An analysis of a full-scale drilled pier foundation in expansive soil

Material Information

An analysis of a full-scale drilled pier foundation in expansive soil
Rogers, Stephen Curtis
Place of Publication:
Denver, Colo.
University of Colorado Denver
Publication Date:
Physical Description:
xi, 101 leaves : illustrations (some color) ; 28 cm

Thesis/Dissertation Information

Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Civil Engineering, CU Denver
Degree Disciplines:
Civil Engineering
Committee Chair:
Chang, Nien Y.
Committee Members:
Wong, Trevor
Brady, Brian


Subjects / Keywords:
Soil-structure interaction ( lcsh )
Swelling soils ( lcsh )
Foundations ( lcsh )
Piling (Civil engineering) ( lcsh )
Foundations ( fast )
Piling (Civil engineering) ( fast )
Soil-structure interaction ( fast )
Swelling soils ( fast )
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )


Includes bibliographical references (leaves 99-101).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Stephen Curtis Rogers.

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:
45264702 ( OCLC )
LD1190.E53 2000m .R64 ( lcc )

Full Text
Stephen Curtis Rogers
B.S., University of Colorado at Boulder, 1994
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering

This thesis for the Master of Science
degree by
Stephen Curtis Rogers
has been approved
Nien Y. Chang

Brian Brady

Rogers, Stephen C. (M.S., Civil Engineering)
An Analysis of a Full-Scale Drilled Pier Foundation in Expansive Soil.
Thesis directed by Professor Nien Y. Chang.
The literature is full of documented cases of damage from swelling soils on structures
of all types. This demonstrates the low level of designers understandings of the
complex mechanism of soil heave and its effect on foundations such as drilled piers
and post-tensioned slabs. This is demonstrated by the fact that damage to engineered
structures supported on drilled piers still occurs regularly. This may be due to the fact
that most of the research regarding drilled piers has been done in a laboratory, where
duplication of real-world conditions is difficult.
Many factors influence the design and analysis of drilled piers such as active zone
depth, swell index, swell pressure, side frictional resistance, magnitude of uplift
pressure, etc. The lack of understanding of these parameters leads to damage of
engineered structures on piers. This makes field performance study very significant
toward better engineering of structures on expansive soils. The in-situ performance
study presented here was designed to measure uplift on a drilled pier in an actual,
full-scale installation. The researcher designed and installed instrumentation in a

foundation element for a three story office building in Douglas County, Colorado.
Additionally, a second, non-loaded pier, or dummy pier, was installed to check the
ultimate uplift potential of the soil. The parameters that were measured and/or
calculated from those measurements were: l)the strain within the foundation
elements, 2)the uplift force on the piers from the swelling soils, 3)the moisture
variation due to construction disturbances, 4)the swell pressures from the soil, and
5)the coefficient of interface friction between the concrete pier and surrounding soil.
The results show that the non-homogeneity of the soil deposit has contributed to the
difficulty is assessing the uplift generated on the piers. It is shown that the
withholding force within the zone of stable moisture content contributes the majority
of the stabilizing force on the foundation element as the upper layer of soil swells.
Additionally, the results show that active zone depths below structures in Colorado
may reach 9 feet. The underestimation of this depth can lead to significant damage to
engineered structures.
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.

Nien Y. Chang

My sincere thanks go to my advisor, Nien Y. Chang, for his direction and persistence
with me. I also extend my thanks to my committee and the staff of the Graduate
School for their support and words of encouragement. Most importantly, I would like
to thank my wife; her unconditional support has truly made this possible.

1. Introduction......................................................1
1.1 Problems with swelling soils......................................1
1.2 Objectives of research...........................................2
1.3 Significance of research.........................................3
1.4 Scope of study...................................................3
2. Literature Review.................................................5
2.1 Clay Mineralogy..................................................5
2.1.1 Nature of soils.................................................5
2.1.2 Formation of clay minerals.....................................7
2.1.3 Clay-water interaction........................................10
2.2 Distribution of Soil in Colorado................................10
2.2.1 The Front Range................................................11
2.2.2 The Hogback belt..............................................11
2.2.3 The Colorado Piedmont.........................................11
2.2.4 Expansive soil distribution along The Front Range.............16
2.3 Characteristics of swelling soils...............................19
2.3.1 Intrinsic properties...........................................20
2.3.2 Environmental conditions......................................23
2.4 Drilled Pier Research...........................................28
2.4.1 Load carrying capacity.........................................29
2.4.2 Uplift resistance.............................................36

3. Materials and Methods..............................................41
3.1 Experimental Design and Approach..................................41
3.1.1 Purpose.........................................................41
3.1.2 Proposed methods................................................41
3.1.3 Project parameters..............................................43
3.2 Selection of gauges...............................................46
3.2.1 Introduction....................................................46
3.2.2 Environmental considerations....................................48
3.2.3 Durability......................................................48
3.2.4 Constructibility................................................48
3.2.5 Test material...................................................49
3.2.6 Gauge resistance................................................50
3.2.7 Gauge parameters................................................50
3.3 Analysis of pipe capacity.........................................53
3.3.1 Preliminary calculations.........................................53
3.3.2 Analysis of pipe capacity.......................................54
3.4 Laboratory installation and calibration...........................55
3.4.1 Introduction.....................................................55
3.4.2 Preparation................................................... 55
3.4.3 Gauge installation..............................................56
3.4.4 Wiring configuration and installation...........................58
3.4.5 Weather protection..............................................58
3.4.6 Calibration.....................................................60
3.5 Field conditions and installation.................................64
3.5.1 Site overview....................................................64
3.5.2 Pier installation...............................................64
3.6 Field measurement.................................................65
3.6.1 Introduction.....................................................65

3.6.2 Strain measurement.........................................68
3.6.3 Soil moisture measurement..................................68
4. Analysis of Results...........................................71
4.1 General predictions and background...........................71
4.1.1 In-situ conditions.........................................71
4.1.2 Site conditions during construction........................72
4.1.3 Predictions................................................76
4.1.4 Transformed sections.......................................77
4.2 Dummy pier..................................................78
4.2.1 Introduction...............................................78
4.2.2 Results....................................................79
4.2.3 Comparison to predictions..................................85
4.2.4 Conclusions................................................87
4.3 Loaded Pier..................................................88
4.3.1 Introduction...............................................88
4.3.2 Results....................................................89
4.3.3 Conclusions................................................93
4.4 Overall Conclusions.........................................94
5. Conclusions...................................................95
5.1 Summary......................................................95
5.2 Conclusions..................................................96
5.3 Recommendations for future research..........................97

2.1 Silica Tetrahedron and Aluminum Octahedron molecular models.........6
2.2 Graphical representation of bonding configurations in 1:1 clay
(Kaolinites) and 2:1 clay (Smectite) structures..................9
2.3 Front Range Urban Corridor landforms area map.......................12
2.4 Location map of Front Range Urban Corridor..........................17
2.5 Geologic map of the Front Range Urban Corridor......................18
2.6 Graph of interface friction angle, a, versus shear strength in clay.33
2.7 Free body diagram showing the forces from uplift and withholding on
a rigid pier...................................................37
3.1 Testing apparatus diagram ........................................47
3.2 Wiring schematic diagram..........................................59
3.3 Photos of testing apparatus prior to installation.................61
3.4 Test gauge calibration curves.....................................63
3.5 Photos of testing apparatus installation.......................66,67
3.6 Plot of moisture versus depth at before and after construction....70
4.1 Site map of subject site with superimposed test boring locations..73

4.2 Boring log for test boring closest test apparatus installation with soil
identification legend.............................................74
4.3 Test results of microstrain versus days elapsed for dummy pier .......80
4.4 Test results of microstrain versus days elapsed for loaded pier.......90

2.1 Standard clay mineral group chemical and physical composition
2.2 Movement required to mobilize full skin friction capacity.29
2.3 Movement required to mobilize full end bearing capacity.30
4.1 Summary of laboratory results.............................75

1. Introduction
1.1 Problems with Swelling Soils
Swelling soils present major problems for engineers throughout the United States and
the world. A 1982 report by the Federal Emergency Management Agency (FEMA)
totaled losses due to swelling soil damage at $798.1 million for the year 1970. The
losses were predicted to reach $997.1 million for the year 2000. The Department of
Housing and Urban Development (HUD) estimated that the losses due to shrink-swell
of soils were $9 billion in 1981. (Nelson, 1992) These numbers make it impossible to
ignore the need for more research for better design practices for foundations.
In Colorado, swelling soils along the Front Range corridor have created problems for
years. With the continued expansion of the population more and more sites that were
considered unsuitable for construction are slated for development. Without good
design practice this expansion could create significant problems for years to come.
The problem with swelling soils is that the expansive pressures generated are much
greater than most structures exert back as dead load. This is especially true for light
structures such as residential homes and small offices as well as for infrastructure
such as roads and bridges. Additionally, these soils typically swell in a very non-

uniform manner, creating differential heave that cracks foundations and pavements
Foundation design over the years has evolved to accommodate these problems. The
most common foundation design used to carry structural loads and avoid swelling soil
induced uplift is the cast-in-place drilled pier. This structure is designed to carry the
load of the structure through the zone of swelling soil to more stable strata below.
Additionally, due to dead load and skin friction, the pier is designed to resist uplift
from the soil in the zone of unstable moisture nearest the surface. This type of
design, however, entails many assumptions and empirical results. While these have
advanced the practice of drilled pier design, many problems are still encountered,
highlighting the need for additional research.
1.2 Objectives of Research
This study was developed to check some of the assumptions that designers routinely
use in designing drilled pier foundations. In particular the objective of this thesis is to
provide an understanding of the performance of a drilled pier in expansive soil in a
full-scale environment. The results of this research will hopefully give some
quantitative measurements for the soil pier interface friction angle, a. Additionally,
this work will provide information about the effect of swell of a non-homogenous soil
deposit on an actual foundation. Particularly this will include the effect of a

completed structure on the moisture content of the soil deposit and the effect of
commercial construction practices on the performance of the drilled pier.
1.3 Significance of research
To date, much research has been done on drilled piers, especially as this affects
performance under swelling conditions, on laboratory scale models. (Woodward,, 1972) This study will be one of the very few to evaluate real-world
conditions. Also, as this instrumentation will be installed in an permanent structure,
the results from this experimentation provides valuable feedback and even early
warning of foundation problems for the building owner. This may constitute the first
design build monitor relationship of this scope done in this area.
1.4 Scope of this study
Under this research, a testing apparatus for monitoring strain in a drilled pier was
designed for installation into a foundation element for a three-story office building in
Douglas County, Colorado. This apparatus was designed so as not to affect the
structural integrity of the foundation element.
Additionally, a second testing apparatus was installed in an unloaded pier in close
proximity to the loaded one. This was done to measure free-field heave and
provide a datum against which the results from the loaded pier could be compared.

Gauges were installed on both apparatus to measure strain at two-foot intervals to
measure the effect of the non-homogeneity of the soil deposit and non-uniformity of
uplift. Additionally, moisture samples were taken before and after construction of the
building to compare the effects of moisture on the pier performance.
Before the results were presented, an in-depth review of the literature was done to
provide background for the results and guidelines for presenting the data.
Finally, the results are presented and variances from expectations are discussed. A
brief presentation of suggested additional research is included to convey the need for
additional information.

2. Literature review
2.1 Clay Mineralogy
The evolution of a clay particle is complex but the understanding of this process is
critical in evaluating the behavior of expansive soils. To begin, one must look at the
chemical makeup of the basic building blocks of soils.
2.1.1 Nature of soils
Little (1995) reports that the soil in the top 10 miles of the earths crust are made up
primarily of oxygen (47.3%), Silicon (27.7%), and Aluminum (7.8%). Most of the
balance is made up of metals such as Iron (4.5%), Calcium (3.5%), Sodium (2.5%),
Potassium (2.5%) and Magnesium (2.2%). The primary building blocks of soils are
the silica tetrahedron and the aluminum octahedron. These two building blocks use
the three primary minerals available in the earths crust. Graphical representations of
these two configurations are shown in Figure 2.1. The silica tetrahedron forms into
frameworks through strong primary valence bonding by sharing the oxygen atom.
This well bonded framework is the primary makeup of stable materials such as quartz
and feldspars. Quartz is an extremely stable soil, one that will typically maintain its
chemical composition after weathered into extremely small particles. Feldspars,

O Oxygen atom O Hydroxyl
# Silicon atom # Aluminum or similar metallic ion
Aluminum Octahedron (b).
Figure 2.1 Silica Tetrahedron and Aluminum Octahedron
(after Little, 1995)

conversely, rarely maintain their composition when weathered to such a degree and
typically produce some type of clay mineral.
The difference between quartz and feldspar is that quartz is a pure silica tetrahedral
composition. Feldspars result from a portion of the silica atoms within the framework
being replaced by an aluminum atom. As the aluminum atom has a plus 3 valence
compared to the plus 4 valence of the silicon atom, a net negative charge results in the
framework. This imbalance is corrected by the adsorption of a single charge cation
such as potassium, calcium or sodium. The feldspar is a weaker, more weatherable
mineral due to the increased size of the aluminum atom with respect to the silicon
atom and the relatively weak bonds that exist between the single charged cations at
the surface.
2.1.2 Formation of clay minerals
As previously mentioned, the weathering of feldspars is the primary source for clay
minerals. Additional sources are the weathering of micas and limestone. The highly
swelling soils of the montmorillonite group are specifically formed from the
weathering of volcanic ash. (Friedman, 1993) Hydrolysis is the most common form
of weathering. The dissociated ions present in water will replace the cations present
in the feldspar resulting in a dissociated layer of water around the feldspar mineral.
The resulting layers between frameworks become unstable and the mineral breaks

apart in sheets. These tetrahedral and octahedral sheets may bond directly together
through strong hydrogen bonds or they may be held together by the weaker bonds
between the free cations associated with the water layer around the plate. The two
types of bonding are demonstrated in Figure 2.2. To explain why the different
formation may occur, one must look at another action happening during the
weathering process.
During weathering, some clay minerals experience a phenomenon know as
isomorphous substitution. In solution, some of the silicon atoms in the tetrahedral
sheet or aluminum atoms in the octahedral sheet may be replaced by metallic ions of
a lower positive charge. This, of course, results in a negative charge at the surface of
the particle. Those minerals that experience very little of this substitution are left to
form strong hydrogen bonds with each other, resulting in a stable material. Kaolinite
is an example of this. In a kaolinite, roughly 1 in 400 silicon atoms will be replaced
by another less positive ion. Conversely, in a smectite mineral, 1 in 6 aluminum
atoms may be replaced by magnesium, resulting in a very negatively charged particle.
In comparison, the negative charge of a smectite particle may be 10 times that of a
kaolinite particle. (Little, 1995)

Basic Units of the 1:1 Mineral Kaounite (a) are Linked with Relatively Strong Hydrogen
Bonds Which Retain a Hich Degree of Moisture Stability among Layers While the Basic Units of the
2:1 Smectite Mineral (b) are Linked by Weak Cation attraction. The Efficiency of This Linkage is a
Function of the Type and Concentration of the available Cations.
Figure 2.2 Bonding Configurations of Kaolinite and Smectite Groups
(after Little, 1995)

2.1.3 Clay water interaction
The clay particles in the smectite group, due to their high negative charge, pick up
many single charged cations at the surface which in turn form weak bonds with other
smectite particles. The cation linkage is a very weak bond and results in particles
with prominent planes of cleavage and very high surface areas. These weak bonds
are easily broken by the intrusion of the H+ ion in water. This coupled with the
extremely high relative surface area of these clay minerals yields a highly plastic,
unstable compound. With some smectite clay particles having surface areas
approaching 800 m per gram of material, this material has been documented to hold
nearly 7 times its dry unit weight in water. (Little, 1995) Montmorillonite is a typical
example of such a mineral. Bentonite, a montmorillonite clay, is common in
Colorado and is the source of much of the swelling soil damage here.
2.2 Distribution of Soil in Colorado
The Front Range Urban Corridor is the area bounded by the Rocky Mountains to the
west and the Colorado Piedmont section of the Great Plains on the east and centered
on Metro Denver. The boundary to the west is occupied by a number of sharp-crested
hogbacks while the Denver Basin to the east is much more subdued topographically.
(Hansen, 1982) An analysis of the formation of these features provides an insight
into the soil conditions designers deal with today in Colorado. Figure 2.3 shows the
various landforms within the Front Range Urban Corridor.

2.2.1 The Front Range
The mountains to the west known as the Front Range were formed from uplift and
subsequent erosive resistance of the rock formations under the sedimentary cover of
the High Plains. The sharp peaks rise 1450 to 1525 meters above the neighboring
Piedmont to the east. The mountains to the west rise higher and contain less
pronounced peaks and valleys. (Hansen, 1982)
2.2.2 The Hogback Belt
The Hogback Belt sits immediately to the east of the Front Range. It rises up to 300
meters above the adjacent Piedmont to the east. This formation is essentially a large
rock outcropping with the sedimentary cover eroded away. It is separated from the
Front Range by eroded valleys. The incline of the hogback is steepest nearest the
mountain front but the dipping rock beds exist to less a degree eastward under the
Piedmont to the boundary of the High Plains.
2.2.3 The Colorado Piedmont
The Colorado Piedmont occupies the space between the Hogback Belt and the
western boundary of the High Plains. The eastern boundary of the Piedmont is about
113 km southeast of Denver, therefore the Denver area is actually on the Colorado

Front Range
Hogback belt
Colorado Piedmont
Piedmont lowlands
Floodplains and tenaccs
Mantled lowlands
Piedmont uplands
Bench-aod-valley uplands
Low rolling uplands
Complex uplands
Composite uplands
Figure 2.3 Landforms of The Front Range Urban Corridor
(after Hansen, 1982)

Piedmont, not the High Plains as widely believed. The Piedmont was formed by
erosion of the High Plains when this formation extended to the mountain front. Some
areas, such as the Palmer Divide between Castle Rock and Colorado Springs exist
near the original ground surface elevation. (Hansen) The subdued topography has
been created by localized deposition of fine material between outcroppings. It is
these deposits that create problems with swelling soils. The Colorado Piedmont can
be divided into two general classifications: lowland and uplands. Piedmont Lowlands
The Piedmont lowlands are divided into Flood Plains and Terraces and Mantled
lowlands. The flood plains and terraces exist along the lower reaches of the South
Platte river and its tributaries. The flood plains are typically 1.5 to 3 km wide while
the accompanying terraces may be as wide as 3 km. The terraces are
characteristically very flat while the flooplains are varied in width and surface
irregularities due to erosion.
The mantled lowlands exist downstream of the floodplains and terraces, exclusively
to the northeast of the Denver Metro area. This area is has very low surface
irregularities due to large eolian deposits of sand and silt. (Hansen)

2.2.32 Piedmont Uplands
Hansen described the Uplands as all that area above the floodplains and terraces and
accompanying mantled lowlands. There are several formation types that are
identified as Bench and Valley uplands, Low Rolling Uplands, Complex Uplands,
and Composite Uplands. The transitions between each are typically gradational. Bench and Valley Uplands
The Bench and Valley Uplands exist nearest the hogback belt. These consist benches
ranging from 3 to 16 km wide that slope gradually to the east. They are randomly
marked by gullies and trenches from erosion and may have locally steep, smooth
slopes. Where the bench has been separated from the mountain front by erosion it
forms a mesa or butte. The tops of these benches, mesas, and buttes are typically
capped with gravel left behind from streams flowing over these surfaces in the
geologic past. Low Rolling Uplands
The low rolling uplands exists from the northern boundary of the Front Range south
to a point roughly 24 to 29 km south of Denver. They are bounded on the west by the
Bench and Valley Uplands and extend to the border of the High Plains. The
topography of this formation is gentle due to deposits of eolian sand and silt. The

bedrock is exposed locally and exists at varying depths below the surface elsewhere.
Local relief is generally less than 30 meters. (Hansen) Complex Uplands
As the name implies the Complex Uplands consist of a diversity of landforms. The
surface is generally controlled by variations in the bedrock elevation. This formation
extends from the southern boundary of the Low Rolling Uplands south across the
Palmer Divide. The surface is generally rocky or consists of residuum from the
weathering of the bedrock itself. Very little mantling from eolian deposits is present.
The formation is generally divisible further into two categories consisting of mesas
and buttes and ridges between drainage basins. (Hansen) This formation is similar to
both the Low Rolling Uplands and the Bench and Valley Uplands but the higher
elevation has precluded the eolian sedimentation and pronounced drainage basins. Composite Uplands
The Composite Uplands exist south from Colorado Springs to the border of the Front
Range Urban Corridor. This landform contains characteristics of all three of the
previous types.

2.2.4 Expansive Soil Distribution Along
The Front Range
There are 4 sedimentary bedrock formations that contain swelling clays that lie under
the majority of the Front Range area. These are the Pierre Shale, the Laramie
Formation, the Denver-Arapahoe Formation, and Dawson-Arkose Formation.
Figures 2.4 and 2.5 show the Front Range area and the soil distribution within that
zone. Pierre Shale
Hansen describes the Pierre Shale as Olive-gray to brown marine shale, siltstone, and
silty sandstone with bentonitic layers interbedded. Thickness of this formation varies
from 2,600 meters in the north to 1,100 meters in the south. This material exists from
central Colorado Springs south through eastern Pueblo to Canon City. It is also found
on the north side from Roxborough Park to the west side of Green Mountain, from
Golden to Boulder west of Colorado Highway 93, and from Boulder northeast to
Longmont, Loveland, Fort Collins and Windsor. The varied layers produce from low
to very high swell potentials. (Friedman) Laramie Formation
The Laramie Formation consists of intermixed sandstone and claystone. The
sandstone is fine grained, white to light-tan, and tightly compacted. The claystone is

e m e to m
Figure 2.4 Location Map of Front Range Urban Corridor
(after Hart, 1974)

Figure 2.5 Geologic Map of the Front Range Urban Corridor
(after Hart, 1974)

dark gray and carbonaceous. Thin, lignitic coal beds exist in the lower sections. The
sandstone possesses low swell potential while the claystone may possess moderate to
very high swell potential. (Hart, 1974) This formation underlies most of the areas
east of Fort Collins, south through Greeley to Brighton and along the east side of The
Denver Metro area.
2.2A.3 Dawson-Arkose and Denver-Arapahoe Formations
These two formations exist primarily through the Denver Metro area. Where the two
exist together, the Dawson-Arkose lies on top of the Denver-Arapahoe. Hansen
describes the Dawson as white to yellowish-gray or brown arkosic conglomerate or
coarse sandstone, siltstone, and olive-brown to variegated claystone. The Denver
Formation consists of brown to olive claystone, siltstone, sandstone, and
conglomerates derived primarily from andesitic volcanic rock. This formation
contains montmorillonite and is largely responsible for swelling soil damage in the
Denver Metro Area.
2.3 Characteristics of Swelling Soils
The science behind the clay-water interaction is still not completely understood.
What is evident is that unstable clays of the smectite group as explained in section 2.1
of this paper have a strong affinity for water. The charge imbalance within the
molecules is satisfied by the addition of positive H ions from the water molecule and

the amount of surface area available for this ion exchange to take place is so high that
volume increases result. (Little) Snethen (1975) and Nelson (1992) explain that there
are two categories of properties that affect the degree of swelling of a certain soil:
intrinsic and environmental. Intrinsic properties are characteristics within the soil
structure itself that lead to swelling and environmental properties are external
2.3.1 Intrinsic Properties
The following sections detail those soil characteristics that influence swelling. Many
factors have been identified as having and affect on swell potential to differing
degrees. This section will focus on 4 primary factors: clay mineralogy, soil water
chemistry, soil suction, and soil density. Other factors that have been identified are
temperature, cementation, permeability, and soil fabric. (Friedman, 1993) Clay Mineralogy
As discussed earlier in section 2.1, clay minerals are composed of plate-like structures
with negatively charged surfaces. The negative charges are balanced by cations in the
surface water attaching to the surface of the plates by electrical charges.
The electrical interparticle force field is a function of both the negative
surface charges and the electrochemistry of the soil water. Van der Waals surface
forces and adsorptive forces between the clay crystals and water molecules also
influence the interparticle force field. The internal electrochemical force system must

be in equilibrium with the externally applied stresses and capillary tension in the soil
water. The capillary tension is often called matric suction. (Nelson, 1992)
The amount to which this action occurs depends on the specific clay mineral size.
Montmorillonites, vermiculites, and some mixed-layer clays are typically most
susceptible to swelling due to the extremely small mineral size. The amount of
surface area available for cation absorption increases with decreasing particle size.
(Friedman, 1993) Table 2.1, from Nelson, 1992, shows the typical size to surface
area relationships of three major clay minerals.
Table 2.1 (from Nelson, 1992)
Mineral Group Basal Spacing A0 Interlayer Bonding Specific Surface (m2/g)
Kaolinites 14.4 Strong Hydrogen Bonds 10-20
Elites 10 Strong Potassium Bonds 65-100
Montmorillonites 9.6 Very weak van der Waals bonds 700-840 Soil Water Chemistry
As important as the soil mineralogy is the chemical composition of the soil water.
Swelling of the soil will be less with higher concentrations of dissolved cations or
with cations with a higher valence. Mitchell states that the fewer particles required to
balance the negative charge the less volume of water will be absorbed. For example,
a magnesium cation with a valence +2 will be cause less swelling that a Na cation
with a valence of +1. (from Nelson, 1992)
21 Soil Suction
Soil suction is a measure of a soils affinity for water with respect to its relative level
of saturation. Suction is measure as two distinct types: osmotic and matric. Osmotic
suction is due to differing salt concentrations between the soil and the water and is
intependent of water content or surcharge on the soil. Osmotic suction accounts for a
small percentage of swelling soils and is not considered here. Matric suction is
defined as the difference between the pore water pressure in the soil voids and air
pressure in those same voids. (Nelson, 1992) In a partially saturated soil, the
interface between air pockets and soil water experiences an inbalance of force based
on what level of saturation the soil is in. When the air pressure is greater than the
water pressure being exerted by the soil water, suction occurs, drawing water into the
void. Plasticity and Density
Plasticity of a soil, commonly measured in terms of Atterberg limits, measures a clay
soils ability to remain intact under increasing water content. A highly plastic soil
will retain its structure after addition of high amounts of water. This is an indication
of a large surface area being available to retain water. This plasticity is directly
related to swell potential of an expansive clay.

Density of a soil is also an important factor for swell potential. Soils with a higher
dry density will typically exhibit higher swell and swell pressures due to the more
compact nature of the particles and, therefore, the higher degree of interaction during
saturation. (Friedman, 1993) The electrical forces between the particles are directly
related to the proximity and orientation of the particles to one another (Nelson, 1992).
Chen (1959) also has reported that the swelling of highly compacted soil can further
influence the swell potential of the soil by changing the soil fabric or interparticle
relationships (Friedman, 1993).
2.3.2 Environmental Conditions
An expansive clay will only shrink or swell if it is subjected to a varying level of
moisture. Therefore, the environmental conditions imposed on a clay deposit are just
as important when considering swell as the intrinsic properties of the soil itself.
Environmental factors are initial moisture content, moisture variations, and stress
conditions. These are further broken down in the following sections. Initial Moisture Condition
From the principles presented in the previous section logic says that an expansive soil
with a high affinity for water and a low natural water content will tend to draw more
water than one with a higher natural water content. The less water present in the soil
voids the less the water pressure within the void and the higher the matric suction.

Conversely, an expansive soil that is very highly saturated will be more susceptible to
shrinking than one with less moisture as the extra pore water is forced out under
loading or during dry periods. (Nelson, 1992) Moisture Variations
The variations in soil moisture are caused by several factors. The change in moisture
is primarily confined to a certain depth below the surface commonly referred to as the
active zone. The depth of this zone will be determined by the factors outlined below. Climate
Expansive soils are most active in semi-arid regions where evaporation exceeds
rainfall. Areas with distinct wet and dry seasons are most susceptible to damage as
the seasonal variation in moisture is highest in these areas. The depth of the active
zone varies with the specific soil and climate but has been reported to be between 5
and 7 feet in semi arid regions by Snethen, 1975 but up to 14 feet in developed
locations. (Sikh, 1994) The reason for this variance is due to the effect that manmade
structures have on soil profiles. Effects of Manmade Structures
Structures have a large effect on the swell potential of soils. Houses and buildings
create a differential in the degree of exposure to natural water variations. Surface

drainage around a structure can concentrate water infiltration to certain areas and
irrigation systems for landscaping can add additional water to the soil profile.
Additionally, the temperature differential from outside the structure to under the
structure can also create moisture migration under the structure. (Nelson, 1992)
Finally, excavation and backfill around foundations can remold the clay and alter the
soil structure, increasing expansiveness or permeability, allowing moisture migration
to greater depths. Permeability
Soils with higher permeability due to natural deposition are more susceptible to
swelling as the water is allowed easier access to soil voids. Compacted soils have
greater permeability at lower moisture contents and will experience a decrease in
permeability as the moisture content approaches optimum. This is due to the fact
that, at optimum moisture, the voids are small and paths for water migration are
reduced. (Friedman, 1993) Vegetation
Trees and shrubs will remove water from the soil in localized areas, causing the soil
to be differentially wetted. (Nelson, 1992) Further, Huzjak (1988) noted that when
this vegetation is removed and structures are built, the moisture that was being drawn

from the soil by the roots of the plants will accumulate below the surface and create
differential heave in those areas (from Friedman, 1993). Stress Conditions
Volume change in a soil is directly related to a change in stress state. Several specific
factors are listed below. Stress History
An overconsolidated soil will be more expansive than a normally consolidated soil at
the same void ratio due to the release in stress and bonding between the particles.
Laboratory testing by Mitchell (1976) and Kassiff and Baker (1971) has also shown
that repeated wetting and drying will tend to reduce swell potentials, although the
effect is minimized after a certain number of cycles (from Nelson, 1992). In situ conditions
The initial stress state of a soil will affect its behavior under loading and during
saturation. The initial effective stress relates to the degree of saturation and, in turn, a
soils tendency to attract water. In-situ stress must be evaluated to truly evaluate the
response of soil under future loading.
26 Soil Profile
The soil profile details the thickness and location of various soil types within a given
deposit. The location and orientation of swelling soils within a deposit will greatly
influence the behavior of that deposit as a whole. Holland and Lawrence (1980)
stated that the most expansive deposits are those with expansive soils extending from
the surface to a depth below the active zone. Conversely, those deposits with
swelling soils only at depths below the active zone are unlikely to experience any
movement, (from Nelson, 1992) In some cases, orientation of the profile, either
man-made or naturally can affect the performance as well. For example, the steeply
dipping bentonite beds west of Metro Denver are more susceptible to swelling due to
their orientation. The deposit is made up of alternating layers of expansive bentonite
and sedimentary sands. The deposit is inclined on edge allowing water from the
surface to access the bentonite at great depths by moving down through the highly
permeable sand. Loading Condition
The loading condition over an expansive soil will determine how much heave will be
experienced at the surface. An external load to a soil deposit will act to balance the
interparticle repulsive forces within the soil and reduce the net swell. (Nelson, 1992)
The concept of drilled pier foundations uses this principle to a degree by

concentrating structure loads over smaller areas to overcome the extremely high
forces generated by the soil.
2.4 Drilled Pier Research
Drilled pier use dates back to the middle of the 19th century after the introduction of
Portland cement (Friedman, 1993). In the Denver area drilled piers have been used to
support structures for 40 years or more. However, only in the last 10 to 15 years has
there been a lot of attention paid to how well standard design practice accommodates
for the effect of swelling soils on drilled pier foundations. In the 1970s, designers in
the Denver area designed drilled piers assuming an active zone depth of 5 to7 feet.
After years of research and forensic evidence it is apparent that this is inadequate.
The inadequacy of the former design methods related primarily to uplift resistance of
the drilled pier, an aspect of foundation design most applicable in areas of swelling
soils. Consummately, designers today regularly consider the active zone to be at least
10 feet deep in horizontally stratified deposits.
Most of the research done over the years on drilled pier design can be divided into
either an analysis of the ultimate load carrying capacity of the pier or an analysis of
the uplift resistance (or capacity) of the pier. The research will be presented in that
order in the following pages.

2.4.1 Load Carrying Capacity
A load imposed on a drilled pier will be transferred to the soil through some
combination of two distinct mechanisms: skin friction and end bearing. The amount
of the total load carried by each mechanism is dependent on several items including
geometry of the pier, the properties of the soil immediately adjacent the pier surface,
the magnitude of the load imposed, and the time elapsed from construction. (Burland,
The major difference between the two mechanisms is the amount of movement
required to mobilize full capacity. Researchers have published varying results about
the magnitude of movement for each but they are agreed that full skin friction
capacity will be mobilized with less movement than that required for full mobilization
of end bearing capacity. It is the skin friction analysis that relates to the action of a
swelling soil on a foundation and thus will be the focus of this overview. Results
from instrumented load test performed by different researchers are presented in tables
2.2 and 2.3.
Table 2.2 Movement Required to Mobilize Full Skin Friction Capacity
Va (6 mm) Whitaker and Cooke per Woodward et. al (1972)
0.4 (10 mm) Vesic per Woodward et. al (1972)
5% of pier diameter Whitaker and Cooke, Berezantezev per Bowles (1977)

Table 2.3 Movement Required to Mobilize Full End Bearing Capacity
10 % 25 % of pier diameter (cohesive soils)
Woodward et. al (1972)
10% 30% of pier diameter (cohesive soils)
Kulhawy (1984)
8 % 10 % of pier diameter (cohesionless soils)
Woodward et. al (1972) Skin Friction Capacity
Skin friction attributable to load carrying capacity should only be considered for
portions of the pier installed below the active zone. This is due to the fact that areas
of the pier in the active zone may be subjected to swelling soils acting to push the pier
upwards, counteracting any downward force. This concept will be explained further
in later sections. The difficulty in evaluating the skin friction capacity of pier is
determining what value of interface friction to use. This determination is different for
cohesionless or cohesive soils. Cohesionless Soils
In a cohesionless soil, the side shear capacity is expressed in terms of conventional
friction along the shaft. The load supported by friction along the skin is given by the
following equation:
In most normally consolidated soils,/can be expressed in terms of p, the coefficient
of friction between the pier material and surrounding soil, and a/ the normal force
Qskin ~ f Askin
(Chen, 1988)

acting on the pier surface. The normal force is expressed in terms of K, the lateral
pressure coefficient, and av\ the effective vertical stress. The above equation then
Qskin = K cv Jl Askin (Chen, 1988)
The frictional capacity of the pier is directly proportional to the horizontal force due
to overburden stress in the soil. Much research has been done to try and quantify the
value of the coefficient of friction, ji. Vesic (1967) states that there exists a formula
for determining this value that shows an increase to a constant value reached at a
depth equal to 10-20 foundation diameters (from Kulhawy, 1984). However,
Kulhawy (1984) states that this is not the case. He states that the value will increase
with depth until the full side shear resistance is mobilized. He gives approximate
relationships of this value to the internal friction angle of the soil depending on the
type of pier foundation. For concrete soil piers he states that the interface friction
angle is equal to the internal angle of friction of the soil (Kulhawy, 1984). This
research parallels a similar search for a frictional resistance value for piers in
cohesive soils, one that is complicated when one considers swell capacity of these
soils. Cohesive Soils
The research of drilled shafts in cohesive soils is most applicable to the problems
concerned with swelling soils on drilled pier foundations as the interaction between

the cohesive soil and the pier surface is similar in both uplift and axial load capacity
analyses. The side shear on a shaft in cohesive soils is typically expressed in terms of
adhesion, which is related to the cohesion of the clay around the pier. The adhesion
term is independent of overburden stress but is influenced by swell potential,
something that will be discussed in the uplift resistance evaluation portion of this
section. An expression for total skin resistance in cohesive soils is as follows:
Qskin = Askin a Cu Winterkom, 1975
Where Cu is the undrained shear strength of the soil and a is a reduction factor.
Research has shown the value of a to be between 0.30 and 0.60 but the parameters of
these tests vary widely. The following paragraphs present some of the research
pertaining to evaluation of an appropriate value for a for axial load capacity.
Woodward et al. (1972) provides a summary of the work of a number of researchers.
The reported values of a range from 0.30 to 0.62 for stiff clay, and include values of
0.64 for tests performed in massive shale and gravelly glacial till with cobbles. He
presents test results that show that the value of a increases as undrained shear
strength decreases. This information is presented in Figure 2.6.

_ Obstnrad orntgi odhaaton
Undraintd tlinr tlrtngth
Relationship between reduction factor a and undrained shear strength.
Figure 2.6 (after Woodward, 1972)

DAppolonia et al., in results presented by Winterkom and Fang (1975) state that a
varies from 0.30 to 0.60 for stiff clays. They also agree that the value for a will
decrease as undrained shear strength increases.
Vesic (1977) states that values have been documented between 0.30 (short piles) and
0.60 (long piles) with an average value of 0.45 for bored, cast-in-place piles in
London clay. He states that the data contains significant scatter and that conclusions
based on these results may be of limited value.
Jubenville and Hepworth (1981) report values of a between 0.33 and 0.52 for test
piers installed at Fort Carson, Colorado, near Colorado Springs. These piers were
installed in the Pierre Shale formation and relate most directly to the testing done
within this research.
Chen (1988) quotes Mohan and Chandra as stating that the frictional resistance of
bored concrete piles in medium to hard clays is about half the undisturbed shear
strength of the clay along the pile surface (a = 0.5)
Chen (1988) indicates that the current standard of practice in the Colorado Front
Range area is to assume the skin friction value for the hard clays and claystone

bedrock materials is equal to one-tenth of the end bearing assigned to the bedrock
based on the Denver rule of thumb procedures.
This one-tenth relationship is also discussed in the Jubenville and Hepworth (1981)
paper. They caution that the one-tenth rule should be used very carefully as it
ultimately relates back to data obtained from the standard penetration test. They
indicate that the relationship is likely valid for stiff to very stiff clays and shales with
strengths similar to very stiff clays, where the penetration test is applicable.
However, as the strength of the clay or shale increases beyond that point, penetration
test blow counts typically increase significantly, or reach refusal, and the usefulness
or validity of the one-tenth rule for determining side friction becomes questionable.
In a similar report, Burland et al. (1966) also related the adhesion to the undrained
shear strength by way of the a factor. However, they outline that this value will be
significantly less than 1.0 for two reasons. First, the adhesion between the pier and
soil will only be about 80% of the shear strength of the clay and second, the
immediate zone around the pier will be softened by infiltration of water during
concrete placement. They state that this zone may be 1 Vi to 3. This can lead to an
initial lag in development of skin friction capacity.

In expansive soils, the determination of potential side resistance from skin friction
becomes more complex due to the magnitude of the horizontal normal forces that can
be generated by the swelling soils. The adhesion developed may be different due to
the lateral expansion of the soil. In most cases, the lateral swelling pressures far
exceed the horizontal stress one would compute using the effective vertical stress and
the lateral pressure coefficient. The magnitude of these forces is critical in evaluating
the uplift resistance of a pier in expansive soils.
2.4.2 Uplift Resistance
The analysis of the uplift resistance of a drilled pier foundation involves two
mechanisms. These are uplift force and withholding force. The uplift force occurs
within the active zone where moisture variation occurs and withholding forces occur
below this depth where the soil is stable. A graphic representation of the forces on a
rigid pier is shown in Figure 2.7. Uplift Force
Several researchers have presented data on the uplifting action of a swelling soil on a
drilled pier. The accepted general equation shows that uplift force is proportional to
the swell pressure of the soil (horizontal), the depth of the active zone, the surface
area of the pier, and the frictional stresses acting between the pier and soil. The area
of the most research around this action relates to the skin friction that should be used

(L-I, )_______ < i DEPTH OF
Figure 2.7 Forces on a Rigid Pier
(after Nelson, 1994)


when relating swell pressure to uplift forces. Chen (1988) defines the skin friction as
where ai is a coefficient of uplift between the pier and the soil and as is the swelling
pressure in terms of effective stresses. ONeill (1988) expressed this same skin
friction as:
/u = a.2 gs tan (from Nelson, 1992)
where a.2 is a factor greater than 1 to account for soil structure and disturbance and <|)r
is the angle of internal friction of the soil for residual strength in terms of effective
stress, (from Nelson, 1992). Nelson (1992) states that the two equations are the same
if ai = 02 tan 4>r. Knowing that the residual angle of internal friction for clays is on
the order of 5 to 10 degrees and using data from ONeill showing 1.0 < GC2 < 1.3 from
laboratory tests, one can see that ctj would vary from about 0.9 to 0.25. This means
that, during design, approximately 9% to 25% of the swell pressure should be used to
calculate uplift force on the drilled pier.

Chen (1988), from laboratory experimentation, proposed the following equation for
uplift force on a drilled pier:
U = 27tr/u(D-d) Chen (1988)
Where U = Total uplift force
r = radius of the pier
D = depth of pier
d = depth of penetration into stable moisture zone
u = lateral swelling pressure
/= coefficient of uplift force (also termed a)
His results showed that the uplifting pressure exerted through the soil-pier interface
was approximately 15% of the vertical swelling swelling pressure. This compares
with the range of 9% to 25% as calculated by ONeill. Withholding Force
The uplift forces generated by the swelling soils in the active zone are counteracted
by withholding forces in the zone of stable moisture content. These withholding
forces are made up of the dead load of the pier and the structure above as well as the
skin friction between the pier and soil. Chen (1988) quantifies this force as follows:
Fw = 7tr2p + 27trsd
Where: 7t r2 = pier end area
p = unit dead load pressure
2 7t r = pier circumference
d = depth of pier in anchoring zone
s = unit skin friction between pier and soil

In this type of analysis, the unit skin friction is similar to those values found by
researchers for skin friction for load capacity analyses (Section as the impact
of swelling pressure does not apply.
Chen (1988) states that the actual frictional resistance can be affected by a number of
factors. These include the roughness of the pier shaft and the strength of the concrete
as well as the geometry of the shaft. A belled pier will generate more withholding
force due to its larger surface area at the base. Uplift tests on modeled piers
(Lutenegger, 1994) show that failure is through perimeter shear. No large shear cone
appears to develop. This is supported by results from Burland et al. (1966) that show
that the zone immediately around the pier/soil interface is weakened during
excavation and concrete placement creating a failure zone that parallels the surface of
the pier.

3. Materials and Methods
3.1 Experimental Design and Approach
3.1.1 Purpose
The purpose of this research is to develop a method of measuring uplift force from a
swelling soil on a drilled pier foundation in an actual installation. The complexities
of this are several. First, the cost of doing a full scale installation is very high.
Second, instrumentation (not to mention researchers) are fragile with respect to the
rigors of the construction industry, protection of the components during installation is
difficult. Additionally, duplicating a real scenario in the field is very difficult.
Before we examine how this research deals with these complexities, we shall examine
the basic theory behind the proposal.
3.1.2 Proposed Methods
The plan was to install strain gauges at a certain interval through the active zone
within the pier to measure uplift force, through strain relationships, in the pier. The
distribution of the gauges allowed localized uplift measurement and the development
of a correlation between the location of strain experienced and the actual depth of the
active zone. (This was an additional bonus, even though its benefit can technically

only be realized in design of structures in the same or similar geological conditions.)
To ensure representative results, the gauges would have to be installed in manner that
would cause them to experience the same force inside the pier that the soil was
imposing on the outside surface of the pier. When this was achieved, the only
information required in determining uplift force was the elasticity of the composite
section of the drilled pier, accounting for any reinforcing steel installed during
construction. Additionally, however, measurement of the variation of the soil
moisture was critical in correlating the observed movement with increases or
decreases in moisture content. This parameter was measured at two points
throughout the process: prior to construction and after structural load had been
imposed. This was done due to the lack of availability of an economical method to
accurately measure soil moisture during monitoring. The scope of this research,
unfortunately, was limited at times by the size of the budget.
The decision was made, for reasons to be described later, to use a steel pipe installed
in the center of the drilled pier equipped with strain gauges. This method created the
need to introduce a new element into the elasticity of the drilled pier. The
calculations for this section are found in the appendix.

3.1.3 Testing Apparatus Design
This research is possible due to a connection between academia and industry. The
primary researcher, through his contacts with a construction firm, was able to
integrate the installation of the instrumentation into a project being built for a third
party owner in an area of swelling soils. The proposal, therefore, was only limited by
one complexity: the protection of the gauges. It is this parameter, therefore, that
drove many of the decisions made on the method of implementation herein. Instrumentation Program
The first problem with installation of anything in concrete is the concrete itself. The
hydration and curing process of concrete can destroy gauges and wiring due to the
chemical reactions taking place. Additionally, the placing of the concrete, especially
in a drilled pier where the product is dropped from above, can turn many hours of
installation and calibration into wasted effort with one impact from a piece of
aggregate. These concerns led us to the idea of using some type of pipe where the
wiring and instrumentation could be installed inside, away from the hazards of the
concrete altogether.
The first plan was to use couplings, approximately 4 or 6 inches long, and install
gauges on the inside of these and then assemble them to longer sections, eventually
into a piece of pipe as long as the pier. This pipe would be installed in the center of

the concrete pier prior to concrete placement and be affixed with some sort of
anchoring device to ensure the pipe would compress or extend with the concrete. The
pipe would need to be anchored in two ways: continuously through the active zone
and somewhere inside the stable zone. This would ensure that localized movement
within the active zone could be measured by the closest gauge and that the pipe
would be sure to experience the same withholding force that the soil-concrete
interface below the active zone provided the pier itself. The installation of the gauges
inside the couplings was desirable due to the fact that everything would be isolated
from the concrete. The properties of the relatively thin-walled pipe would also ensure
that strains experienced on the inside surface of the pipe could be assumed to be the
same as those being imposed outside.
A problem was encountered with this method. The standard product available in a
steel pipe consisted of couplings that were threaded continuously on the inside
surface. This prevented an acceptable bond between the gauge and the steel. Again
the budget helped to make the decision that the process would use only standard parts
and avoid the cost of custom fabrication. Several additional options were considered,
including installing the gauges inside 2-foot long sections of pipe and threading these
together with shorter couplings. This was easier due to the fact that the standard pipe
was threaded on the outside to accept the inside-threaded couplings. This method
created a concern about the translation of the force. The gauges would be installed

inside the end of the pipe, in an area that would eventually be overlapped by the
coupling on the outside. This introduced a rather complex analysis of how axial force
in the pipe would be transferred through the threads in the coupling and how much of
that force would be seen by the pipe itself at that location.
This concern led to an idea to use welded construction, eliminating the concern with
the threads. The stress concentrations developed around the weld could be avoided
by using full penetration welds. This idea was complicated by the temperature
tolerance of the installed gauges as well as the skill of the researcher (or lack thereof)
at welding steel pipe. This apparatus would be tough to assemble in the field, which
was the only practical place to assemble a 20-foot long piece of pipe.
All these concerns led to the examination of another method entirely. Research with
MicroMeasurements, the manufacturer of the gauges and associated accessories,
turned up a product called M-coat F that was capable of protecting installed gauges in
extreme atmospheres. After having been assured by the product representative that
this would include concrete, the obvious solution was pursued.

3.2 Selection of Gauges
3.2.1 Introduction
The final configuration of the instrumentation consisted of gauges installed on the
outside of 2-foot long sections of pipe joined by threaded couplings to form a section
long enough to stretch the length of the drilled pier installed. The threaded couplings
allowed easy assembly on site and the 2-foot sections were easy enough to handle in
the laboratory for gauge installation. The gauges were installed in the center of the
length of the pipe to remove them from any stress concentrations occurring at the
couplings. To further protect the installation, holes were drilled in the pipe a short
distance away from the gauge, again to avoid stress concentrations, and the wiring
inserted into the pipe and run through here to the surface. This, along with some
temporary protection of the short length of wire exposed between the gauge and the
hole, eliminated the concern of the concrete placing operation tearing away the
wiring. The couplings installed between each section of pipe, due to the difference in
diameter, were assumed to provide enough of an embedment feature to cause the pipe
to act compositely with the concrete. A diagram of the final setup is shown in Figure
3.1. The configuration of gauge installation as well as pipe selection is outlined here
During gauge selection, the MicroMeasurements representative as well as the
company provided information provided much assistance. According to their

Figure 3.1 Testing Apparatus

literature, consideration of 5 factors is necessary in selection of the appropriate gauge
for any application. These are: environment in which installed, durability of the
gauge, constructibility of the entire application, material to which the gauge is to be
bonded, and the resistance of the gauge.
3.2.2 Environmental Considerations
The environment parameter relates to what kind of exposure the gauge will see during
its monitoring life. This includes moisture content as well as chemical
concentrations. According to advice from MicroMeasurements, embedment in
concrete was to be considered an extreme environment, due to the high presence of
water initially and the chemical reaction taking place as the concrete hydrates. This
led the decision toward a fully encapsulated gauge, that is a gauge where the grid
pattern is fully protected by polyimide.
3.2.3 Durability
Durability was obviously important after definition of the environment as extreme.
An additional concern was the curved surface of the pipe where the gauge was to be
installed. This led to the use of a gauge with a flexible backing material that would
conform to the surface of the pipe and be more durable during installation.

3.2.4 Constructibility
The question of constructibility actually did the most to narrow the field. Two
additional types of gauges, both with less required preparation prior to field
installation were originally considered. Weldable gauges are commonly used when
installing on steel members, commonly rebar; and direct embedment gauges are also
used quite commonly in concrete applications. For both, the constructibility of the
entire setup was difficult. With the weldable gauges the installation of the gauge,
leadwires and associated protection would all have to be done in the field as that was
where the reinforcing steel was being constructed (the 20-foot lengths of steel being
used made it impractical to transport the reinforcing bars from the laboratory to the
field). The direct embedment gauges were much easier in that the leadwires were
pre-attached and protection was not an issue as the gauges were completely enclosed.
The problem with this method was the inability to install the gauges at a precise depth
and orientation at depth inside a 24 diameter hole. Both of these methods were
discarded for the use of conventionally bonded strain gauges due to constructibility
3.2.5 Test Material
The material being bonded to was schedule 40 steel. This made selection easy as
most of the gauge series were capable of this with some type of adhesive.

3.2.6 Gauge Resistance
The final parameter was resistance of the gauge. The higher the resistance, the more
sensitive the gauge would be with minimal voltage input. This is important in
locations where the strain is small and/or the heat generation from higher voltage
inputs can be a problem. In this case, the strain was predicted to be about 0.1%, not
very large, but the accuracy with which this would need to be measured was not high.
The inconsistency of the soil type encountered in the original site investigation
suggested that the results would indicate general trends, at best. Additionally, the
installation within the concrete shaft would provide a large heat sink for dissipation of
the extra heat created with higher voltage input. Due to the higher cost of the higher
resistance gauges and this preceding information, the decision was made to use a low
resistance gauge.
3.2.7 Gauge Parameters
Once the general parameters were set, the use of MicroMeasurements Tech Note 505
drove some more specific decisions about the type of gauge. Specifically these
parameters were backing material, grid foil alloy, gauge length, gauge pattern, and
self-temperature compensating (STC) number.
50 Backing Material and Grid Foil Alloy
The backing material and the grid foil alloy were considered together, as
MicoMeasurements standard series designations consist of certain, predetermined
combinations based on performance. The alloy is the material that will actually be
calibrated to determine the strain in the material being measured. For this application
the most common type of alloy, constantan, was selected due to its high strain
sensitivity and its ability to be delivered in various STC numbers, (this parameter
will be discussed shortly) Additionally, the commonality of this type of alloy made it
available in many different options at a reasonable price. The backing material
selected was determined by the previously discussed need for a fully encapsulated
gauge. These two parameters led to selection of one of the CEA series gauges, the
most common type for general purpose static and dynamic strain measurement. Gauge Length
The gauge length is dependent on the amount of space available. The size of the pipe
allowed a common size of VS to be installed. This made the constructibility easier. Gauge Pattern
The specific gauge pattern was selected based on the desired measurement.
Specifically what was to be measured was the axial strain in the longitudinal
direction. As the devices were to be installed on a round surface, ensuring true

longitudinal alignment would be difficult. Additionally, installing the pipe in the
drilled pier in perfect alignment with the shaft would be very difficult, if not
impossible. Due to these conditions and the desire to obtain as close to a true
longitudinal strain measurement a decision was made to incorporate a Poisson gauge
configuration. The Poisson configuration consists of a pair of gauges, one oriented
90 degrees from the other. When wired together in a half or full bridge arrangement
the two gauges will average the stresses to produce an output equal to one of the
primary stresses. In this case, a MicroMeasurements Poisson gauge was useful here.
This is a double gauge installed on one carrier backing, each oriented 90 degrees
apart. Due to the low price of this gauge relative to buying two separate gauges, this
project was designed to use two gauges per installation location. This resulted in a
full bridge installation. Self Temperature Compensation
Self temperature compensation is the ability of an installed gauge to correct for
temperature effects on the strain behavior of the test subject material. Different
materials respond differently when subjected to changes in temperature.
MicroMeasurements manufactures gauges that are pre-calibrated to compensate for
these changes based on the type of material. This selection was quite simple as the
pipe to be used was standard schedule 40 steel pipe. The STC number for carbon
steel is 6 so the type of gauge was selected accordingly.

3.3 Analysis of Pipe Capacity
3.3.1 Preliminary calculations
To determine the size of pipe required some preliminary calculations were done using
the parameters given by the soils report to determine what potential total uplift could
be experienced. The equation was as follows:
U = 2 7i r a u (D d) (Chen, 1988)
Where: U = uplift force
r = radius of pier
a = coefficient of uplift between the pier and soil
u = swell pressure, from soils report
(D-d) = depth of active zone
A value of a equal to 0.15 was used (Chen, 1988). This value was also within the
range found by ONeill (1988) who reported finding values of a between 0.10 and
0.25, with the clayey soils having values nearer the lower end of the range due to the
internal shearing between particles. A swell pressure of 12 ksf was found from
interpreting the oedometer tests reported in the soils report (see section 4.1.1) from
borings near the pier to be instrumented. The piers to be instrumented were 18 in
diameter or 9 radius. Finally, the depth of the active zone was assumed to be 10
feet. This was an assumption made from the beginning of this experiment and was
designed to be checked later after results were obtained.

3.3.2 Analysis of pipe capacity
The calculation for soil-induced uplift force was made as follows:
U= 2 k (.75) (0.15) (12,000) (20-10) = 84,823 lbs
The schedule 40 pipe to be used for this project had a yield strength of 36,000 psi so
the objective was to find a cross section large enough to prevent exceeding this value
under the calculated maximum load. Accounting for the contribution of the area of
concrete the formula was as follows:
Ftotal= (^allowable A)pjpe + f t ACOncrete
Through trial and error, a standard pipe size of 2 W\ with a cross sectional area of
1.72 in2, encased in an 18 diameter concrete pier produced a total capacity of
136,920 pounds. Of this, the concrete pier contributes 75,000 pounds and the pipe
contributes 61,920 pounds. This analysis also assumed 3000 psi concrete, and
industry standard, and a tensile strength of concrete equal to 1/10th the compressive
strength per standard structural theory. Now that the materials to be used were
selected, the preparation was begun.

3.4 Laboratory Installation and Calibration
3.4.1 Introduction
The benefit of using a reputable company such as MicroMeasurements was their
plethora of technical documentation and advice for gauge installation. These
procedures were followed religiously to ensure proper performance of the gauges
under extreme conditions for an extended period of time. As mentioned previously,
the pipe was acquired in 2 foot segments to make handling in the laboratory easier.
The gauges, adhesives, preparatory solutions, and sealants were all acquired from
MicroMeasurements directly. The University of Colorado at Denver provided the
tools necessary as well as the laboratory facilities. The complete installation took
approximately 18 hours, spread over two weekends. The specifics are detailed here.
3.4.2 Preparation
Each section of pipe was prepped by marking two longitudinal lines spaced equally
around the diameter to act as guide marks for the centerline of the gauge installations.
A hole was drilled 90 degrees from this plane for inserting the wiring into the pipe for
conveying it to the surface and to remove it from risk of damage from concrete
placing operations. The hole was drilled 90 degrees away to prevent creating a stress
concentration in the line of either gauge setup. Drilling the holes first prevented
potential damage to a gauge from slipping of the drill bit on the rounded surface.

After each pipe was marked and drilled the surface preparation began. This was
extensive due to the extremely rough surface by nature of schedule 40 steel pipe. The
gauges would need an extremely smooth surface to adhere to due to the small
dimensions of the gauge components. To achieve this the areas where the gauges
would attach were first ground smooth with a drum sander attachment on a hand held
drill until the surface protectant was removed and the resulting surface was smooth to
the touch. After this the localized surface was hand sanded with increasingly finer
grains of sanding paper until a fine grit was used with a minor amount of neutralizing
agent for lubrication. After a fine grit sanding paper could be used without picking
up residue from the surface the prep was completed. This procedure was repeated for
each of the 10 lengths of pipe, 2 areas per pipe. Incidentally, this procedure left the
section of pipe exposed to potential corrosion so the area being prepped was
maintained large enough to ensure proper adhesion of the entire backing of each
gauge but small enough to be sealed by the final protective covering.
3.4.3 Gauge Installation
The installation of the gauges was the next step. The adhesive being used was a
product called AE 15, a high strength adhesive designed for medium range strain and
extreme durability conditions. This adhesive consisted of a two-part solution, once
mixed the resulting mixture had a shelf life of approximately 2 hours so as much
preparatory work as possible would be done on each piece prior to mixing the

adhesive. (One package of adhesive consisted of 5 pots for a total cost of
approximately $60 so wasting one mixed pot equaled $12, consequently the concern).
Prior to actual adhesion, the remaining steps were degreasing, conditioning, and
neutralizing the surface. All three steps were performed in sequence and all contact
with the surface was done with sterile gauze pads and cotton swabs, used only once.
Once the surface was prepped the adhesive was conservatively applied and the gauge
installed using a piece of cellophane tape as a temporary carrier to prevent handing of
the gauge directly with the hand. After adhesion, a rubber gasket was installed over
each pair of gauges and a pipe clamp installed around the entire installation and
tightened. The rubber gasket was used to distribute the force from the pipe clamp
evenly across the entire gauge. Once all the gauges were installed accordingly, they
were placed in an oven and left overnight at approximately 100 degrees Centigrade to
accelerate and improve the bonding process.
Once the adhesive had cured overnight the clamps and gaskets were removed and the
cellophane tape removed very carefully to prevent tearing off the gauge. Out of 20
gauges installed, 6 failed to adhere and needed to be re-prepped and new gauges
installed in the same manner as before. Of these, four in particular failed to adhere
again and this resulted in one section of pipe being installed without gauges due to
time and material constraints. (Once a gauge was installed and failed to adhere, reuse
was not possible).

3.4.4 Wiring Configuration and Installation
As mentioned previously, the gauge installation was selected to develop a full
Wheatstone bridge configuration. This resulted in a total of eight wires being run to
each gauge set location. The final configuration and wiring schematic is shown in
Figure 3.2. All MicroMeasurements guidelines were followed for soldering of
connections. Wiring from each side of the pipe was inserted through the hole drilled
in the pipe and spliced together into one run of four wires that would ultimately lead
to the surface were connection would be made to the testing device.
3.4.5 Weather Protection
Due to the expected exposure of this gauge installation, considerable effort was
invested in the protection of the both the gauges and the wiring. A
MicroMeasurements product called M-Coat F was used to protect the gauges
themselves after all the wiring was installed and tested for continuity. This system
consisted of a layer of liquid rubber mastic (M Coat B) applied to the wiring, Teflon
tape applied over the electrical surfaces, a butyl rubber sealing patch, a neoprene
rubber pad, and a layer of aluminum tape sealed at the edges again with liquid rubber
mastic (M Coat B). The installation process was lengthy but considered a necessity
to prevent damage during pier installation. The entire outer surface was protected
with a layer of duct tape to prevent any wiring being pulled loose or damaged during

Figure 3.2 Wiring Schematic

transportation to the site and installation in the pier. The wire splices done to
complete the full bridge circuit were all contained within the pipe and were protected
by installing shrink wrap tubing over the spliced section and sealing the ends with M
Coat B. Once the pipe was fully assembled, the top was closed with duct tape as a
temporary measure to prevent wet concrete from being deposited inside the pipe.
Pictures of the completed installation just prior to installation in the pier are contained
in Figure 3.3.
3.4.6 Calibration
After all the wiring and protection was complete and prior to assembling the sections
of pipe, each 2 foot length was calibrated in the laboratory using a load cell. Each
pipe section was placed in the cell longitudinally and connected to the testing module
to be used for the entire project. The cell was used to exert an axial compressive
force on the pipe section in exactly 1000 pound increments. The testing module was
designed to convert the return voltage from the gauge set into a reading of
microstrain, after the operator entered some basic data about the configuration. This
included the gauge factor as well as the bridge configuration. These readings of
microstrain were unique for each gauge but, when plotted, the calibration curves
showed identical slopes within a few percent. These curves can be see in Figure 3.4.
The true strain in the pipe was determined with the following equation:
e = F / E Area (Nilson, 1991)

Figure 3.3 Photos of Testing Apparatus prior to Installation

The area used for this calculation was the area of the pipe cross section. The true
strain is plotted to determine the multiplier for each gauge to convert measured strain
to true strain. This information is detailed in Chapter Four of this report. Half Bridge Calibration Readings
Two of the gauge sets installed were done so in half bridge configurations due to
complications in the installation. These two sets produced calibration curves slightly
different from the full bridge installations. Gauge set 1A is shown with a slope
roughly half that of the full bridge configurations. The multiplier for this gauge set
was 1.75. Gauge set IB was similar to the full bridge readings but slightly lower.
The multiplier for this gauge set was 2.50. Full Bridge Calibration Readings
The remainder of the gauge sets were all installed as full bridge configurations. Only
four of the 7 gauge sets are plotted for clarity as the results were nearly identical. The
multiplier used for these gauges was 2.77.

Figure 3.4 Calibration Curves
Gauge 1A (1/2 bridge)
--Gauge 3A (full bridge)
Gauge 5A (full bridge)
Gauge 1B (1/2 bridge)
Gauge 3B (full bridge)
-Gauge 5B (full bridge)
4True strain
1000 2000
4000 5000
Load (lbs)

3.5 Field Conditions and Installation
3.5.1 Site overview
The site where these two piers would be located was located in Northern Douglas
County in the Inverness Business Park. A detailed soils report was prepared for this
property by CTL/Thompson Geotechnical Engineers. Detailed information of the site
conditions was presented previously in section 1.3 of this report.
The loaded pier was located on the outside building line to simulate worst case
conditions with respect to natural and irrigation moisture migrating down from the
surface. The unloaded pier was located approximately 9 feet away under the building
footprint. The first floor of the structure was to be an elevated slab so access to the
lead wires would be possible throughout construction.
3.5.2 Pier Installation
The piers were installed on March 15th, 1999. The steel pipe was transported to the
site in two pieces and threaded together on site. Prior to installation in the ground a
hole was drilled through the pipe approximately 1 foot from the bottom of the un-
instrumented section. A steel rod was inserted through the hole and secured in place.
This was done to prevent accidental slippage of the pipe and ensure that the pipe and

concrete pier would act compositely. Pictures of this installation are shown in figure
3.5. After the piers were drilled and the reinforcing steel cage set, the pipe was
lowered into the hole with a forklift and held in place with a brace while the concrete
was placed. The installation of both pipes was completed in less that one hour. The
top of the pipe was left approximately 6 inches above the top of concrete and the
wiring securely taped and concealed while the concrete cured. No water was
encountered during drilling and installation of either pier.
3.6 Field Measurement
3.6.1 Introduction
As was mentioned earlier, the first floor of the structure was elevated so access to the
lead wires from the piers was accessible throughout construction. The installation
was designed to last approximately 6 months so that enough time could pass to allow
wetting to occur to significant depth. The measurements taken throughout this time
period would, hopefully, show approximately what depth that was. As the
installation was done in March, these piers saw the end of the wet winter weather as
well as the seasonal monsoon season. This was beneficial in simulating a worst
case uplift situation. In addition, as this was new construction, the piers saw the
worst case as some of the uplift likely occurred prior to the structure being fully

Figure 3.5 Photos of Testing Apparatus Installation

Figure 3.5 Photos of Installation (continued)
ftt'V ***'

3.6.2 Strain Measurement
The days after the piers were installed and the concrete had cured overnight readings
were taken on each gauge set. This was done most importantly to record a relative
datum of microstrain for each gauge set and it also demonstrated how the gauges had
fared through the most demanding portion of the installation. All gauges reported
fully functional. For approximately 6 months, readings were taken regularly: about
every week to 10 days for the first 3 months and about every 3 to 4 weeks thereafter.
The regularity was decreased later due to several factors: 1) the structure was more
than 90% loaded, 2) the gauge readings had begun to level off, and 3) the change in
moisture in the environment was low through this period so no major movement was
expected. Information from each gauge set was recorded for each date with some
brief information detailing weather patterns between the last reading and the current
sample. Tabulations of results for each pier are presented in Section 4 of this report.
3.6.3 Soil Moisture Measurement
Prior to site construction the geotechnical report conducted by CTL/Thompson
reported soil moisture values at varying depths in the immediate area where the piers
were ultimately installed. As soil expansion is directly related to the change in
moisture, some knowledge about this variation was necessary to fully interpret the
results being measured from the strain gauges. Unfortunately, a moisture cell that
could be installed with the pier to measure this on a regular basis was cost prohibitive

for this project. The only reliable method known was a system used regularly by
agricultural researchers costing $2000 or more to set up. Additionally, having the
installation in a 3 foot-tall crawlspace eliminates the possibility of excavating or
drilling directly down on a regular basis to check moisture. As a compromise, prior
to the hardscape being placed around the building and after the structure had been
erected for about 6 weeks, a drill rig was brought in to drill out samples from outside
the building. Using a continuous flight auger and a little geometry, the rig was set up
to drill at a 45 degree angle from outside the building to depths of approximately 2, 5
V2, and 7 feet below the surface at or near the pier installation location under the
building. Samples were removed from the augers upon and moisture content tests
done in the laboratory. This enabled the author to measure the moisture change from
before construction to a point roughly 6 months later. The moisture content values
before and after construction are shown in Figure 3.6. While not a detailed record of
moisture variation, the two points did provide enough information to check some

Figure 3.6 Moisture Variation Curves
Moisture, w%

4. Analysis of Results
4.1 General Predictions and Background
4.1.1 In-situ Conditions
For this project, two 18 diameter piers were instrumented under a three-story office
building in Northern Douglas County, Colorado. CTL/Thompson did a geotechnical
observation and survey of the specific construction in November of 1998 and results
of that report were used to predict the results of this report. The material recovered
from the borings showed that the site consisted of sandy clay to clay deposits ranging
in depth from 2 to 12 feet overlying weathered claystone and/or weathered sandstone.
The weathered material ranged in thickness from 0 to 5 feet and overlay hard to very
hard claystone bedrock. Moisture tests showed natural moisture contents ranging
from 13.7% to 29.0%. The material appeared to be roughly horizontally stratified
with some occurrences of water at depths ranging from 8 to 14 feet below the surface.
Swell tests were also conducted on the samples and showed the site consisted of
predominantly low to moderately swelling soils. Percent swell values ranged from
0% to 3 Vi % with one sample measuring 7%. The location showing the highest swell
was chosen for the installation of the instrumented piers due to this being the apparent
worst case scenario. The value for moderate swell was used in lieu of the very high

swell value because this data point seemed to be errant compared to the consistency
of the other values across the site. The two piers were located on the southwest
comer of the building near boring #TH-1. A boring log superimposed over the
footprint of the building is shown in Figure 4.1. The boring log for boring TH-1 is
shown in Figure 4.2. The summary of the laboratory test results are shown in Table
4.1. During construction, some specific site factors contributed to the expectation of
results beyond what is described in the soils report. Those factors are described in the
next paragraph.
4.1.2 Site Conditions During Construction
This project was conducted from March through September of 1999 and the moisture
variation over that period was documented in the research. Over the first 2 months
following the installation, the site received several accumulations of snow high in
moisture content. At this same time, the construction of the building above left the
area where the piers were installed in the shade and, hence, prevented this buildup of
moisture from drying out as quickly as would normally be experienced in an open-air
location. Subsequently, 3 months after installation, the ground surface around the
pier installations was still damp.
Additionally, starting two weeks after the installation and continuing for
approximately 7 weeks the structure above was being loaded to its ultimate deadload

Figure 4.1 Site Map with Boring Locations
(from CTL/Thompson geotechnical report)

XI Dummy Pi6R
X2L LoKbSb PlER_

Figure 4.2 Boring Log (from CTL/Thompson report)
TH-1 TH-2 TH-3
EL =5011.1 EL =5814.1 EL =5806.3
P 27/12
P 50/10

TH-1 .9 13.7 11B 7.2 1.000 38.000 CLAY. SANDY (CL)
TH-1 19 20.6 105 1.8 1.000 5.400 50 29 82 CLAYSTONE
TH-1 29 24.2 101 3.0 1.000 12.000 CLAYSTONE
TH-2 4 15.6 103 -0.1 1.000 47 31 68 CLAY. SANDY (CL)
TH-2 9 26.9 92 0.0 1.000 52 34 80 CLAY. SANDY (CH)
TH-2 19 26.7 93 1.7 1.000 4.300 70 44 98 CLAYSTONE
TH-3 14 33.4 B7 1.1 1.000 4,500 CLAYSTONE
TH-3 24 17.8 105 2.2 1.000 4.800 CLAYSTONE
TH-1 9 21.0 100 0.3 1.000 1.600 0.050 CLAYSTONE
TH-1 14 29.0 91 12.800 CLAYSTONE
TH-1 19 24.4 97 1.5 1.000 5.200 CLAYSTONE
TH-5 4 17.1 106 0.8 500 1.800 3.94 CLAY. SANDY (CL)
TH-5 9 27.5 93 1.3 1.100 4.500 4.24 WEATHERED CLAYSTONE
TH-5 14 24.3 99 2.1 1,800 25.000 4.14 CLAYSTONE
TH-5 19 23.0 104 0.2 2.400 3.300 4.11 CLAYSTONE
TH-5 24 18.9 108 0.7 3.000 7.400 4.17 CLAYSTONE
TH-fl 4 25.4 97 1.0 1.000 FILL. CLAYSTONE. SANDY
TH-6 14 23.1 103 0.2 1.000 59 36 97 CLAYSTONE
TH-6 24 22.3 101 1.8 1.000 4.600 61 39 98 CLAYSTONE
S-1 0 TO 10 15.7 48 26 48 SANDSTONE
S-2 0 T010 11.5 52 33 70 CLAYSTONE
S-2 9 25.7 97 1.8 200 5.700 CLAYSTONE
S-3 0TO10 13.7 48 31 62 CLAYS/C LAYSTONE
S-3 4 12.3 111 2.9 200 CLAY. SANDY (CL)
S-1 0 TO 10 14.0 58 41 80 CLAYS/C LAYSTONE
S-4 4 9.8 101 3.5 200 7.800 CLAY. SANDY (CL) '

Table 4.1 Summary of Laboratory Results

with steel, concrete decks, and precast concrete skin system. This loading served to
moderate any uplift forces in the loaded pier but not in the dummy pier. The progress
of the structure loading is plotted next to the gauge information from the loaded pier
in Figure 4.4.
After three months of this activity, a boring was done from outside the footprint of the
building to obtain soil moisture measurements to determine what effect the rain and
snowfall had on the moisture content of the soil. This information proved crucial in
evaluating the results against what was predicted from calculations.
4.1.3 Predictions
The high amount of moisture left by the snowfall and the shading of the structure
indicates that higher swells were to be seen near the surface. As the porosity of the
clay was very low, the moisture ponding on top would only be expected to permeate a
few feet below the surface in such a short period of time. The values of uplift could
be expected to be less than that calculated for the loaded pier and nearer that
calculated on the dummy pier. Using the equation developed by Chen (1988) for
uplift from a soil on a drilled pier foundation a value for total uplift was estimated as
84 kips. Details of this calculation are contained in section 3.3 of this report. This
value was used as a benchmark to measure the degree to which the soil performed to
its maximum swell potential. Ultimately, knowing the swell potential of the soil from

laboratory tests, the intent of this research was to use the results from the
instrumented piers to check the assumptions made for the uplift calculation to begin
with. This comparison would have to take into account the actual increase in
moisture content to determine how much saturation was achieved.
4.1.4 Transformed Sections
Before the detailed readings from either pier are evaluated the transformed section for
each needs to be explained. The calibration curves developed in the laboratory
testing (Section 3.4.6) are used here to relate strain in the pipe to the actual load being
imposed on the total pier. To do this, the elasticity of the pipe needs to be related to
the total elasticity of the pipe/pier configuration through a transformed section
calculation. The equation for this was as follows:
^transformed = (Arebar + Apjpg + Aconc/10) from Nilson, 1991
The area of the concrete in the pier is divided by 10 due to the fact that the elasticity
of concrete is one tenth that of steel. The above equation therefore yields an effective
area of steel for the entire section. The dummy pier was installed without reinforcing
steel and the loaded pier contained six number 6 bars. This gives an effective area of
steel for the dummy pier of 26.7 in2 and an effective area for the loaded pier of 29.1
in2. These effective areas are used to relate true strain in the pier, given by the
calibration curves in Figure 3.4 and the associated multipliers, to actual load exerted
by the soil on the pier. The assumption here was that the strain in the pipe measured

by the gauges was the same as the strain in the entire pier. Therefore the following
equation was used:
Force from Soil Bsteel Areatransformed Ctrue
The Force measured is shown on the right hand vertical axis of each pier gauge data
4.2 Dummy Pier
4.2.1 Introduction
The dummy pier was installed without any deformed steel reinforcing embedded in
the concrete and without any imposed load from above. The pier was denoted pier
B for data acquisition sake but is considered here first for a few reasons. This pier
effectively provided a baseline for which to compare the loaded pier. The only factor
to consider when evaluating the results was the uplift of the soil. As no calculations
needed to be made for the contribution of the reinforcing steel or the weight of the
structure, this pier constituted the simplest evaluation. However, as the results
showed, the effect of the relative elasticity of the steel pipe was an important,
although unanticipated, factor to consider as well. All this provided some
experience prior to evaluating the more complex results from the loaded pier.

4.2.2 Results
Figure 4.3 shows the results from the gauge readings on the dummy pier. The series
are labeled with numbers corresponding to the pier and their position on the pipe from
top to bottom. A gauge pair was located every 2 feet starting at 1 foot from the top of
the pipe. Therefore, the last gauge pair, 5B, was designed to be installed at a depth of
9 feet below the surface. This was designed to be near the assumed depth of the
active zone of 10 feet. As installed, the top of the pipe on the dummy pier was 5
inches above the top of the concrete so the actual locations are 5 shallower than
designed. This variation had negligible affect on the results herein. All values in the
plot are normalized to zero with respect to the stress state the day after installation,
after the concrete had cured.
The majority of the gauge data fit into consistent curves that seemed appropriate for
the result expected. However, one data point in particular on gauge set IB seemed an
impossible value. This series is plotted with a dotted line and the value in question is
marked. The gauge reading corresponded to an uplift force of 137.6 kips. As this
gauge installation was only 7 below the surface of the concrete it is impossible for
this short section of pipe to develop this much uplift. The results are analyzed
neglecting this data point.

Figure 4.3 Dummy Pier True Strain
Days Elapsed
Compression Force, kips Tension Shallow effects
Gauge set IB was nearest the surface and, as expected, shows a high tension level
over the first few readings. This is most likely due to the water ponding on the
surface and permeating to a small depth. Unfortunately, this gauge stopped
responding to measurements after2 months. This may have been due to a level of
strain higher than the design threshold of this device or too much moisture resulting
in an effective submersion that shorted the gauge. The peak reading constituted
roughly 88.4 fie of tension higher than the in-situ state at commencement of the
experiment. Using the calibration numbers in Figure 3.4 and the ratio developed from
the transformed section analysis this translates to roughly 68 kips of uplift force. This
peak occurred 1 week after a period of roughly 2 weeks of wet weather, indicating
that the moisture had penetrated to the shallow depth of the first gauge set. The
original calculation yielded a total uplift force of 84 kips, this result shows that the
assumptions for this calculation may be conservative. Overall Effects
On an unloaded pier in swelling soils one would expect to see readings of pure
tension, relative to the post-installation stress state. However, as the results in Figure
4.3 show, gauge sets 3 and 4 showed initial readings of compression. This is where
overall effects need to be considered. Gauge sets 3 and 4 show compression for the
first two and three readings, respectively. This could be due to forces being exerted

from the lower section of the pier. This appears to be possible, as the tensile readings
from gauge set five are higher than those from gauge sets 1 and 2, indicating that the
section of pier between these gauge sets is being placed in compression. To further
ground this theory, as the tensile force in gauge set 5 drops off after the second
reading, gauge set 3 actually moves upward into tension. Gauge set 4 does the same
shortly thereafter. At this point, with a few minor exceptions, all gauge readings
appear to climb at a similar rate. Slippage and/or Passive Failure
The overall trend of the gauge readings seems to be climbing slightly in tension, yet
there are several spikes in the graph that this researcher believes may highlight an
interesting phenomenon. At three times, April 18th, May 16th, and May 30th, the
readings of one or more gauge sets appears to drop suddenly. The time elapsed
between readings here is roughly 7 to 10 days so the dramatic change, especially at
depth, does not agree with common soil theory. Between April 6th and April 18th
gauge set 5 experiences a rapid decrease in tension. Between May 9 and May 16
gauge sets 1,3, and 5 all experience the same phenomenon. Finally, between May 23
and May 30 gauge set 3 again experiences a rapid decrease in tension. This may be
due to slippage between the pier and soil. This slippage may be explained by the fact
that, as a soil swells its volume increases and density decreases accordingly. This
decrease in density contributes to a reduction in strength of the soil. Additionally, the

water from the concrete during and immediately after placement may have reduced
the interface friction value in the layer of soil immediately adjacent to the pier
Another phenomenon that may be at work here, however, is passive failure within the
soil. As the uplift from the soil causes the pipe to exert upward pressure back on the
soil deposit, passive failure may be occurring. This failure would be exhibited as a
cone failure around the pier up to the surface. This phenomenon may be more
applicable nearest the top of the pier where the failure plane would be shorter. No
significant signs of ground movement were observed during testing.
It can be seen that the sections of pipe in question both reach a higher level of tension
before slipping the second time. This may be due to the shear strength of the clay
being mobilized during the first slippage. If this is true, then the ultimate uplift force
being exerted on the pier is not limited by the swell pressure of the soil but rather by
the interface friction angle, a. This action further complicates the analysis of uplift
on a drilled pier by highlighting how differences in the stratification of the soil and
the method of installation can change the interface friction angle at various points
along the shaft.
83 Moisture Content
Moisture content measurements were taken on June 23rd, approximately 3 months
after the start of the experiment. The majority of the snow and rainfall stopped in mid
May but June experienced the usual evening thunderstorms and occasional short,
although heavy, rainfalls. The moisture content was measured at depths of 2, 5 Vz,
and 7 feet below the surface. The drilling was done from outside the building at a 45
degree angle to obtain soil as close to the actual pier location as possible. The results
showed an increase in moisture content of 6.5% and 7.5% at 5 Vz and 7 feet
respectively when compared to pre-construction levels. No moisture information
from 2 feet was available from testing done prior to construction. The moisture
variation curves are shown in Figure 3.6.
The optimum moisture content for this soil was reported to be approximately 20% by
CTL from soil testing during construction. These values were from testing done near
the surface but this variance is negligible for this analysis due to the relative
consistency of the soils. This showed that the soil had been saturated beyond
optimum so a logical assumption can be made that it would have been approaching
the zero air voids curve. Additionally, this highlights that the soil at depth had a
natural moisture content well below optimum indicating a potentially high value of
matric suction in this zone. This seems to agree with the relatively high strain seen in
gauge set number 5 when compared to the other gauges.

4.2.3 Comparison to Predictions
As expected, the gauge set nearest the surface did experience the highest uplift force.
The high moisture content near the surface due to the standing water and snow
combined with the short distance to this section of the pier quite likely contributed to
this. At the end of the testing period, well after completion of the shell of the
structure, all the points appear to be converging near an average value of
approximately 70 pe. This corresponds roughly to 55.8 kips of uplift force from the
calibration curves and the transformed section.
One must be careful how this value is compared to the computed results, however.
The equation used here to calculate uplift force is intended to yield a total force on the
pier from swelling soils. As the gauges were installed every two feet the area of soil
affecting stress on that particular section of pipe would only be equal to 27tr times 2
feet, as opposed to the total active zone depth assumed of 10 feet. If one assumes that
the pier is moving upward slightly with the uplift force then the total force from uplift
would be the sum of the values from each individual gauge location. The maximum
value of uplift on the pier as calculated would therefore be 55.8 kips times 5 gauge
locations, or 277.5 kips. This is assuming that, if gauge set number 1 were still
reporting, those values would also trend toward 70 p£. The final reading from this
gauge set, as shown in Figure 4.3, is roughly 85 pe, so this is reasonable.

At this point, one must compare these results to the assumptions using the original
U = 27trccu(D-d) (Chen, 1988)
Where: U = uplift force
r = radius of pier = 0.75 feet
a = coefficient of uplift between the pier and soil = 0.15
u = swell pressure, from soils report =12 ksf
(D-d) = depth of active zone =10 feet
The only truly known value from above is r. The pier diameter was effectively
constant after drilling. From laboratory tests, Chen suggested the use of 0.15 for a
but other researchers have found varying results, typically showing higher values for
a. At this point, the assumption will be made that 0.15 is a reasonable value and the
corresponding value for swell pressure will be back-calculated. This can then be
compared to the swell pressure given by the soils analysis and differences evaluated. Iteration for swell pressure
The calculation for swell pressure is important due to the fact that the above equation
relates to uplift truly by using a value for horizontal swell pressure whereas typical
industry testing, this project included, provides a value for vertical swell pressure.
Using a value of 277 kips for uplift and maintaining the other assumptions, one finds
a horizontal swell pressure value equal to 39.2 ksf. The original value used, from the

soils report, was 12 ksf. Several researchers (Parcher and Liu, 1965), (Komomik and
Zeitlen, 1965), (Josh and Kattie, 1978) have found that horizontal swelling pressures
typically equal or exceed vertical swelling pressure so this result agrees with those
findings. In particular, Josh and Kattie found that horizontal swell pressures are
higher than vertical pressures for a finite period after initial wetting and then tend
toward a constant value nearly equal to that of the vertical pressure. The results from
this experiment, particularly the early peak and eventual lower constant value
measured for uplift, are also consistent with Josh and Katties findings.
However, another possibility should be considered here. The original value of u=12
ksf was the highest value of swell in the general range of results from the soils
testing. There was one value of 38 ksf that was judged to be an anomaly due to its
extremely high nature when compared to all the other results. This sample was taken
at 9 feet deep at the location of the test boring closest to the test piers. At this point, it
appears that this value of swell may be accurate as the testing apparatus only
measured swell to +/- 9 feet deep as well.
4.2.4 Conclusions
At this point it seems that the original assumption of a = 0.15 may be valid. It is,
however, possible that the average swell pressure across the pier was less than 38 ksf
and closer to the typical values found across the site. These values ranged in swell

percentage from 0 to 3.5% and from 0 to 25,000-psf swell pressure. If a minimum
swell pressure of 12,000 psf is used then one finds 0.15 < a < 0.48. While this result
is not out of the range found by other researchers (section 2.4.2), a narrower range is
desired. This information will be reconsidered after evaluation of the next pier.
4.3 Loaded Pier
4.3.1 Introduction
The loaded pier was installed approximately 9 feet away from the dummy pier and
on the perimeter of the building. This pier was part of the designed foundation
system for the building and, as such, contained reinforcing steel throughout its entire
length. It also, of course, held a steel column exerting a load from the structure above
into the pier. Two additional unknowns in the effect of reinforcing and the actual
load from the structure over time add to make this analysis more complex than the
last. Fortunately, however, the exact effect of the reinforcing steel can be neglected
and the column load can be predicted based on erection sequence within a few
percent. The assumption made here is that the fact that the reinforcing steel and steel
pipe are both constructed of A3 6 steel means that the strain shown in the pipe by the
gauges is the same as that in the reinforcing steel. In effect, this means that the
reinforcing steel and the steel pipe both serve to provide tensile capacity to the pier