 529 13.5
Element 528 20
Element 527 13.3
Element 526 12.4
Element 525 9.4
Element 524 12.4
Element 523 16.1
Element 522 7.6 Bottom of Foundation
Maximum shear stress in structure 21.2
85


The maximum shear stress for the rectangular structure was 235.4 psi
located at element 219 in the upper corner of the structure while the maximum
shear stress for the curvilinear structure was 21.2 psi and was located at element
864 on the bottom of the structure (Figures 8.7 and 8.8).
Fringe Levels
2.354e+02
2.276e+02l
2.199e+02ll
2.121e+02ll
2.043e+02 J
1.965e+02 J
1.887e+02_
1.809e+02_
1.731e+02_
1.653e+02_
1.576e+02_
1.498e+02_
1.420e+02_
1.342e+02_
1.264e+02_
1.186e+02_
1.108e+02_
1.030e+02_
9.525e+01 _
8.746e+01 _
7.968e+01 _
7.189e+01 _
6.410e+01 _
5.631e+01 _
4.853e+01
4.074e+01 II
3^96e+01 J
2.616e+01 !
1.738e+0l]
9.588e+00 ]
1.800e+00 H
Figure 8.7: Shear Stress along Window Area of Rectangular Foundation Fringe
Plot
86


LS-DYNA KEYWORD DECK BY LS-PREPOST
Time 19.99
Contours of Maximum Shear Stress
min=1.41687, at elem# 547
maxn21.203, at elem# 864
Fringe Levels
2.120e+01
2.054e+01 ]
1.988e+01 J
1.922b+01 J
1.8566+011
1.791e+01 H
1.7256+01 _
1.659e+01 _
1.693b+01 _
1.5276+01 _
1.4616+01 _
1.3956+01 _
1.3296+01 _
1.263e+01 _
1.197e+01 _
1.1316+01 _.
1.0656+01 J
9.991e+00_
9.331e+O0 _
8.672e+00 _
8.012e+00 _
7.353e+00_
6.693e+00_
6.034e+00 _
5.374e+00
4.715e+00j
4.056e+0ol
3.3956+00 M
2.736e+00 II
2.076e+00 II
1.417b+00 II
Figure 8.8: Shear Stress along Window Area of Curvilinear Foundation Fringe
Plot
Evaluating the shear stresses for the two structures around the area of
discontinuity at the window, the shear stresses are greatest above the window for
the rectangular structure and at the upper corner of the window for the curvilinear
structure (Figures 8.9 and 8.10).
87


Figure 8.9: Shear Stress around Window Rectangular Structure
88


Full Text

PAGE 1

THE PERFORMANCE OF A CURVILINEAR VERSUS A RECTANGULAR BASEMENT FOUNDATION DESIGN IN EXPANSIVE CLAY SOILS b y Michael James Gardiner B.A., The Colorado College, 1985 B.S., University of Colorado, Denver, 2001 A thesis submitted to the Faculty of the Graduate School of the University of Colorado i n partial fulfillment of the requirements for the degree of Master of Science Civil Engineering 201 3

PAGE 2

ii This thesis for the Master of Science degree b y Michael James Gardiner has been approved for the C ivil Engineering Program by Nien Y i n Chang Chair Brian T. Brady Yail J immy Kim April 10, 2013

PAGE 3

iii Gardiner, Michael J ames (Master of Science, Civil Engineering) The Performance of a Curvilinear Versus a Rectangular Basement Foundation Design in Expansive Clay Soils Thesis directed by Professor Nien Yin Chang A BSTRACT Shallow foundation design in expansive soils has generally been approached in the industry using a typical pier/beam or spread footing/foundation wall and reinforcement design. The addition of supporting piers anchored in stable soils and excavated expansion areas under beam elements have modi fied a traditional foundation design for expansive soils. This traditional approach to foundation design uses designs that mitigate around the swelling effects rather than designing to take advantage of or resist the imposed forces. This research define s a new shallow foundation design that uses a curvilinear structure to take advantage of the forces exerted on the foundation by the expansive forces of the soil. In addition, the design allows for a cast in place or precast implementation. The curviline ar foundation design was modeled using LS DYNA Finite Element analysis and compared to a traditional rectangular foundation design using 3D models In addition the soil and concrete models were evaluated by comparing the LS DYNA model results ( deflection shear, tensile / compression) for both foundation designs

PAGE 4

iv Al so included in the research is the analysis of a discontinuity (window) in the structural design and the effects of the discontinuity on the structure. The form and content of this abstract are approved. I recommend its publication. Approved: Nien Yin Chang

PAGE 5

v DEDICATION I dedicate this thesis to my wife Jerre and express my gratitude and appreciation for her support, encouragement and sacrifice throughout the extent of my studies. She has been the foundation on which I have built my life and I am truly blessed to have such a wonde rful partner. I would also like to dedicate this thesis to my father and mother, John and Viola Gardiner for instilling in me perseverance to complete what I have started and for their support. I would also like to include a special dedication to Jean D urham for her continuous encouragement throughout the many years of research and study.

PAGE 6

vi A CKNOWLEDGEMENT I would like to express my sincere appreciation and heartfelt thanks to my advisor Professor Nien Yin Chang for his continuous guidance, s upport and unwavering encouragement throughout my studies and the completion of this research. The catalyst for this research began with Dr. Chang during coursework inexhaustible pat ience and personal commitment allowed me the freedom and time to manage multiple priorities during this course of study. In the end I have the privilege of calling Dr. Chang my friend. I would like to express my gratitude to Dr. Brian Brady and Dr. Jimmy Kim for serving on my defense committee and providing me with their valuable inputs and comments to improve the content of this thesis.

PAGE 7

vii TABLE OF CONTENTS C hapter 1. Introduction 1 1.1 Purpose of the Study 2 1.2 Scope of the Study 2 1.3 Organizational Outline 3 2. Propert ies of Expansive Clay Soils 6 2.1 Introduc tion Expansive Soils 6 2.2 Soil Classification 7 2.2.1 Gradation of Soils 9 2.2.2 Atterberg Limits 13 2.2.3 Activity 1 5 2.3 Clay Minerals 1 8 2.3.1 Kaolinite 2 2 2.3.2 Halloysite 2 3 2.3.3 Montmorillonite/Smectite 23 2.3.4 Illite 24 2.3.5 Chlorite 25 2.4 Swell ing Potential of Clay Soils 27 2.4. 1 Consolidometer Swell Test 28 3. Late ral Forces on a Foundation 3 4 3.1 Lateral Earth Pressure 3 4 3.2 Ground Water and the Lateral Earth Pres sure Coefficient 4 1

PAGE 8

viii 4. Ri gid Wall Foundation Design 4 5 4.1 International Building Code 4 5 4.2 Int ernational Residential Code 47 4.3 American Concrete Institute 49 5. New Foundation Design Approach 5 1 6. Finite Element Analysis Founda tion 5 4 6.1 LS Dyna Model 5 4 6.2 Model Configuration 5 5 6.2.1 Element Types 57 6.2.2 Loading and Boundary Conditions 57 6.2.3 Contact Type 6 1 6.2.4 Material Properties 6 2 7. Results 6 5 7.1 General Behavior of the Foundation Structures 6 5 7.2 Wall Displacements of the Founda tion Structures 67 7.3 Vertical Stress Stat e of the Foundation Structures 7 0 7.4 Shear Stress Stat e of the Foundation Structures 7 2 8. Disc ussion of Analyses Results 76 8.1 Displacement s of Foundations 76 8.2 Ver tical Stress of Foundations 79 8.3 Shear Stress of Foundation 85 9. Conclusions and Recomm endations for Further Research 91 References 96

PAGE 9

ix LIST OF TABLES T able 2.1 Clay Activity (after McCarthy, 1998) 17 3.1 Typical values of the coefficient of lateral earth pressure K 0 4 1 6.1 LS DYNA model material properties 6 2 8.1 Lateral D isplacement (in) of Rectangular and Curvilinear Wall Designs 77 8.2 Vertical Stress of Rectangular and Curvilinear Wall Designs 79 8.3 Stress Concentration Factors around Window Rectangular 84 8.4 Stress Concentration Factors around Window Curvilinear 84 8.5 Shear Stress of Rectangular and Curvilinear Wall Designs 85 8.6 Stress Concentration Factors around Window Rectangular 90 8.7 Stress Concentration Factors around Window Curvilinear 90

PAGE 10

x LIST OF FIGURES F igure 2.1 Grain size Classification System After U.S. Army Waterways Experiment Station (1960) and Howard (1977) 9 2.2 Grain Size Distribution (after ASTM International, D2487 11) 11 2.3 Plasticity Index versus Liquid Limit (ASTM D2487 11, Standard Practice for Classification of Soils for Engineering Purposes) 15 2.4 Probable Clay Expansion as Estimated from Classification Test Data (after Holtz, 1959) 1 6 2.5 Characteristics of Common Clay Minerals (after Mitchell, 1976 ) 16 2. 6 Silica Tetrahedron and Silica Tetrahedral Molecules (after Grim, 1968) 20 2. 7 Alumina Octahedron and Alumina Octahedral Molecules (after Grim, 1968) 21 2. 8 Schematic Diagram of Kaolinite (after Lambe, 1953) 22 2. 9 Schematic Diagram of Montmorillo nite (after Lambe, 1953) 24 2. 10 Schematic Diagram of Illite (after Lambe, 1953) 25 2. 11 Schematic Diagram of Chlorite (after Mitchell, 1976) 26 2.1 2 Free Swell Oedometer Test Results 3 1 2.1 3 Correction for Sample Disturbance (Fredlund, et.al., 1980) 33 3 .1 At rest Earth Pressure 3 5 3.2 Wall Movement for Active Earth Pressure 3 6 3.3 Wall Movement for Passive Earth Pressure 3 7 3.4 Relationship between Vertical and Horizontal Soil Stress 38 3.5 Subsurface Stresses Soil in At rest Condition 4 3

PAGE 11

xi 3.6 Subsurface Stresses Soil in At rest Condition With Vertical Surcharge and Influenced by the Water Table 4 4 5.1 Rectangular Structure and Soil Backfill 5 1 5.2 Curvilinear structure and soil backfill 5 2 6.1 Rectangular Structure Finite Element Model 5 5 6.2 Curvilinear Structure Finite Element Model 5 6 6.3 Structural Loading of Rectangular and Curvilinear Foundations 58 6.4 Z Body Soil Loading 6 0 7.1 Deformed Rectangular Basement Structure (5X Displacement Factor) 6 5 7.2 Deformed Curvilinear Basement S tructure (5x Displacement Factor) 66 7.3 Lateral Deflection of Rectangular Basement Structure 67 7.4 Lateral Deflection of Rectangular Basement Structure Fringe Plot 68 7.5 Lateral Deflection of Curvilinear Basement Structure 68 7.6 Lateral Deflection of Curvilinear Basement Structure Fringe Plot 69 7.7 Z Stress (Vertical) Plot of Window Area of Rectangular Foundation 7 0 7.8 Z Stress (Vertical) of Window Area of Rectangular Foundation Fringe Plot 7 1 7.9 Z Stress (Vertical) Plot of Window Area of Curv ilinear Foundation 7 1 7.10 Z Stress (Vertical) of Window Area of Curvilinear Foundation Fringe Plot 7 2 7.11 Shear Stress Plot along Window Area of Rectangular Foundation 7 3 7.12 Shear Stress along Window Area of Rectangular Foundation Fringe Plot 7 4 7.13 Shear Stress Plot along Window Area of Curvilinear Foundation 7 5

PAGE 12

xii 7.14 Shear Stress along Window Area of Curvilinear Foundation Fringe Plot 75 8.1 Lateral Deflection of Rectangular Basement Structure Fringe Plot 78 8.2 Lateral Deflection of Curvilinear Basement Structure Fringe Plot 78 8.3 Z Stress (Vertical) of Window Area of Rectangular Foundation Fringe Plot 8 0 8.4 Z Stress (Vertical) of Window Area of Curvilinear Foundation Fringe Plot 8 1 8.5 Vertical Stress around Window Rectan gular Structure 8 2 8.6 Vertical Stress around Window Curvilinear Structure 8 3 8.7 Shear Stress along Window Area of Rectangular Foundation Fringe Plot 86 8.8 Shear Stress along Window Area of Curvilinear Foundation Fringe Plot 87 8.9 Shear Stress a round Window Rectangular Structure 88 8.10 Shear Stress around Window Curvilinear Structure 89

PAGE 13

1 1. Introduction Expansive (swelling) soils are extremely common in the Front Range area of Colorado and can be found on almost every continent across the globe. The destructive effects caused by expansive soils have been reported in numerous countries such as the United States, Canada, Australia, China, Israel, South Africa and India (Nelson and Miller, 1992; Steinberg, 1998). It has been widely reported that losses due to expansive soils have been measured in several billions of dollars yearly ( Nelson and Miller, 1992 ) The cost of repairing damage caused by swelling soils amounts to more than the cost for all other natural hazards combined. This is especially true of light structures, pavements and service piping. Expansive soils are capable of mobilizing huge vertical and lateral pressures which, in turn, become a hazard primarily to structures and pavements built on top o f the expansive soil or within the volume of expansive soils that are subject to moisture changes. The damage may not manifest itself immediately, depending on the soil composition, moisture history, future moisture/desiccation cycling and type of foundat ion construction. Up until now, mitigation in the industry has followed two primary paths: mitigation of existing structures by adjusting drainage and/or underpinning; and mitigation of the design by using pier and beam foundations with drainage impleme nted at the foundation to prevent soil expansion. To a lesser extent, changing the soil properties by chemical mixing, removal of the offending soils or mixture of the soils with more suitable soil has been implemented in the industry.

PAGE 14

2 Although the curren t industry standards for mitigation of light structures constructed in expansive soils, as described above, offer piece of mind against catastrophic damage, long term stability is not guaranteed. Over time, severe drought and flooding cycles can directly influence the foundation performance and exceed the design mitigations implemented. In addition, changes in soil chemistry can occur during periods of high moisture exposure negating the benefits of soil treatment methodologies. Over time, changes in soi l drying can occur due to rises in the ambient temperatures and/or the growth of vegetation within the soil mass. With unprecedented weather events taking place all over the globe, the design of structures must rely less on mitigation of moisture intrusio n and more on foundation designs that take advantage of the potential forces mobilized by the soil. 1.1 Purpose of the Study The purpose of this study is to introduce a foundation design that accounts for and reacts to the pressures generated by expansive soils. It allows for changes in the moisture content of the expansive soil and is designed to accommodate the resulting forces. This approach is consistent with the intent of the International Building Code and the American Concrete 05: B uilding Code Requirements for Structural Concrete. 1.2 Scope of Study The primary scope of this study is to examine the current design and construction practices for light foundation designs relative to expansive clay soil pressures with changing soil prop erties. Included in the study is a curvilinear

PAGE 15

3 foundation design approach/solution that works with the in situ soil conditions to resist the changing lateral pressures and soil heave. The study includes a Finite Element Model and analysis of a new founda tion design compared to a traditional rectangular foundation design used in industry today. 1.3 Organizational Outline A brief description of each chapter in this study is presented below. Chapter 1 introduces the issue of expansive soils as they relate to foundation design and damage. Chapter 2 presents a review of clay soil properties and experimental techniques used in determining swell potential and lateral swell pressure. The risk of foundation movement relates to the amount of vertical and horizontal heave/swell that will occur. Heave depends on more than just the percent swell of the soil. Calculations of predicted heave must also take into account the stress or surcharge applied to the soil when the soil is inundated with water. Various methods are widely used in the industry to classify swell potential and determine soil properties related to unsaturated clays. One method commonly used to determine the expansion potential of a soil is based on the index properties (Holtz and Gibbs 1956; Holtz and Kovacs 1981). This requires knowledge of the clay content and the plasticity index. These properties can be determined by performing a gradation test including the Atterburg limits. Another method widely used for estimating the expansion potential of a clay soil uses soil classific ation

PAGE 16

4 information. Seed et al. (1962) developed a classification chart method (activity) based on the amount and type of clay particles in the soil. In addition, experimental methods are also used to determine swell induced strains, swell potential and s well pressure. These are typically accomplished by means of a consolidation swell type test such as ASTM D4546, One Dimensional Swell or Collapse of Cohesive Soils. Chapter 3 addresses the application of lateral pressures resulting from the soil mass, sur charge and the water table. It also describes the methods that are used to define the lateral pressure profile on a structure. Chapter 4 reviews the current design practices that are used for foundation wall design as described in the International Buildi ng Code (International Code Council, 2005), the International Residential Code for One and Two Family Dwellings (International Code Council, 2005), and the American Concrete Institute, 318 05: Building Code Requirements for Structural Concrete. Chapter 5 presents a new, curvilinear foundation design approach to effectively use the properties of an expansive soil to achieve long term survivability and serviceability of the structure. This includes, as an assumption, changing soil conditions that prove to b e problematic to traditionally designed foundations including moisture and climatic changes, soil chemistry changes and changes in drainage.

PAGE 17

5 Chapter 6 presents the Finite Element Analysis approach for the new foundation design using LS DYNA software (Liver more Software Technology Corporation). The analysis parameters for structural properties of the foundation and loading due to the soil are presented. Also discussed is a traditional rectangular foundation design for comparison. Chapter 7 presents the fin dings of the analysis. It compares the new curvilinear foundation design to the traditional rectangular design, evaluating wall stress versus applied loading. Chapter 8 presents the discussion of the results and a comparison of performance between the tra ditional rectangular foundation design and the curvilinear design. Chapter 9 presents the summary and conclusions of the research/analysis and recommendations for future, related research.

PAGE 18

6 2. Properties of Cohesive Soils 2. 1 Introduction Expansive Soils Clay soils are often described as cohesive, fine grained soils having plasticity and containing clay minerals such as kaolinite, halloysite, mon t morillonite, illite, chlorite and vermiculite ( Holtz & Kovacs, 1981 ) However, not all fine grained soils are cohesive and/or clay. Silts, for example, are classified as fine grained and granular but are not cohesive and are not plastic. For clay soils, grain size distribution has little influence on the properties of the clay whe reas for granular soils the grain size distribution and the grain shape can have marked effects on the properties of the soil. Additionally, water content is relatively unimportant (with a few exceptions) for granular soils but has a definite influence on clay soils. Silts are fine grained and granular but are not plastic and are non cohesive. Their strengths, like sands, are essentially independent of water content. Clay minerals owe their unique properties and behavior to some very distinct characteristics. Clay minerals are extremely small particles (< 1 m diameter) that are electrochemically active. They are affected by the quantity and type of clay minerals pr esent, the moisture content, the type and chemistry of the soil water surrounding the clay particles, the arrangement, soil density and specific surface area of the clay particles. In a mixed clay and soil mass, as the clay content increases, the behavior of the soil mass is increasingly governed by the clay fraction properties. As the clay content approaches and exceeds approximately 50%, the sand and silt grains in the mixed clay/soil mass are

PAGE 19

7 ass behavior ( Holtz & Kovacs, 1981 ). 2.2 Soil Classification The purpose of soil classification is to provide for a common means of determining or predicting the behavior of soils and/or evaluating soils for engineering purposes. There are numerous soil cl assification systems in use. In the United States, the Unified Soil Classification System (USCS) is the most widely used soil classification system for structural considerations (Howard 1977) while the American Association of State Highway and Transporta tion Of ficials ( AASHTO ) classification system is typically used for pavement design. The Unified Soil Classification System (USCS) was initially developed by Casagrande in1948 and later modified by Casagrande in 1952. Within the USCS (Figure 2.1) soil m aterials are classified into three main groups: C oarse g rained, F ine grained and P eat (highly organic soils) depending on the predominant particle sizes and make up within the soil matrix. Soils are identified within the three major groups primarily on th e basis of particle sizes and changes to the soil properties and volume when interacting with water. Coarse grained soils ( sands and gravels) contain particles that are visible to the naked eye (larger than about 0.003 in. [0.075 mm]) and are generally described as cohesionless with engineering behavior primarily influenced by the composition of particle sizes, particle shape and relative density. Coarse grained soils are further defined within the USCS as greater than 50% (by dry mass) retained on the number 200 Standard Sieve with a mesh opening of 0.075mm.

PAGE 20

8 Subdivisions within this classification system are largely based o n particle size: gravels (75mm to 4.75mm) and sands (4.75mm to 0.075mm). Both sands and gravels are further subdivided into four secondary groups (GW, GP, GM, GC; SW, SP, SM, SC). The four secondary classifications are based on whether the soils are well graded, poorly graded, contain silt size particles or contain clay size particles. Fine grained soils include silts and clays containing particles that are not visible to the naked eye. Fine grained soils are those composed primarily of silt and clay sized particles smaller than 0.075 mm. Fine grained soils are defined as having 50 percent or more (by dry mass) of soil particles passing through the number 200 Standard Sieve. Silts and clays are largely distinguished based on the plasticity properties of t he soil, as measured by the soils' Atterberg Limits Both silts and clays are further subdivided into three secondary groups (ML, CL, OL; MH, CH, OH). The three secondary classifications are based on the inorganic and organic nature of the soil and on its plasticity. Silts may be either cohesive or cohesionless and are granular material s with sizes falling between sand s and clay s. Silts may occur as a soil or as suspended sediment Clays, on the other hand, are cohesive soils, with engineering behavior primarily influenced by plasticity and cohesion. Soils containing high natural organic content comprise the third major group. Peats (o rganic soils) can be of extremely low strength and high compressibility, depending on organic content and composition, and geologic history.

PAGE 21

9 Fig ure 2.1 : Grain size Classification System (a fter U.S. Army Waterways Experiment Station (1960) and Ho ward (1977) ) 2.2.1 Gradation of Soils Gradation tests are performed on a soil to determine the particle size distribution which is used in the classification of a soil. The gradation of a soil has a major effect on its mechanical and hydraulic properties and enables an evaluation of engineering characteristics such as permeability, strength, swelling potential, and susceptibility to frost action. The tests consist of two types: sieve analysis for coarse grained soils (sands, gravels) and hydrometer analy sis for fine grained soils (silts, clays). Materials containing both types of soils

PAGE 22

10 (sands/gravels and silts/clays) are tested by both methods and the results are merged to create one particle size distribution result. Gradation of coarse grained soils co nsists of a mechanical grain size analysis. The analysis consists of taking an oven dried soil sample and subjecting it to a series of standard sieves with progressively smaller openings while mechanically shaking the sieves. Once complete, the amount of material retained on each of the sieves is weighed. The total percentage passing each sieve is determined and the data plotted on a semilogarithmic graph of grain size versus percent finer by weight (Fig ure 2.2). Based on the results of the particle size distribution testing, soils can be classified as poorly graded (uniform), when it contains a narrow distribution of particle sizes or well graded, when the soil has a wide range of particle sizes. The flatter the grain size curve the larger the range of particle sizes found in the soil and the steeper the curve the fewer the particle sizes. Generally speaking, a well graded soil has a curve that is smooth and contains particles over a relatively large range of sizes while a poorly graded soil has a curve where a high portion of the soil particles contain sizes within a narrow band. If particles of large and small sizes are present with a low proportion of particles in the intermediate sizes the soil is categorized as a gap graded soil (McCarthy, 1998).

PAGE 23

11 Figure 2.2 : Grain Size Distribution (after ASTM International, D2487 11) A hydrometer analysis is performed on soils finer than the No. 200 sieve (0.07 5 mm) since a sieve analysis is impractical for small diameter particles (grains). The hydrometer analysis is a sedimentation process where the rate of settlement of a soil in water is measured as an indication of particle size. The test is based on Stoke terminal velocity of fall depends on the grain diameter and the densities of the grains in suspension and of the fluid. The particle diameter can be determined from knowledge of the distance of fall and the time. Stokes law does not apply to particle sizes below 0.0002 mm as these particle sizes are influenced by Brownian movement ( U.S. Army Corps of Egineers 1998). Interpretation of the gradation analysis focuses on the range of particle diameters found in the sample. This information can be readily determined from

PAGE 24

12 the semi logarithmic grain size distribution curve (Figure 2.2). The particle size reading the particle size fr commonly used in calculating uniformity coefficients are the percentage smaller than 10%, 30% and 60% and are denoted D 10 D 30 and D 60 respectively. A s a measure of the gradation of a soil, the coefficient of uniformity ( C u ) is used to D 60 size of the soil (the particle size in mm where 60% of the soil particles are finer than) to the D 10 size (the particle size in mm w here 10% of the soil particles are finer than). The uniformity coefficient ( C u ) is calc ulated as the following ratio: 2.1 Where: D60 = soil particle diameter at which 60% of the mass of a soil sample is finer and D10 = the diameter at which 10% of the mass of a soil sample is finer The D 10 is often referred to as the effective particle size and is utilized in many empirical methods to characterize the soil as a whole, particularly with regard to hydraulic conductivity. Generally, the higher the value of the coefficient of uniformity (C u ) the greater the range of particle sizes in the soil sample. Another quantity that may be used to judge the gradation of a soil is the coefficient of curvature, designated by the symbol C c The coeffi cient of curvature is defined as the following:

PAGE 25

13 2.2 Where: D60 = soil particle diameter at which 60% of the mass of a soil sample is finer, D10 = soil particle diameter at which 10% is finer and D30 = soil particle diameter at which 30% of the mass of the soil is finer. A well graded soil is defined as having a good representation of all particle sizes from the largest to the smallest and the shape of the grain size distribution curve is considered "smooth." In the USCS, wel l graded gravels must have a C u value > 4, and well graded sands must have a C u value > 6. For well graded sands and gravels, a C c value from 1 to 3 is required. Sands and gravels not meeting these conditions are considered poorly graded. 2.2.2 Atterberg Limits Atterberg limits are limits of moisture content (mass of water in the soil to the mass of the solid particles) used to define fine grained soil behavior. In engineering practice, three of the limits (the liquid, plastic and shrinkage limits) are co mmonly used. The Liquid Limit (LL) is the water content, in percent, that defines where the soil changes from a viscous, fluid state to a plastic state Above this point the soil behaves as a liquid, while below this point the soil behaves as a plastic ma terial. The Liquid Limit can be measured using the (Casagrande) liquid limit device.

PAGE 26

14 The Plastic Limit (PL) is defined as the water content, in percent, where the soil changes from a plastic state to a semi solid state Above this point the soil behaves as a plastic material, while below this point the soil behaves as a semi solid The Plastic Limit is also the moisture content at which a soil crumbles when rolled into a thread of 1/8 inch in diameter (Das, 2002). The Shrinkage Limit (SL) is defined as t he moisture content where the soil volume will not reduce further if the moisture content is reduced. Above this point the soil behaves as a semi solid, while below this point the soil behaves as a solid. Plasticity Index (PI) is defined as the difference between the moisture content at the Liquid and Plastic Limits. This represents the range of water content where a material behaves plastically ( Das, 2002 ). 2.3 Since the PI is determined from Atterberg L imits test ing on the fraction of soil that passes the no. 40 sieve (0.425 mm), a correction factor is applied for soils that contain a large fraction of particles coarser than the no. 40 sieve. Fine grained (cohesive) soils can be classified either as low or high compressibility materials based on the results of the Atterberg Limits tests. By plotting the Plasticity Index versus the Liquid Limit the classification can be determined graphically (Figure 2.3 ).

PAGE 27

15 Figure 2.3 : Plasticity Index versus Liquid Limit (ASTM D2487 11 Standard Prac tice for Classification of Soils for Engineering Purposes ) The A Line separates clay classifications and silt classifications, while the U Line represents an approximate upper limit of LL and PI combinations for natural soils. 2.2.3 Activity A variety of soil engineering properties have been correlated to the liquid and plastic limits as well as being used to classify fine grained soils according to the Unified Soil Classification System. Knowledge of the Atterberg limits for a cohesive soil and the natu ral moisture content can tell a good deal about its geologic history and engineering performance (Figure 2.4). 0 10 20 30 40 50 60 -10 0 10 20 30 40 50 60 70 80 90 100 110 Plasticity Index (PI) Liquid Limit (LL) Plasticity Chart for Classification of Fine Grained Soils ML or OL MH or OH CL or OL CL or ML

PAGE 28

16 Figure 2.4 : Probable Clay Expansion as Estimated from Classification Test Data (after Holtz, 1959) The presence of small amounts of certain types of clay minerals can have significant impacts o n the soil properties. The identification of the type and amount of the clay minerals present can help in determining or predicting the he clay minerals present (McCarthy 1998). Indirect methods are available to determine information about the type and effects of clay minerals in a soil that are relatively easy to perform and give qualitative, if not quantitative results (Figure 2.5) Figure 2.5 : Characteristics of Common Clay Minerals (after Mitchell, 1976 )

PAGE 29

17 One such method is the clay Activity and is determined from a relationship of the plasticity of the clay and the quantity of clay sized particles. For a given amount of clay min eral the plasticity resulting in a soil varies for different clay minerals. The Activity, defined as 2.4 Where the percent by weight finer than 2 m is equal to that portion of the soil (by weight) consisting of particles <0.002 mm. This information can be determined by the hydrometer analysis (ASTM International D422 63 ). Three classes of clays according to activity suggested by Skempton (1953) are inactive (A The clays with the highest activity have the most potential for expansion. Table 2.1 : Clay Activity (after McCarthy, 1998) Activity Classification <0.75 Inactive Clay 0.75 1.25 Normal Clay > 1.25 Active Clay Typical values of activities for various clay minerals range from q uartz (activity = 0) to sodium montmorillonite (activity = 4 to 7) at the extremes while

PAGE 30

18 Illite (activity = 0.9) and calcium montmorillonite (activity = 1.5) fall somewhere in between 2.3 Clay Minerals The cohesive properties of natural soils are related to the presence of clay minerals (e.g., kaolinite, halloysite, montmorillonite, illite, chlorite and vermiculite). The most important of these clay minerals assoc iated with expansion are kaolinite, montmorillonite and illite. All are very small crystalline substances of hydrous alum i nosilicates (phyllosilicates or layered silicates ). The crystalline substances are a result of weathering (physical or chemical) of parent rock materials, primarily igneous and sedimentary rocks (G.W. Donaldson, 1969). Three of the most important clay mineral groups are the kaolinites (generally non expansive), the mica mineral group which includes the illites and vermiculites (can be expansive), and the smectites which includes montmorillonite (can be highly expansive). The swelling potential varies widely with the type of clay mineral and is generally ranked in the following order from most to least expansive: montmorillonite, illite kaolinite (Lambe and Whitman, 1969). Clay minerals are generally constructed of stacks of two types of sheets: silica tetrahedral sheets and alumina octahedral sheets. Each sheet is only angstroms thick (1.0000e 10 m ) and can be thousands of angstroms wi de in each of their lateral dimensions. The individual crystals can only be observed with an electron microscope and their structure has been observed using x ray

PAGE 31

19 diffraction. Each family of clay mineral type consists of vertical stacks of these elementar y plates or sheets in differing arrangements. Each plate or sheet has a repeating atomic structure consisting of the two primary crystal sheets. The silica tetrahedral sheets are constructed of individual tetrahedron molecules with the general composition of SiO 4 (in some instances contain substitutions of aluminum ions for silica ions) The alumina octahedr al sheets are constructed of individual octahedron molecules with the general composition of Al(OH) 6 (Gibbsite) with substitutions of either iron or ma gnesium (Brucite) for aluminum ions. If all of the anions are hydroxyls and at least 2/3 of the cation positions are filled with aluminum the mineral is labeled gibbsite. If the cation positions are filled with magnesium instead of the aluminum then the mineral is labeled brucite. The arrangement of the stacking of these sheets, with the various metal ion substitutions throughout the crystal lattice, makes up the different clay mineral types ( Holtz and Kovacs, 1981 ). The basic tetrahedral sheet is a comb ination of silica tetrahedral molecules. The tetrahedral molecule consist s of four oxygen atoms at each of the corners of the tetrahedron with a silicon atom at its center (tetrahedr al s heet F ig ure 2. 6 ). The oxygen atoms at the base of a single tetrahedr on molecule are combined to form continuous sheets with the base oxygen atoms arranged in a single plane.

PAGE 32

20 Figure 2. 6 : Silica Tetrahedron and Silica Tetrahedral Molecules ( after Grim, 1968 ) The basic octahedral sheet is a combination of octahedral molecules. The octahedral molecule consists of six oxygen atoms or hydroxyl molecules (Oxygen Hydrogen molecule) with aluminum, magnesium, iron, or other atom at its center (single octahedral molec ule F ig ure 2. 7 ). The oxygen atoms or the hydroxyl molecules are positioned such that two planes are formed within the sheet. Substitutions of the cations within the octahedral sheet are common and produce different clay minerals. Oxygen Silicon

PAGE 33

21 Fig ure 2. 7 : Alumina Octahedron and Alumina Octahedral Molecules (after Grim, 1968 ) A ll clay minerals contain the two basic tetrahedral and octahedral sheets and are stacked together in various combinations with differing substitutions of cations within the molecules to form the minerals. The various types of clay minerals result from the stacking of the specific te trahedral and octahedral sheets and the type of chemical bonding between each sheet. The variety of mechanical and physical properties of different clays is a result of the u nbalanced electrical charges that are a result of certain cationic substitutions and the overall net electronegative polarity resulting from the sheet stacking arrangements This includes differences in cohesion water absorption and expansion found in the different clay minerals The more common clay minerals found in clay soils wit hin the Unites States are described in the following sections. Hydroxyls or Oxygen Aluminum, Magnesium, etc.

PAGE 34

22 2.3.1 Kaolinite Kaolinite minerals are formed of repeating layers of one tetrahedral (silica) sheet and one octahedral (alumina or gibbsite) sheet and are classified as 1:1 minerals. The two layers are held together by hydrogen bonding (Hydroxyl ions on the octahedral sheet to oxygen atoms on the tetrahedral sheet) forming a single layer ( F ig ure 2. 8 ). The strong hydrogen bonding resists swelling stresses when water is present and also allows the construction of large crystal structures. These mineral layers are approximately 0.72 nm (7.2 A) in thickness and extend laterally in both directions indefinitely. Kaolinite crystals are constructed of repeating layers of the 0.72 nm mineral layer. It is not uncommon to have kaolinite crystals 70 to 100 layers thick ( Holtz and Kovacs, 1981 ). The cation exchange capacity of Kaolinite is very low (3 to 15 meq/100 gm) owing to little substitution within the mineral sheets. Figure 2. 8 : Schematic Diagram of Kaolinite (after Lambe, 1953) 7.2 A Repeating struc ture Repeating structure Al Si Tetrahedral sheet typ. Octahedral sheet typ. Gibbsite typ. Silica sheet typ.

PAGE 35

23 2.3.2 Halloysite Halloysite is another 1:1 clay mineral but differs from Kaolinite due to water entrained between the mineral layers. This hydration results in deformations of the crystal structure and random stacking of the mineral sheets giving it a tubular structure. The water can be removed from t he mineral layers by heating or air drying but is irreversible. The Halloysite mineral sheet will not rehydrate upon addition of water to the clay mineral ( Holtz and Kovacs, 1981 ). When hydrated, the clay exhibits a 1 nm spacing of the layers and when de hydrated the mineral layers are 0.7 nm in spacing. Halloysite naturally occurs as small cylinders which average 30 nm in diameter with lengths between 0.5 and 10 micrometers. The cation exchange capacity depends on the amount of hydration, as 2H 2 O has 5 1 0me q /100g, while 4H 2 O has 40 50me q /100g. 2.3.3 Montmorillonite/Smectite Montmorillonite, or Smectite, is a 2:1 mineral containing two silica tetrahedron sheets and one alumina (gibbsite) octahedron sheet. The octahedron sheet is located between the two silica tetrahedron sheets forming a single layer ( F ig ure 2. 9 ). The tips of the silica tetrahedrons form a bond with the hydroxyls of the alumin forces. The bonding forces are weak and allow water and exchangeable ions can e nter the layers. Typical thickness of a 2:1 mineral layer is approximately 0.96 nm (9.6 A) and extends laterally in both directions indefinitely ( Holtz and Kovacs, 1981 ). Due to the very small size and the affinity for water, Montmorillite is highly expa nsive

PAGE 36

24 depending on the initial and final moisture contents. The cation exchange capacity of smectite is in the range of 80 to 150 meq/100 gm. Figure 2. 9 : Schematic Diagram of Montmorillonite (after Lambe, 1953) 2.3.4 Illite Illite is another 2:1 mineral similar to Montmorillite with the interlayers bonded with a potassium atom. Illites consist of one octahedral sheet sandwiched between two silica tetrahedral sheets. The potassium atom fits into the hexagonal space created i n the silica tetrahedron sheet and bonds the mineral layers together ( F ig ure 2. 10 ). The resulting charge is somewhat balanced by the potassium atoms in the hexagonal space between the layers. This makes the potassium essentially non exchangeable since bo nding is very strong. Ionic substitutions do occur in Illite, usually in the silica tetrahedral layers. 9.6 A Gibbsit H 2 O and cations in interlayer region Repeating structure Repeating structure Si Al Si n H 2 0 layers and exchangeable cations Tetrahedral sheet typ. Octahedral sheet typ. Gibbsite typ. Silica sheet typ.

PAGE 37

25 This strong bonding reduces the potential for expansion by preventing water intrusion between the layers allowing the layers to remain essentially con stant (Mitchell, 1993). The cation exchange capacity of Illite is in the range of 10 to 40 meq/100 gm. Figure 2. 10 : Schematic Diagram of Illite (after Lambe, 1953) 2.3.5 Chlorite Chlorite is a 2:1:1 mineral consisting of a silica tetrahedron sheet, an alumina octahedral sheet, a silica tetrahedron sheet followed by either a gibbsite or brucite sheet ( F ig ure 2. 11 ). Unlike other 2:1 clay minerals, chlorite's interlayer space i s made up of Mg2+ or Fe3+ more commonly referred to as the brucite like layer. Chlorite can be missing an occasional brucite or gibbsite layer and also have considerable isomorphous substitution This may lead to a higher 9.6 A Gibbsit Repeating structure Repeating structure Si Al Si Potassium ion Tetrahedral sheet typ. Octahedral sheet typ. Gibbsite typ. Silica Sheet typ. K K

PAGE 38

26 s usceptib ility to swelling due to water enter ing between the sheets. Generally, it is signif icantly less active than montmorillonite ( Holtz and Kovacs, 1981 ). Cation exchange capacity for silt size chlorites varies from 4 to 32 me q/100 gm and for q/100 gm. Cation exchange capacities for ites are essentially the same. Figure 2. 11 : Schematic Diagram of Chlorite (after Mitchell, 1976) 14.0 A Repeating structure Repeating structure Si Si Tetrahedral sheet typ. Octahedral sheet typ. Gibbsite or Brucite Silica Sheet typ. Brucite Brucite Al

PAGE 39

27 2. 4 Swelling Potential of Clay Soils One of the most important considerations in determining risks to structures due to expansive soils is identifying the vertical and lateral swell potential of the soil and the resulting swell pressures. One dimensional tests are by far the most widely used method to estimate expansive soil swelling potential and pressures. This is due largely to the simplicity of the procedures and the availability of the testing equipment. Standard test methods for evaluating the potential for one dimensional heave / set tlement and swell pressure of cohesive soils is described in ASTM D 4546 (One Dimensional Swell or Collapse of Cohesive Soils). Three oedometer / consolidometer tests useful for measuring potential swell / settlement and swelling pressure can be described as follows: 1) Free swell test A seating pressure (e.g., 0.01 tsf) is applied to the test sample in a n oedometer / consolidometer the sample is inundated with water and allowed to swell vertically until primary swell is complete. The sample is loaded following primary swell until its initial void ratio / height is obtained. The total pressure required to reduce the test sample hei ght to the original void ratio / height prior to inundation is defined as the swell pressure. 2) Overburden swell test A vertical pressure exceeding the seating pressure is applied to the test sample in a n oedometer / consolidometer and the sample is inundate d with water. The test sample may swell, swell then contract, contract, or contract then swell. The vertical pressure is

PAGE 40

28 typically chosen to replicate the in situ overburden pressure and may include structural loads depending on the purpose of the test. 3) Constant volume swell test A seating pressure and additional vertical pressure (typically equivalent to the in situ overburden pressure) is applied to the test sample in a n oedometer / consolidometer and the sample is inundated with water. Additional ve rtical pressure is applied as needed or removed to maintain a constant void ratio / height of the test sample. A consolidation test is subsequently performed and the total pressure required to maintain a constant void ratio / height of the test sample is the measured swell pressure. This measured swell pressure is corrected to compensate for sample disturbance by using the results of the subsequent consolidation test. A suitable correction procedure is similar to that for estimating the maximum past pressu re. The procedures outlined typically use a rigid soil confining ring in a n oedometer / consolidometer apparatus to measure the vertical stress and strain components of swell. 2.4.1 Consolidometer Swell Test e Dimensional measuring free swell, swell pressure and the magnitude of one dimensional swell or collapse of compacted or intact cohesive soils. The test methods can be used to m easure the magnitude of one dimensional wetting induced swell or collapse (hydrocompression) under different vertical loading, as well as the magnitude of

PAGE 41

29 vertical swell pressure and the magnitude of vertical free swell. It can also be used to obtain data for stress induced compression following wetting induced swell or collapse. All of the methods involve the use of a one dimensional oedometer / consolidometer apparatus to laterally restrain the soil s ample and allow for access to free water. Three alter native methods are described to determine the swell behavior and measure the swell parameters of the soil. The three testing procedures for determining the swelling pressure of a soil can be described as: Method A wetting after loading testing of multip le samples differing surcharge loading performed on compacted or natural soil samples followed by inundation with free water. Method B single point wetting after loading testing of a single sample of natural soil, or a embankment followed by inundation with free water. Method C loading after wetting test after completion of the Method A or B testing increments of addit ional vertical loads are applied to the sample and the load induced deformations are determined. The results would apply to situations where new fill and/or additional structural loads are applied to the ground that has previously gone through wetting ind uced heave or settlement. Typically, two classes of testing are performed, free swell and either a free swell test followed by consolidation or a continuous consolidation upon

PAGE 42

30 inundation of water to keep a constant volume. In the free swell test, a sample is subjected to an applied load and allowed to swell freely. The resulting final volume changes are plotted against the corresponding applied loads or stresses. The stress corresponding to zero volume change (reloading the sample to the initial void rati o / height) is termed the swelling pressure ( Hardy et al, 1962 ). The swelling pressure may be further defined as the pressure that prevents either a positive or negative volume change. In the free swell followed by consolidation or continuous consolidati on upon inundation with water, the soil sample is inundated with water and allowed to swell freely with a given load applied. The soil is gradually consolidated back to its original void ratio / height in the conventional manner of a consolidation test pr ocedure. The swelling pressure is defined as the stress necessary to consolidate the specimen back to its original volume or the minimum stress required to prevent swelling (ASTM D4546 08) The swell pressure ( P s) is the applied load required to prevent s well strain ( L ) divided by the cross sectional area of the specimen (A): 2.5 Stability is assumed to occur when no further applied load is required to restrict vertical strain. Swell strain is determined by measuring the resultant heave after reaching stability (no further increase in sample height with time) upon inundation of water

PAGE 43

31 to the sample. The heave is defined as the change in height of the sample and the swell strain in the change in height divided by the initial height of the sample: 2.6 The test results are commonly plotted as shown in Fig ure 2.1 2 vertical height or void ratio versus the log of the vertical load (or stress) applied. Figure 2.1 2 : Free Swell Oedometer Test Results 0.7 0.75 0.8 0.85 0.9 0.95 1 1.05 1.1 1,000 10,000 100,000 1,000,000 Void Ratio (e) Pressure (psf) Free Swell Oedometer Test Initial Swell Consolidation Curve Loading Rebound Curve Unloading Swelling Pressure

PAGE 44

32 The actual stress paths followed during the test can be traced beginning with water inundation initiating the swell of the soil followed by loading of the sample to reach the original height/void ratio. At this point the pressure required (Ps) to resist the swelling can be determined. To empirically account for sampling disturbance, Fredlund et al. (1980) defined a correction procedure which could be applied to the data to give a corrected swellin g pressure. The correction procedure is a modification of the Casagrande type of geometrical construction as shown in Figure 2.1 3 Using the constant volume oedometer test plot (void ratio versus logarithm of total pressure) a correction can be establishe d by identifying the point of maximum curvature of the curve. The point is found immediately past the uncorrected swelling pressure. From that point, horizontal and tangential lines can be drawn and the resulting angle bisected. The intersection of a li ne parallel to the rebound portion of the curve and the bisector indicates the corrected swelling pressure.

PAGE 45

33 Figure 2.1 3 : Correction for Sample Disturbance (Fredlund, et.al., 1980) T he potential swelling pressure and volume change of a soil sample can be determined from laboratory tests. The results can be applied to actual foundations with reasonable accuracy when the stress distribution of the fo undation soil is understood and the effects of seasonal variations and the movement of moisture beneath the foundation is known. The oedometer tests in which a soil sample is subjected to estimated surcharge loads will help understand the in situ potent ial expansion and pressure.

PAGE 46

34 3. Lateral Forces on a Foundation Lateral forces applied to a structural foundation from the surrounding soil are directed so as to apply a load perpendicular to the plane of the wall. This loads the wall in flexure (bending) and it must act as a one way slab, beam, or two way slab, depending on the design. Determination of the magnitude and orientation versus the depth of these forces are crucial to the development of a safe and economic design. For a proper struct ural design, these lateral forces are generated from three sources: the soil used for backfill, the water table depth in the backfill and the surcharge loads (if any) at the top of the backfill in proximity to the wall. 3.1 Lateral Earth Pressure The magni tude of the lateral soil pressure that can develop in a soil mass is a function of several factors: the type of soil, the strength of the soil, the stress strain properties of the soil, the unit weight of the soil, the drainage conditions of the soil, the water table depth, and the amount and direction of wall movement when subjected to the lateral soil pressure. In defining the stress state of the wall three conditions may exist: (1) The wall is restrained from moving either toward the soil mass or away from the soil mass and the lateral earth pressure on the wall is defined as the at rest earth pressure. No deformations or displacements are occurring in this stress state. This is described as an at rest condition.

PAGE 47

35 (2) The wall may tilt or translate away from th e soil mass where a triangular soil wedge behind and adjacent to the wall may fail. This lateral earth pressure in this condition is defined as an active earth pressure. (3) The wall may tilt or translate into the soil mass where, with sufficient movement, a soil wedge may fail. This lateral earth pressure in this condition is defined as a passive earth pressure. The at rest earth pressure condition can be explained by means of F ig ure 3.1, where length A B is a structural retaining wall that supports a reta ined soil mass. The backfill is horizontal, typical of a structural basement wall. If the structural wall does not move (rotate) or yield either toward or away from the retained soil, the horizontal lateral earth pressure at any depth to which the wall w ill be subjected is called the lateral earth pressure at rest. The total force per unit length of the wall is equal to 0 Fig ure 3 1 : At rest Earth Pressure H = height Structural Basement Wall A B

PAGE 48

36 If the structural wall tends to move (rotate) or yield away from the retained soil mass, creating a plastic deformation in the soil mass, the lateral earth pressure at any depth to which it will be subjected is called the active earth pressure. This condition is described in F ig ure 3.2. This active condition can be a result of rotation of the wall about its bottom or top, or by translation of the wall away from the retained soil mass. a ) magnitude, direction, and location per unit length of the wall depends on several factors including soil type, shear strength, backfill incline and stiffness of the structural wall and foundation. Fig ure 3.2 : Wall Movement for Active Earth Pressure If the structural wall tends to move (rotate) or yield into the retained soil mass, creating a plastic deformation in the soil mass, the lateral earth pressure at any depth to which it will be subjected is called the passive earth pressure. This condition is described in F ig ure 3.3. This passive condition can be a result of

PAGE 49

37 rotat ion of the wall about its bottom or top, or by translation of the wall into the p ) magnitude, direction, and location per unit length of the wall also depend on several factors including soil type, shear strengt h, backfill incline and stiffness of the structural wall and foundation. Fig ure 3.3 : Wall Movement for Passive Earth Pressure In this research the foundation wall is modeled as a two way slab with sufficient support at the base and top of the foundatio n wall to prevent either active or passive earth pressures and is therefore modeled using an at rest earth pressure. For an at rest condition, vertical stresses existing in a soil mass at a given depth Z below the ground surface where the water table exist s below the depth Z of the soil element is the weight of the overburden and can be written a s 3.1

PAGE 50

38 w soil is the unit weight of the soil mass for a homogenous, isotropic mass of infinite extent (McCarthy, 1980). Fig ure 3.4 : Relationship between Vertical and Horizontal Soil Stress The relationship between the vertical stress and the horizontal stress (Figure 3.4) is determined by calculation of the coefficient of lateral earth pressure at rest K 0 h and the vertical stress v 3.2 Where v above equation (Jky, 1944) is an empirical approximation and is defined as the H = height Z v = soil z h = K 0 v

PAGE 51

39 at rest condition for normally consolidated sands designated K 0 which can be written: 3.3 For normally consolidated clays, the coefficient of lateral earth pressure K 0 in the at rest condition is approximated by Brooker and Ireland (1965) as 3.4 where is the drained peak friction angle of the soil. In addit ion, Brooker and Ireland experimented with normally consolidated clay soils and determined the value for K 0 may be approximated with the Plasticity Index (PI) as 3.5 This for a Plasticity Index between 0 and 40 and 3.6 for a Plasticity Index between 40 and 80.

PAGE 52

40 Where overconsolidated clays are concerned K 0 can be defined as follows: 3.7 Where OCR is defined as the overconsolidation ratio. A soil is described as normally consolidated when the preconsolidation pressure equals the existing vertical overburden pressure (the soil has not experienced anything greater than the current overburde n stress). If the preconsolidation pressure is greater than the existing vertical overburden pressure the soil is described as being overconsolidated (the soil had experienced a pressure that is larger than the current overburden condition). The overcons olidation ratio can be determined by the ratio of the preconsolidation stress divided by the existing vertical effective overburden stress 3.8 p vo is the effective vertical overburden pressure (Holtz & Kovacs, 1981). As can be seen from the OCR p = vo ), soils p > vo ) and finally a soil that has p < vo ). Typical values have been tabulated for K 0 and are shown in T able 3 1.

PAGE 53

41 Table 3.1 : Typical values of the coefficient of lateral earth pressure, K 0 3. 2 Ground Water and the Lateral Earth Pressure Coefficient The discussion of lateral earth pressure from a natural soil perspective is applicable only if the soil remains in the current state and water is not introduced to the system. Since the presence of groundwater and specific groundwater levels can fluctuate over time, the lateral earth pressure coefficient K 0 is not a constant for a soil deposit or for a given time period. One way to address this variability is to express the lateral earth pressure coefficient in terms of effective stre ss. The equation now becomes 3.9 This expresses the in situ soil stress state in terms of effective stresses to allow for independence from the presence and level of the groundwater table. If the water table level changes the lateral earth pressure coefficient K 0 remains constant as long as we remain in the current soil layer with a constant density

PAGE 54

42 (Holtz & Kovacs, 1981). The coefficient of lateral earth pressure at rest, K 0 is sensitive to many things including the density of the overlying soil layer(s) and the geologic stress history of the soil itself. Where the soil is completely submerged, soil is below the groundwater table, the intergranular (effective) stress betwee n soil particles is reduced by the magnitude of the water pressure at that specific depth as follows: 3.10 where u is equal to the water pressure at the same depth. The presence of water acts to reduce the amount of load the soil intergranular particles accept by taking on some of the load itself. For a normally consolidated soil that is located above the groundwater table (water below depth H, F igure 3.5) the lateral at rest pressure acting against the wall would increase with de pth uniformly. The resulting distribution (Figure 3.5) would be represented by a triangle with the maximum pressure existing at the base of the triangle and equal to 3.11 and the resultant lateral force per unit of wal l length is 3.12

PAGE 55

43 Figure 3.5 : Subsurface Stresses Soil in At rest Condition In the case where the soil is below the groundwater table, the intergrannular or effective stress between the soil particles is reduced by the hydrostatic pressure of the water taken at the same depth 3.13 Where u is the magnitude of the water pressure at depth Z. The net effect of the soil being within the water table (submerged) is that compared to a non submerged soil given the same conditions, the lateral soil pressure is less.

PAGE 56

44 However, since the soil is located below the water table (submerged) you have to a ccount for the hydrostatic pressure of the water on the wall. The total effect on the wall lateral pressure due to the hydrostatic pressure and the lateral soil pressure imposes a larger total lateral force than a non submerged soil (Figure 3.6) Fig ure 3.6 : Subsurface Stresses Soil in At rest Condition w ith Vertical Surcharge and Influenced by the Water Table

PAGE 57

45 4. Rigid Wall Foundation Design The purpose of a reinforced concrete basement wall is multifold. It basement of the structure, and supports the structural loads which typically react vertically. Design inf ormation for basement foundations is scattered across multiple sources leaving design engineers to navigate and assimilate the appropriate information for a complete design. Most of the sources used today use prescriptive design practices and are related back to the Unified Soil Classification system for backfill loads. Typically, the design of basement walls uses a combination of the International Building Code (IBC), the International 8, Building Code Requirements for Structural Concrete 4.1 International Building Code The International Building Code (IBC), within multiple sections, outlines a prescriptive approach to basement wall design for walls that are supported at the top (floor diaphragm) and at the bottom (Keyed or doweled into a footing). Section 1610, Table 1610.1 provides soil lateral loads based on the Unified Soil Classification (USC) system. If the top of the wall is restrained from horizontal movement the design uses an at rest pressure. Surcharge loading is to be included in the analysis and if expansive soils are a concern a note indicates that the lateral pressure may increase. Section 1802 outlines foundation and soils investigations with some discussion on expansiv e soils.

PAGE 58

46 Section 1802.3.2 outlines the provisions for classifying an expansive soil as follows: 1) Plasticity Index (PI) greater than or equal to 15 in accordance with ASTM D4318 2) Greater than 10% of the soil particles passing a #200 sieve (75uM) in accordance with ASTM D 422 3) Greater than 10% of the soil particles less than 5uM in size in accordance with ASTM D4829 4) The Expansive Index greater than 20 in accordance with ASTM D4829 One caveat is included in the criteria that if #4 is true, the provisions describe d in #1, #2 and #3 are not required. Section 1804 outlines the load bearing pressure, lateral pressure and lateral sliding resistance capacity for soils (Table 1804.2) using the UCS system. Section 1805.2 outlines footing options and depths and section 1805.5 continues with the foundation wall design for concrete and masonry wall designs. Table 1805.5 (1) provides the minimum wall thickness for plain concrete and plain masonry walls. Tables 1805.5 (2,3,4) provide the vertical reinforcement requirements for 8 inch, 10 inch, and 12 inch wall thicknesses based on an unbalanced backfill and three groupings of soil types based on the UCS system. Included in the table are reinforcement location, grade of reinforcement, minimum concrete strength and alternate sizes of vertical reinforcement available for use. The provision of the code gives the required amount of vertical reinforcement to resist lateral forces or pressures

PAGE 59

47 given a general soil type. The design approach is based on a one way behavior of the ba sement wall. The wall acts as a vertical beam pinned at the top and bottom with the beam length equal to the height of the wall and the beam width equal to the length of the foundation wall. No explicit benefit is given to support at corners or bump outs The vertical pinned pinned beam design assumes a triangular soil loading profile, exerting pressure on the exterior surface of the basement wall. The resulting beam reacts in tension on the interior, unsupported wall and compression on the exterior of the beam. Reinforcing steel is prescribed on the interior side of the beam due to the tensile forces present at that location. The footing is assumed to provide no rotational resistance and is designed primarily to resist the gravity loading of the struc way of the floor. As part of the prescriptive approach, reinforcement amount of backfill being retained. Section 1805.8 addresses foundations in expansiv e soils but is limited to the design of slabs on grade and mitigation of the soils as opposed to the actual design of a structure to resist the forces applied. Section 1907 describes the details of the reinforcement to include hooks, the placement of rein forcement, protection, etc. and primarily references ACI 318 for the actual design details. 4.2 International Residential Code The International Residential Code, Chapter 4, addresses foundation design for one and two family dwellings. Section R401.4.1 re ferences Table R401.4.1 where presumptive load bearing values for foundation materials are

PAGE 60

48 given. Values in the table are based on the UCS system. Section R401.4 includes a note to direct soils testing for expansive, compressible, shifting or other unkno wn soil conditions without any distinct criteria specified. Section 402.2 outlines the minimum specified compressive strength of concrete for the type or location of the concrete construction against the weathering potential for the application. Section R403 continues with the sizing of the footings, described in section R403.1.1, and includes the minimum width of concrete or masonry footings in Table R403.1(1). Expansive soils, as related to the footing, are addressed in section R403.1.8 by referencing the International Building Code (IBC), section 1805.8. It gives the same provisions for classifying an expansive soil as defined in section 1802.3.2 of the IBC, namely testing for the Plasticity Index, soil fines and the Expansion Index. Section R404 out lines prescriptive solutions for the top reactions for a foundation wall based on UCS system for varying soil unbalanced backfill heights (Table R404.1(1)) and the maximum plate anchor bolt spacing based on unbalanced backfill height using the UCS system ( Table R404.1(2)). The section also specifies the maximum aspect ratio (l/W) for unbalanced backfill height based on the UCS system (Table R404.1(3)). Concrete foundation wall minimum vertical reinforcement size and spacing is addressed in Table R404.1(5) and is based on the unbalanced backfill height and the soil classification from the UCS system. Some of the vertical reinforcement design is referenced to the ACI 318 spec. The design approach outlined in the IRC is also based on a one way behavior of t he basement. The wall acts as a vertical beam pinned at the top and bottom with the beam length equal to the

PAGE 61

49 height of the wall and the beam width equal to the length of the foundation wall. Similar to the IBC, no explicit benefit is given to additional support at corners or bump outs. The vertical pinned pinned beam design assumes a triangular soil loading profile, exerting pressure on the exterior surface of the basement wall. 4.3 American Concrete Institute The American Concrete Institute ACI 318 (200 5) Building Code Requirements for Structural Concrete does not specifically address the design of basement foundation walls as a single design section It must be developed through the use of multiple sections. The design process typically begins with the definition of the structural loads outlined in Chapter 9 for various load combinations. Lateral soil loads are not addressed and must be determined using other geotechnical sources. Chapter 14 of the code specifically addresses the design and analysi s of concrete walls The design must satisfy s ections 14.2 (general requirements) and 14.3 ( minimum reinforcement requirements ) plus 14.4, 14.5 or 14.8 which provide design methods (only one of which is used in a given design problem). The shear design of the wall must satisfy requirements in s ection 11.10 ( s pecial provisions for walls) requirements. Vertical and horizontal reinforcement is designed in accordance with Chapter 14, section 14.4 for walls designed as compression members, All of the design literature requires that the top and bottom of the wall be restrained (pinned) and to satisfy the prescriptive design aids it must act as a pinned pinned connection. This is accomplished at the base by means of a concrete slab (floor) opposing movement of the base of the wall. In addition, the

PAGE 62

50 top is restrained by the floor diaphragm by means of the connection of the sill plate to the concrete wall, the connection of the sill plate to the floor framing and the framing and stiffness of the floor diaphragm itself.

PAGE 63

51 5 New Foundation Design Approach T wo structural designs were used in th is analysis. The first wa s a rectangular structure (Figure 5.1) that represents a typical two story, single family home. The structure measures 30 feet on its side by 50 feet in length. The wall maximum height was chosen as 9 feet with a maximum unbalanced backfill height of 8 feet. The thickness of the structural wall was chosen to be 12 inches conditions, wall height and the soil pressures exerted on the structure. The backfill soil was assumed to be CL at 60 psf per foot of depth. N o allowance was given for granular fill against the structural basement wall or drainage for the soil against the exterior wall. Figure 5.1 : Rectangular Structure and Soil Backfill

PAGE 64

52 The second structure was chosen to be a curvilinear structure in the shape of an ellipsoid (Figure 5.2). The minor axis was chosen to be a 15 foot radius and the major axis was chosen to be a 25 foot radius. This was to model the curvilinear structure as close to the same dimensions as the rectangular structure. The same wall thickness of 12 inches was used and the height of the structural wall was also 9 foot high. The soil backfill type, height and drainage was also chosen to be the same as the rectan gular structure at CL at 60 psf per foot of depth 8 foot backfill height and no drainage, respectively. Fig ure 5.2 : Curvilinear structure and soil backfill The following assumptions were made in the design of the structures: Each structure has a full basement floor 3 inches thick and tight against the bottom of the foundation with adequate stiffness to model a pinned joint

PAGE 65

53 Joists are used at the top of the wall and connected in such a manner to provide for full span support of the top of the wall the top s surface acts as diaphragm and is the pinned top support for wall T he lateral soil loading is identical on all sides of the structure (no unbalanced loading) Structure resides in a low seismic region and therefore seismic for ces do not control the design. The backfill against the structure on all walls is horizontal no sloped backfill. The total service level vertical l oad on wall = 1.5 kips/ft.

PAGE 66

54 6. Finite Element Analysis Foundation 6 1 LS DYNA Model LS DYNA is a general purpose finite element program with the ability to simulate highly nonlinear, transient dynamic finite element analysis using explicit or implicit time integration for complex real world problems. The software was developed by the Livermore Software Technology Corporation (LSTC), Livermore California. The LS DYNA code easily handles highly nonlinear, transient, dynamic finite element analys is using explicit time integration. Being a alyzing high speed, short duration events where inertial forces are important. The software code is especially applicable to soil structural interaction problems involving nonlinear soil materials with potentially large deformations. LS DYNA (version 971 Revision 7600.1224) contains numerous concrete material constitutive models that can be used for this analysis. When little is known about the concrete material properties, options to generate material constants given the unconfined compressive strength as an input can be very useful (e.g. MAT 159 CSCM). In this research, MAT Elastic 001 was e effects are not needed since loading of the concrete structure is achieved over relatively long spans of time. Since great variability exists in the field when

PAGE 67

55 constructing concrete foundations, defining concrete properties based on finely detailed inpu t parameters is n ot warranted. 6 2 Model Configuration concrete foundation designed in accordance with Chapter 4 of the 2006 International Residential Code. The foundation was designed with a wall thickness of 12 inches and a height of 9 feet (108 inches). Two foundation types were modeled, a typical rectangular structure (Figure 6.1) with a length of 50 feet (600 inches) and a width of 30 feet (360 inches) and a curvilinear structure Figure 6.1 : Rectangular Structure Finite Element Model (Figure 6.2) shaped in the form of an ellipse with a major diameter of 25 feet (300 inches) and a minor diameter of 15 feet (180 inches).

PAGE 68

56 Figure 6.2 : Curvilinear Structure Finite Element Model The curvilinear structure was designed to fit within the same footprint of the rectangular structure. The designs also included an escape window as defined in the 2006 International Residential Code, with a minimum of 5.7 ft 2 opening, 24 inch minimum heig ht and a 20 inch minimum width. The window was placed in the center of the longest wall to evaluate the stress state of the elements around the window. The soil backfill was modeled after a clay soil (Unified Soil Classification System of CH) with a back fill height of 8 feet (96 inches) around the structure with uniform depth of fill around the entire structure. The unbalanced backfill height as defined in the 2006 International Residential Code was calculated as 8.0 ft. The top of the concrete foundati on was modeled as a pinned joint, supported by a fixed diaphragm (framed floor) at the top of the wall structure. Motion was restricted in the plane of the wall in the lateral direction but allowed to move relative to the vertical. No rotational limitati ons

PAGE 69

57 were imposed on the structures. The bottom of the wall was restrained by a continuous 3 inch thick, cast in place concrete floor at the base of the foundation. The wall is restrained on the exterior by the soil mass. It was modeled as a pinned joint with rotational freedom at the upper and lower joints. Reinforcement used in the models was based on minimal vertical reinforcement outlined in the nes a minimum spacing of 18 inches on center for #4 vertical reinforcement. This results in an area of steel of 0.093% for a #4 bar. In the models, the incorporation of reinforcement was accomplished by using the Rule of Mixtures to define the composite properties rather than incorporating discrete beam elements in the model to account for the reinforcement. The models were limited to 10,000 total elements due to the licensing of the LS DYNA software. 6.2.1 Element Types All models and parts used constant stress solid brick elements for the concrete foundation and soil backfill material. The LS DYNA default solid element uses a single integration point. In addition, the model used two discrete parts for the concrete structure and the soil backfil l. 6.2. 2 Loading and Boundary Conditions Two types of loading were used in the modeling of the concrete foundations, a vertical structural load and an earth pressure load on the structure. The vertical structural load was applied to the nodes at the top of the

PAGE 70

58 foundation over a time interval of zero to 0.1 seconds and continuing constant to 10 seconds while applying the full vertical load (Figure 6.3). Figure 6.3 : Structural Loading of Rectangular and Curvilinear Foundations The specific load applied to the rectangular foundation nodes was 1500 lb/lf and was determined for each node by the following equation: 6.1 0 100 200 300 400 500 600 700 800 0.0 5.0 10.0 Force Pounds Time Seconds Rectangular Foundation Node Loading Curvilinear Foundation Node Loading

PAGE 71

59 The specific load applied to the curvilinear foundation nodes was 1500 lb/lf and was determined for each node by the following equation: 6.2 The soil loading was accomplished using two approaches 1. The first was the application of a gravity load to the soil mass surrounding the foundation utilizing the Load_Body_Z card This imposes a body force due to gravity on the soil mass The acceleration factor used was 1g or 32.2 ft/s 2 (386.4 in/s 2 ) in the model. 2. The second soil loading was the application of a lateral load to the soil mass in the X and Y directions to simulate the pressure applied by an expansive soil. The soil mass was extended 10 feet from the foundation walls on all sides of the foundation. This facilitated the placement of lateral loads on the structure in the X and Y directions which was important for the load application on the foundation walls The applied lo ad in both the X and Y directions was 5,000 psf. Both load ing condition s were applied from zero load to max loading at 4.0 seconds and continuing constant to 10 seconds to ensure stability of the loading (Figure 6.4).

PAGE 72

60 Figure 6.4 : Z Body Soil Loading The boundary conditions used for the models included co nstraining translational movement (X, Y and Z directions) for all of the soil and concrete foundation nodes located at the bottom of the models. This ensures that the foundation and soil cannot mov e due to the applied loading. In addition, the top of the concrete foundation was constrained in the X and Y direction to model the Joists used at the top of the wall and connected in such a manner to provide for full span support of the top of the wall the joists and floor acts as diaphragm and is the pinned top support for wall. 0 100 200 300 400 0.0 5.0 10.0 Z Acceleration In/S 2 Time Seconds Z Body Soil Loading ZBody Soil Loading

PAGE 73

61 6.2. 3 Contact Type LS DYNA defines a contact by identifying (using parts, part sets, segment sets, and/or node sets) what is to be checked for potential penetration of a slave node through a master segment. To accomplish this, LS DYNA searches for penetrations, using a numb er of different algorithms, at each time step throughout the analysis. When a penetration is found (penalty based contact) a force proportional to the penetration depth is applied to resist, and ultimately eliminate, the penetration. In this analysis a t wo way treatment of contact was chosen between the concrete wall and soil. This type of contact calls the subroutines twice which checks the slave nodes for penetration the first time and checks the master nodes for penetration through the slave segments the second time The treatment is therefore symmetric and the definition of the slave surface and master surface is arbitrary since both are called in the subroutine and the results end up the same. This results in an increase of computation time due to t he extra subroutine calls. The interface between the concrete wall and soil was modeled using the Surface_To_Surface contact card. The objective of the contact definition is to eliminate any penetration between the interfacing surfaces. The Slave Segmen t Set and Master Segment Set were set up to use the Part I.D. with the Master defined as the Soil. The Static Friction coefficient was defined as 0.4 and the Dynamic Friction coefficient was defined as 0.2 ( ETL 1110 3 446, 1992, Department of the Army ). The friction coefficients were based on a wet clay soil

PAGE 74

62 concrete forms are smooth and similar to a troweled surface. There is general agreement on these values in the existing l iterature although the friction coefficients are dependent on the type of clay, moisture content, density and the relative surface condition of the concrete. 6.2. 4 Material Properties The concrete properties used for this analysis were based on a normal c oncrete with a compressive strength of 3000 psi, density of 150 lb/ft 3 Ratio of 0.16 and a Modulus of Elasticity of 3.146 x 10 6 psi. In addition, the reinforcing steel used in the model had a minimum yield strength of 60,000 psi and a Modulus of Elasticity of 29 x 10 6 psi. The soil properties used for this analysis was based on a clay soil (Unified Soil Classification System CH) with a Unit Weight of 100 lb/ft 3 a Density of 1.498 x 10 4 lb/in 3 a Modulus of Elasticity of 2175 psi and a Poisso Ratio of 0.3. Table 6.1 : LS DYNA model material properties Material Name Rho (lb/in 3 ) (lb/in 2 ) Concrete /Steel Composite 2.247E 4 3.146E+6 0.16 Soil 1.498E 4 2.175E+3 0.30

PAGE 75

63 The concrete/steel properties were further combined to generate a composite material using the Rule of Mixtures (Callister, 2001). The Rule of Mixtures asserts that the properties of the combined material are a combination of the individual components of the individual materials. Typical composit es have two phases consisting of a matrix (continuous) phase and a dispersed (particulates, fibers) phase. The properties of the composite depends on the specific properties of the phases, the geometry of dispersed phase (particle size, distribution, orie ntation) and the amount of each phase in the composite. Composites are typically classified as particle reinforced (large particle and dispersion strengthened) composites, fiber reinforced (continuous (aligned) and short fibers (aligned or random) composi tes and structural (laminates and sandwich panels) composites. The rule of mixtures calculates an upper limit of the elastic modulus of the composite in terms of the elastic moduli of the matrix (E m ) and the particulate (E p ) phases by the equation: 6.3 where V m and V p are the volume fraction of the two phases The calculated composite elastic modulus for a 12 inch thick foundation wall with #4 reinforcement spaced at 18 inch on center is:

PAGE 76

64 6.4 The resulting Modulus of Elasticity for the composite material used in the model was 3,1 50 ksi

PAGE 77

65 7 Results 7 .1 General Behavior of the Foundation Structures A review of the structural behavior of the rectangular foundation indicates the model is behaving as a pinned pinned structural wall with combined vertical and horizontal loading. As shown in Figure 7.1, the deformed shape shows rigidity at the corners with a corresponding lack of displacement. It also shows the greatest displacement at the center of the longest walls exhibiting bowing to the interior due to the lateral pressure from the soil mass. The shape of the deformation in Figure 7.1 is exaggerated (5X displacement factor) to visually demonstrate the deformations. Figure 7.1 : Deformed Rectangular Basement Structure (5X Displacement Factor) Comparing the deformations of the rectangular foundation wall with that of the curvilinear foundation (Figure 7.2) under the same loading conditions, very

PAGE 78

66 little deformation along the entire wall (5X displacement factor) is observed. The load act s in a ne arly radial direction and is distributed semi uniformly around the periphery of the structure Truly r adial forces acting on a circular cross section result in only compressive forces on the section. There are no bending forces in a completely circular sec tion unless there is a discontinuity such as a window or door The modeled curvilinear structure capitalizes on this loading advantage by balancing constructability and space utilization of the structure with a shape that takes advantage of compressive fo rces to increase the load capacity of the structure. Figure 7.2 : Deformed Curvilinear Basement Structure (5x Displacement Factor) The finite element model indicates that a nearly compressive shape can easily be applied to basement foundation design.

PAGE 79

67 7 .2 Wall Displacements of the Foundation Structures The results of the finite element analysis demonstrate a considerable difference in the mid wall, out of plane lateral displacements between the rectangular basement wall and the curvilinear basement wall (Figure 7.3). Figure 7.3 : Lateral Deflection o f Rectangular Basement Structure The maximum displacement experienced by the rectangular wall was 2.67 inches inward at the top of the window. Similarly, the corresponding node on the opposite wall (without window) shows a deflection of 2.23 inches whic h was the second greatest deflection in the model. The magnitudes of all four elements analyzed at the mid span of the longest wall were very similar as shown in the fringe plot of the rectangular structure (Figure 7.4).

PAGE 80

68 Figure 7.4 : Lateral Deflection of Rectangular Basement Structure Fringe Plot Comparing the results of the rectangular wall design to the curvilinear wall design there are differences to note. The deflections are essentially zero inches (Figure 7.5) for mid span elements at the win dow and on the opposite wall of the curvilinear structure. These locations are representative of the same locations analyzed for the rectangular structure. Figure 7.5 : Lateral Deflection of Curvilinear Basement Structure

PAGE 81

69 All of the displacements within the curvilinear structure were less than 0.1 inches, an order of magnitude less than the rectangular structure (Figure 7.6). Figure 7.6 : Lateral Deflection of Curvilinear Basement Structure Fringe Plot It was also noted that the maximum defle ctions were located along the top of the window, similar to the rectangular structure, although all deflections were below 0.1 inches.

PAGE 82

70 7 .3 Vertical Stress State of the Foundation Structures In addition to the deflection of the walls, the stress state of each foundation type was analyzed. The Z direction stress (vertical plane of the wall) in the area of the window was analyzed between the two wall designs. The results of the rectangular wall structure indicate a compressive stress of 12 psi an d 8 psi in the two elements above the window and increasing larger stresses from 0 psi at the window sill to 66 psi at the bottom of the foundation, all in tension (Figure 7.7). Figure 7.7 : Z Stress (Vertical) Plot o f Window Area o f Rectangular Foundation The maximum stress in the rectangular structure is 412.9 psi in tension (Figure 7.8). This occurs at the lower corners of the foundation.

PAGE 83

71 Figure 7.8 : Z Stress (Vertical) o f Window Area o f Rectangular Foundation Fringe Plot Comparing t he results of the rectangular wall design to the curvilinear wall design the wall stresses are smaller in magnitude with a maximum stress for the Elements at the window opening ranging from 10.9 psi to 1.8 psi (Figure 7.9). Figure 7.9 : Z Stress (Vertica l) Plot o f Window Area o f Curvilinear Foundation

PAGE 84

72 In addition, the curvilinear structure has a maximum stress at 38.2 psi located at the upper right side of the window cut out. Of greater interest is that all of the vertical stresses in the curvilinear structure are compressive stresses (Figure 7.10) with the exception of element #547 at the top of the window opening. The greatest tensile stress is 1.9 psi which is a significant difference from the rectangular wall design. Figure 7.10 : Z Stress (Vertical) o f Window Area o f Curvilinear Foundation Fringe Plot 7 .4 Shear Stress State of the Foundation Structures The shear stress for the rectangular structure was evaluated against the curvilinear structure. The greatest shear stress in t he rectangular structure was determined to be 235.4 psi (Figure 7.12) located toward the upper corner of the structure.

PAGE 85

73 Figure 7.11 : Shear Stress Plot along Window Area o f Rectangular Foundation The highest shear stresses are located in the corner area s where the walls transition from X to Y directions. The lowest shear stress in the rectangular structure was 1.8 psi located on the upper side wall of the rectangular foundation.

PAGE 86

74 Figure 7.12 : Shear Stress along Window Area o f Rectangular Foundation Fringe Plot Com paring the rectangular foundation to the curvilinear foundation the maximum shear stress for the curvilinear structure is 21.2 psi (Figure 7.14) located along the bottom side of the foundation while the minimum shear stress is 1.4 psi and is located above the window.

PAGE 87

75 Figure 7.13 : Shear Stress Plot along Window Area o f Curvilinear Foundation The difference between the maximum shear stress in the rectangular structure and the curvilinear structure is approximately a factor of ten with the highest shear stresses exhibited in the rectangular structure. Figure 7.14 : Shear Stress along Window Area o f Curvilinear Foundation Fringe Plo t

PAGE 88

76 8 Discussion of Analyses Results Results from the analyses of the rectangular and curvilinear structural designs were analyzed and compared to determine the performance advantages of a curvilinear foundation in environments with high lateral soil loading in addition to the vertical structural loading. This included comparing the lateral displaceme nts of the walls, the vertical stress performance, the shear stress performance and the stress concentration distribution around a discontinuity (window) in each design with combined vertical and horizontal loading. 8 1 Displacements of Foundations The fir st observation regarding the lateral deflection of each design was the lateral displacements of the rectangular wall design were greater than 100X of the displacements for the curvilinear design for the same locations. Both foundations experienced the gre atest lateral deflections at mid wall of the long span section of the foundation. Four locations were studied for each design, two around the window and two opposite the window on the solid, continuous long wall.

PAGE 89

77 Table 8.1 : Lateral displacement (in) of Rectangular and Curvilinear Wall Designs This is a significant performance difference between the two structural designs and would be advantages from a concrete cracking performance standpoint. Th is might also have the added advantage of using less reinforcing steel in the design providing better economy of construction, although this would have to be further modeled and analyzed. In addition, the maximum displacement for the rectangular foundatio n occurred at the right, upper corner of the window at node 1637 and measured 2.67 inches while the maximum displacement for the curvilinear structure occurred at node 1137 and measured 0.06 inches at a location above the window (Figures 8.1 and 8.2).

PAGE 90

78 F igure 8.1 : Lateral Deflection of Rectangular Basement Structure Fringe Plot Figure 8.2 : Lateral Deflection of Curvilinear Basement Structure Fringe Plot The results of the lateral displacement analysis demonstrates that the curvilinear foundation design is a much better design regarding lateral deflections

PAGE 91

79 under combined vertical and lateral loading. All of the deflections for the curvilinear structure we re much smaller than the rectangular design with worst case deflections of the rectangular structure 100 times the magnitude of the curvilinear design. 8 2 Vertical Stress of Foundations In addition, the vertical stress performance was analyzed in both the rectangular and curvilinear models. A section of wall was analyzed through the depth of the wall located in the middle of the window opening. The results of the analysis demonstrate that the rectangular structure experiences both tensile and compressive forces within the structure due to the combined vertical and lateral loading (Table 8.2). The top of the foundation remains in compression above the window opening and at the window sill, while the bottom of the foundation below the window remains in ten sion. Table 8.2 : Vertical Stress of Rectangular and Curvilinear Wall Designs

PAGE 92

80 Comparing the performance of the rectangular structure to the curvilinear structure it is noted from the analysis that the curvilinear structure remains in compression through the depth of the foundation at the window, with the exception of the top of the window which is only slightly in tension at 1.8 psi. In addition, the maximum vertical stress in the rectangular structure is located at the left lower corner of the structure at element 1284 and measured 412.9 psi while the maximum vertical stress of the curvilinear structure, located at element 573, in the upper right corner of the window measured 38.2 psi. This represents a difference of greater than 10X in the maximum vertical stress level in the rectangular structure (Figures 8.3 and 8.4) as compared to the curvilinear structure. The maximum vertical stress for both structures was a compressive stress. Figure 8.3 : Z Stress (Vertical) of Window Area of Rectangular Foundation Fringe Plot

PAGE 93

81 Figure 8.4 : Z Stress (Vertical) o f Window Area o f Curvilinear Foundation Fringe Plot Focusing in on the discontinuous area created by the incorporation of the window in both the rectangular structure and the curvilinear structure, the higher vertical stresses peak along the side of the window for the r ectangular structure and at the upper corner of the window for the curvilinear structure (Figures 8.5 and 8.6).

PAGE 94

82 Figure 8.5 : Vertical Stress around Window Rectangular Structure

PAGE 95

83 Figure 8.6 : Vertical Stress around Window Curvilinear Structure The maximum vertical stress in th e area around the window for the rectangular structure was 92.7 psi while the maximum vertical stress for the curvilinear structure for the same area was 38.2 psi. The rectangular structure around the window was primarily in t ension while the curvilinear was in compression. Evaluating the stress concentration factors in that same area around the corner of the window (Table 8.3 and Table 8.4) for both the rectangular and curvilinear designs the stress concentration factors for the rectangular structure range from 0.88 to 190.18 while the curvilinear structure

PAGE 96

84 ranged from 1.04 to 2.71. This represents a huge difference on structural performance relative to the applied vertical stress around the discontinuity of the window openin g. Table 8.3 : Stress Concentration Factors around Window Rectangular Table 8.4 : Stress Concentration Factors around Window Curvilinear Curvilinear Foundation Design Vertical Stress Element ID Window Z Stress (psi) Element ID No Window Z Stress (psi) Stress Concentration Factor 538 13.4 1339 12.9 1.04 547 1.8 1330 13.1 1.14 556 1.9 1321 13 1.15 565 14.2 1312 13.3 1.07 574 27.3 1294 12.6 2.17 573 38.2 1293 14.1 2.71 572 25.9 1292 14.9 1.74 571 25.4 1291 15.9 1.60 The performance improvement around the window in the curvilinear foundation demonstrates the advantage of designing a foundation to remain in

PAGE 97

85 compression throughout the structure and the improvement in performance of crack initiatio n around a high stress concentration area. 8 3 Shear Stress of Foundations When comparing the shear stress performance of the rectangular and curvilinear structures, a vertical section of wall was selected along the side of the window for each structural design. Overall, the rectangular structure exhibited much greater shear stresses for the same locations in the structure, as much as 10X for some locations (Table 8. 5 ). Table 8.5 : Shear Stress of Rectangular and Curvilinear Wall Designs

PAGE 98

86 The maximum shear stress for the rectangular structure was 235.4 psi located at element 219 in the upper corner of the structure while the maximum shear stress for the curvilinear structure was 21.2 psi and was located at element 864 on the bottom of the structure (Figures 8.7 and 8.8). Figure 8.7 : Shear Stress along Window Area o f Rectangular Foundation Fringe Plot

PAGE 99

87 Figure 8.8 : Shear Stress along Window Area o f Curvilinear Foundation Fringe Plot Evaluating the shear str esses for the two structures around the area of discontinuity at the window, the shear stresses are greatest above the window for the rectangular structure and at the upper corner of the window for the curvilinear structure (Figures 8.9 and 8.10).

PAGE 100

88 Figur e 8.9 : Shear Stress around Window Rectangular Structure

PAGE 101

89 Figure 8.10 : Shear Stress around Window Curvilinear Structure Evaluating the stress concentration factors in the same area around the corner of the window (Table 8. 6 and Table 8. 7 ), for both the rectangular and curvilinear designs, the stress concentration factors for the rectangular structure range from 61.2 to 177.8 while the curvilinear structure ranged from 6.3 to 7.8. This represents a huge difference on structural performanc e around the discontinuity of the window opening. Since this area of the structure is most vulnerable, due to the discontinuity of the window, the large reductions in the stress concentration factors demonstrated in the curvilinear design help to mitigate against cracking in this area.

PAGE 102

90 Table 8.6 : Stress Concentration Factors around Window Rectangular Rectangular Foundation Design Shear Stress Element ID Window Z Stress (psi) Element ID No Window Z Stress (psi) Stress Concentration Factor 875 204.1 1406 61.6 3.31 857 207.6 1388 61.3 3.39 839 207.2 1370 61.2 3.39 821 203.4 1352 61.6 3.30 803 148.9 1334 62.3 2.39 802 51.7 1333 98.6 0.52 801 35 1332 177.8 0.20 800 92.5 1331 103 0.90 Table 8.7 : Stress Concentration Factors around Window Curvilinear Curvilinear Foundation Design Shear Stress Element ID Window Z Stress (psi) Element ID No Window Z Stress (psi) Stress Concentration Factor 538 10 1339 6.8 1.47 547 1.4 1330 6.6 0.21 556 1.8 1321 6.6 0.27 565 10.3 1312 6.9 1.49 574 14 1294 6.3 2.22 573 19.9 1293 7.1 2.80 572 14.7 1292 7.4 1.99 571 13 1291 7.8 1.67

PAGE 103

91 9 Conclusions and Recommendations for Further Research In the present study the analysis of a rectangular and curvilinear foundation was undertaken to demonstrate r esults from the evaluation of these structures under combined lateral and vertical loading confirm From the LS DYNA analysis of the rectangular a nd curvilinear structures, it is demonstrated that a basement foundation wall can be constructed in such a manner that all of the internal forces within the foundation are compressive forces. This design, using a curvilinear structure, takes advantage of resist potentially high lateral forces exerted by the soil mass. The advantage to this type of design is decreased lateral wall deflections around the circumference of the foundation leading to a better perf ormance of the foundation even in the presence of swelling soils. The results of this study have led to several significant observations about the performance of rectangular and curvilinear structures under combined vertical and lateral loading. Key obse rvations from the LS DYNA models are the following: The performance of a curvilinear structural design in light building foundations under combined lateral and vertical loading is superior to the traditional rectangular wall design. The curvilinear design remains almost entirely in compression due to the combined loading which takes advantage of the best design properties of the material. The curvilinear foundation design outperforms the rectangular structure by 100X when analyzed for lateral displacements under identical loading. The maximum displacement for the rectangular foundation occurred at the

PAGE 104

92 right, upper corner of the window at node 1637 and measured 2.67 inches while the maximum displacement for the curvilinear structure occurred at node 1137 an d measured 0.06 inches at a location above the window. The rectangular structure experiences both tensile and compressive forces within the structure due to the combined vertical and lateral loading. The curvilinear remains in compression throughout the s tructure, with the exception of the top of the window which is only slightly in tension at 1.8 psi. The maximum vertical stress in the rectangular structure measured 412.9 psi while the maximum vertical stress of the curvilinear structure measured 38.2 psi This represents a difference of greater than 10X in the maximum vertical stress level in the rectangular as compared to the curvilinear structure. The maximum vertical stresses in both structures were compressive stresses. Evaluation of the discontinuo us area created by the incorporation of the window in both structures, the higher vertical stresses peak along the side of the window for the rectangular structure and at the upper corner of the window for the curvilinear structure. The maximum vertical s tress in this area around the window for the rectangular structure was 92.7 psi while the maximum vertical stress for the curvilinear structure for the same area was 38.2 psi. The rectangular structure around the window was primarily in tension while the curvilinear was in compression.

PAGE 105

93 The rectangular structure exhibited much greater shear stresses for the same locations in the structure, as much as 10X for some locations. The maximum shear stress for the rectangular structure was 235.4 psi while the maxi mum shear stress for the curvilinear structure was 21.2 psi. Current mitigation techniques for expansive soils include adj usting drainage underpinning or mitigation of the design by using pier and beam foundations with drainage implemented at the foundati on to prevent soil expansion. These strategies assume a stable soil condition over the life of the foundation and do not take into account the potential changing soil conditions at the foundation wall over time Incorporating a design that resists the fo rces exerted by a swelling soil ensures survivability of the structure even if drainage fails and expansive soils are present at the foundation. Although the current industry standards for mitigation of light structures constructed in expansive soils, as described above, offer piece of mind against catastrophic damage, long term stability is not guaranteed. Over time, severe drought and flooding cycles can directly influence the foundation performance and exceed the design mitigations implemented. In add ition, changes in soil chemistry can occur during periods of high moisture exposure negating the benefits of soil treatment methodologies. Over time, changes in soil drying can occur due to rises in the ambient temperatures and/or the growth of vegetation within the soil mass. With unprecedented weather events taking place all over the globe, the design of structures must rely less on mitigation of moisture

PAGE 106

94 intrusion and more on foundation designs that take advantage of the potential forces mobilized by t he soil. Based on the present work additional analysis is required in the finite element analysis of the two structural types. The following are recommendations for continued research: A better understanding of the sensitivity of the ellipse dimensions ve rsus the introduction of tensile forces within the structural wall would be helpful in refining the design for the optimum use of space and constructability. T he development of an expansive soil model in LS DYNA to provide a more accurate soil model that t akes into consideration the saturation state of the soil and the changing soil pressures as a function of moisture content and density would help in more accurate modeling of the lateral forces. This should be combined with experiments that validate the s oil model and the LS DYNA results. A more accurate modeling the reinforced concrete foundation by incorporating the rebar reinforcement using beam elements within the solid element structure of the concrete. This would replace the smeared properties used in this analysis. The i ntroduc tion of piers for the foundation support would complete the design model for the curvilinear foundation and enable the analysis of a complete foundation system.

PAGE 107

95 The experimental derivation of the static and dynamic coefficients of friction between the subject soil and concrete foundation to more accurately model the physical contact.

PAGE 108

96 REFERENCES American Concrete Institute (2005). Building Code Requirements for Structural Concrete (ACI 318 05) Farmington Hills, MI. American Society of Civil Engineers. (2005). Minimum Design Loads for Buildings and Other Structures. Danvers, MA. Das, B. M. (2002). Soil mechanics laboratory manual. (6 th ed.). New York: Oxford University Press. Donaldson, G. W. (1969). The Occurrence of Problems of Heave and the Factors Affecting its Nature. 2nd International Research and Engineering Conference on Expansive Clay Soils. Texas Press. ETL 1110 3 446, 1992, Department of the Army, U.S. Army Corps of Engineers, Engineering and Design Revision of Thrust Block Criteria in TM 5 813 5/AFM 88 10, Vol. 5, Appendix C, Department of the Army, Washington, D.C. Grim, R. E. (1968). Clay Mineralogy: International Series in the Earth and Planetary Sciences New Y ork: McGraw Hill Book Company. Hardy, R. M. (1965). Identification and Performance of Swelling Soil Types. Canadian Geotechnical Journal, 11 (2),141 161. Holtz, W.G., and Gibbs, H.J. (1956). Engineering properties of expansive clays. Transactions of ASC E, 121 641 663. Holtz, R. D., and Kovacs, W. D. (1981). An Introduction to Geotechnical Engineering. New Jersey: Prentice Hall. International Code Council ( 200 6). International Building Code Country Club Hills, IL. International Code Council ( 200 6). International Residential Code for One and Two Family Dwellings. Country Club Hills, IL. Howard, A. K. (1977). Laboratory Classification of Soils: Unified Soil Classification System. Earth Sciences Training Manual, no. 4, U.S. Bureau of Reclamation, D enver, CO. Jones, D.E J., and Holtz, W.G. (1973) Expansive Soils The hidden Disaster. Civil Engineering 43. Lambe, T. W. (1953) The Structure of Inorganic Soil. Proceedings of the American Society of Civil Engineers 79, (pp. 49).

PAGE 109

97 Lambe, T. W. and Wh itman, R. V. (1969), Soil Mechanics New York: John Wiley & Sons. McCarthy, David F. (1998). Essentials of Soil Mechanics and Foundations Columbus, Ohio: Prentice Hall. Mitchel J. K. (1976). Fundamentals of Soil Behaviour. New York: John Wiley & Sons. Mitchel J. K. (1993). Fundamentals of Soil Behaviour. (2 nd ed.). New York: John Wiley & Sons. Nelson, J. D., and Miller, D. J. (1992). Expansive Soils, Problems and Practice in Foundation and Pavement Engineering New York: John Wiley & Sons. Seed, H. B., Woodward, R. J., and Lundgren, R. (1962) Prediction of Swelling Potential for Compacted Clays: ASCE Journal of Soil Mechanics and Foundations Division SM 3, Part 1, 53 87. Skempton, A.W. (1953). The Colloidal Activity of Clay. Proceedings of the Third International, Conference on Soil Mechanics and Foundation Engineering 1 57 60 Steinberg, M. (1998). Geomembranes and the Control of Expansive Soils in Construction New York: McGraw Hill. U.S. Army Corps of Engineers. (1970). Laboratory S oils Testing EM 1110 2 1906, Appendix V.