RESISTANCE OF A DRILLED SHAFT FOOTING

TO OVERTURNING LOADS,
MODEL TESTS AND CORRELATION WITH THEORY

By
Don L. Ivey,
Kenneth J. Koch
and
Carl F. Raba, Jr.

Research Report Number 105-2

Design of Footings for Minor Service Structures

Research Study Number 2-5-67-105

Sponsored by

The Texas Highway Department

in cooperation with the
Department of Transportation, Federal Highway Administration,
Bureau of Public Roads

July, 1968

TEXAS TRANSPORTATION INSTITUTE

Texas A&M University
College Station, Texas
 ACKNOWLEDGEMENTS
This research was conducted under an interagency contract between the Texas
Transportation Institute and the Texas Highway Department. It was sponsored
jointly by the Texas Highway Department and the Bureau of Public Roads. Liason
was maintained through Mr. D. L. Hawkins and Mr. H. D. Butler, contact repre-
sentatives for the Texas Highway Department, and through Mr. Robert J. Prochaska
of the Bureau of Public Roads.
The theory presented in Research Report 105-l was programmed for the IBM
7094 for analysis of the footings in this study by Mr. Leon E. Travis, III, and Mr.
Kurt A. Schemher.
The opinions, findings, and conclusions expressed in this publication are those
of the authors and not necessarily those of the Bureau of Public Roads.
 SYNOPSIS
Reported in this paper are the results of twenty-eight model tests of drilled shaft
footings subjected to overturning horizontal loads. The models are geometrically
similar but reduced by a factor o.f six compared to the average size of footings used
for mino·r service structures in Texas. The soils investigated range from cohesion-
less sands, through soils with both cohesion and an angle o.f shear resistance, to clay§
with no angle of shear resistance when tested using the unconsolidated-undrained
quick triaxial compression test.
The results o.f the model footing tests are compared with the theory developed
in this study, 1 * It was found that the conventional methods of predicting ultimate
load were conservative by as much as 500% for the cohesionless sands and by as
little as 20% for the clays. The coefficients developed in the new theoretical treat-
ment are evaluated so that the ultimate loads on this type of .footing in any given
soil can be predicted.
*Refers to numbers in selected references.
 TABLE OF CONTENTS
Pages
INTRODUCTION _______________________________________________________ 7
EQUIPMENT AND INSTRUMENTATION _________________________________ 7
General
Loading System
Load Measurement
Rotation Measurement
PLACEMENT OF FOOTINGS AND SOIL CONDITIONS____________________ 8
Easterwood Clay
Trinity Clay
Laho·ratory Sandy Clay
Ottawa Sand
TESTING PROCEDURE _________________________________________________ lo

TESTING RESULTS------------------------------------ ----------------11

COMPARISON OF TEST RESULTS AND THEORY _________________________ l3
SUMMARY------------------------------------------------------------17
SELECTED REFERENCES _______________________________________________ l7
APPENDIX __:_ ___________________________ __: ____________________________ :___18
 NOTATION

c cohesion ( force/length2 ) • *
cp angle of shear resistance (degrees) .
y unit weight, in place o·r w~t (force/length3 ).
y1 modified. unit weight of soil in the direction of the applied load (force/
length3 ).
y2 modified unit weight of soil in the direction opposite the applied. load (force/
length3 ).
P horizontal load applied to footing at some distance H above ground (force).
H height of horizontal load, P, above ground (length).
D depth of footing (length).
d footing diameter (length).
a depth to point of footing rotation (length).
K0 coefficient of earth pressure at rest (dimensionless) .
K1 coefficient of passive earth pressure applied to unit weight term (dimension-
less).
K2 coefficient of passive earth pressure applied to cohesion term (dimensionless) .
K3 coefficient of active earth pressure (dimensionless) .
k unit weight coefficient (dimensionless) .
J1 coefficient of shear stress, vertical footing surfaces (dimensionless) .
J2 coefficient o·f shear stress, footing bottom (dimensionless) .
B earth pressure coefficient modifier (dimensionless) .
*The dimensions of the various symbols are given in parentheses after each definition.
 Introduction
The wide use of drilled shaft footings to support
service structures necessary for the functioning of a
highway system has focused attention on the highly
conservative design procedures presently in use. The
foundations of structures such as signbo·ards, strain
poles, and lighting poles should be designed using fac-
tors of safety consistent with the relative importance of
the particular structure.
As the next necessary step, after the development
of a theoretical treatment for this type of foundation,
a series of model tests was undertaken. The purpose
of these tests was to evaluate certain coefficients intro-
duced in the theoretical treatment and to establish the
degree of precision to be expected in predicting ultimate
loads with the new theoretical treatment. The footings
tested ranged in diameter from 2 to 4 in., and in depth
from 10 to 12 in. Height of load application ranged
from 0 to 42 in. The soils in which footings were tested
ranged from a dry sand with no cohesion to a clay with
no angle of shear resistance. The soils are referred to
in this way on the basis of the unconsolidated-undrained
quick triaxial compression test. This type of test was
used to determine the pertinent soil parameters because
it best simulated conditions of short term loading in the
field. It is recognized that the same soils tested in an-
other way would exhibit different properties. Eight
tests of footings in soils with both cohesion and . an
angle of shear resistance are reported.
This is the second of a series of papers to be writ-
ten concerned with the design of these footings. The
first, Research Report 105-1, reported in detail the
theoretical development of the load prediction equations
which are compared with test data in this paper.

Equipment and Instrumentation

General
In o·rder to conduct these ·tests, it was necessary to
develop systems capable of ( 1) applying a horizontal
force on the footing at a uniform displacement rate
and ( 2) measuring the load acting on the footing· at
known angles of rotation.
wide range of load developed, 3 transducers were used
f<Yr the various tests. The capacities o·f each transducer
were 0 to 50 pounds, 0 to 400 pounds, and 0 to 1000
pounds, respectively.
The transducers were c<Ynstructed of a metal bar
instrumented with a full bridge o.f foil strain gages.

Loading System
A mechanically driven loading machine normally
used for compression testing was modified to apply the
overturning force to the footings. Figure 1 illustrates
the loading system used for these tests. The loading
machine was run at a speed of 0.05 inch per minute but
the pulley system increased this speed to 0.20 inch per
minute at the footing.

Load Measurement
The load applied to the footing was measured by
means of a fo-rce transducer spliced into the cable ap-
proximately 2 feet from the footing. Because of the

CABLE

FORCE

PULLEY
SYSTEM

VERTICAL DRIVE SHAFT

DRIVE SYSTEM

Figure 1. Loading system. Figure 2. Point source of light.

PAGE SEVEN

Figure 3. Placement of ultrG!Violet lamps.
Placement of Footings and Soil Conditions
Easterwood Clay
A series of tests was conducted in the shallow sandy
clay located in the vicinity of Easterwood Airport, Col-
lege Station, Texas. This soil was chosen as a repre-
sentative of soils possessing both cohesion (c) and an
angle o·f shear resistance ( cf>) , as determined by the
unconsolidated-undrained quick triaxial test. After the
test site location was determined by means of a series
of auger borings, the area was leveled and prepared for
the installation of footings.
A detailed procedure was established in order to
install the footings as they would be placed in actual
practice. Special hand augers were designed and con-
structed to drill the footing holes. These augers were
aligned and guided during the drilled operation by a
wooden template placed on top of the ground. The
excavated hole was a right circular cylinder. A heavily
reinforced steel cage was lowered into each hole which
was subsequently filled with a cement mortar made from
Type III cement and 20-30 Ottawa sand. The steel cage
was vibrated with a portable vibrato·r to eliminate air
bubbles o-r voids on the foO<ting surface.
After the footing tests were completed, a total of
nine borings were made with continuous sampling to a
depth of 18 in. Undisturbed soil samples were obtained
in the test area by a truck-mounted rotary drill rig. The
specimens ( 2.8 in. in diameter) were sampled Weith thin-
walled Shelby tubes. After extrusion the samples were'-
PAGE EIGHT
The output voltages from these strain gages were ampli-
fied and recorded on a visicorder, providing a continu-
ous record of the load on the footing.
Rotation Measurement
The position of the footing at known loads was
measured by recording the rotation of a metal pipe that
was attached to the top of each footing. The pipe was
screwed onto a %-inch diameter threaded rod extending
from the top of the concrete footings. The cable from
the loading machine was connected to the pipe by means
of a clamp that permitted the height of pull to be varied
along the entire length of the pole.
Very small holes were drilled in the pipe on 5-inch
centers along a straight line on the upper three feet of
the pipe. These holes provided point sources of light
from two ultraviolet lamps mounted on a carriage that
had been lowered into the pipe as shown in Figures 2
and 3. As the footing was rotated, these light sources
developed lines of movement on light sensitive paper
mounted on a wooden panel adjacent to the pipe. The
lights were turned on and off at specific loads so that
the position of all ten traces could be related. A straight
line was extrapolated through each series of termination
points to find the position of the footing at that specific
load. Because of the use of the light sensitive paper, it
was necessary to perform all the tests in the dark. When
the light sources were turned off, an external triggering
device simultaneously marked the visicorder load record
to denote the load at which the rotations were measured.
wrapped in aluminum foil and placed in cartons for
transportation back to the University's soil laboratory.
Upon reaching the laboratory, the specimens were sealed
with a microvan wax and stored in a humid room fo-r
classification and strength tests.
Comprehensive classification and strength tests were
conducted on the soil samples. The classification tests
included hydrometer analysis and determination of the
Atterberg Indices; the procedures outlined by Lambe2
were followed in these tests. The soil was classified as
a CH material by the Unified Soil Classification System.
A series of unconfined compression and quick triaxial
tests 3 was performed to establish the soil strength param-
eters ( cGhesion and angle of shear resistance) under
simulated field conditions. These tests were performed.
on representative samples from zero to 6 in., 6 in. tG 12
in., and 12 in. to 18 in. depths.
Mohr-Coulomb failure envelopes were constructed
from unconfined compression and quick triaxial com-
pressiGn test results on selected specimens. The average
values determined for cohesion and angle of shear re-
sistance from these tests were 2810 psf and 9o respec-
tively. The average in place density was 133 pcf.
Trinity Clay
A series of tests was conducted in the laboratory
using a soil exhibiting no angle of shear resistance in
the quick triaxial tests. The clay was obtained in dry
powdered form in 25 lb. bags, and mixed with water in
a counter current mixer to obtain a water content of
18%. The soil was classified as a CL material by the
Unified Soil Classification System.
A test bin was constructed in the Civil Engineering
Department's Soil Mechanics Laborato·ry to facilitate
these tests. The bin, 17.5 in. tall and 27.5 in. wide, was
separated into 3 equal parts, each 48 in. long. The bin
had a metal liner surrounded by 6 inch~ of concrete
for rigidity. The clay was placed in 4, in. thick, loose
layers and- compacted with a pneumatic hammer as
shown in Figure 4. The bin was completely filled with
the compacted soil. A constant amount of compaction
effort was applied to each layer in order to develop
uniformity of unit weight and shear strength.
After the compaction of each bin was completed,
holes were drilled and footings were installed as de-
scribed in the Easterwood Clay section. This is shown
in Figure 5.
Figure 5. Footing steel reinforcement before placement
The entire bin was covered with a sheet of thin of mvrta.r.
plastic to form an airtight fit. This prevented the soil-
water mixture from losing moisture through evapora-
tion. A minimum of 14 days was allowed to· elapse tinctive stress-strain characteristics. 4 •5 •6 These soils will
befme the footings were tested to pro·vide sufficient undergo large deformations prior to mobilizing their
time for the mortar to· cure. The compacted soil also full strengths. However, if allowed to "age," they be-
gained strength through thixotropy 3 during this time. come stiffer and more closely approach the stress-strain
Thixotropy can best be described as a gain in strength characteristics of an undisturbed soil.
with time at a constant volume-water content condition. After the footings were tested, representative sam-
Remolded or compacted soils are known to possess dis- ples were ()btained thmughout the depth of the footings
in each bin. Miniature samples were obtained with a
thin-walled core cutter; these samples were 1.425 in. in
diameter and 3.0 in. long. The core cutter was pushed
by hand into the clay and then trimmed out with a knife.
The soil specimen was then extruded from the cutter,
wrapped in aluminum foil, and waxed to prevent mois-
ture bss. Quick triaxial compression tests were con-
ducted on these specimens. The results of these tests
indicated the follo·wing physical and engineering prop-
erties; cohesion (c) = 3790 psf, angle (}f shear resist-
ance ( cf>)= 0, and unit weight in place ( y) _:_ 126 pd.

Laboratory Sandy Clay

The sandy clay encountered in the field tests at
Easterwood Airport was only representative ()f the high-
er shear strength range. Because of this, it was con-
sidered necessary t(} conduct additional tests in c- cf>
soils with lower shear strengths in order that the spec-
trum might be better covered.
The Trinity clay that was described previously was
mixed with a locally available well-graded concrete sand
in different pmportions and at different water contents.
The constituents of these batches are shown in Table 1:

TABLE 1

In Place
Water Unit
Test %Sand %Clay Content Weight
Number by Weight by Weight (%) (pcf)

L-1 67 33 11.9 123

L-2 33 67 16.4 126
L-3 67 33 9.8 132
Figure 4. Compaction of clay m test bin.
PAGE NINE
The same procedures were followed as stated previously duce intermediate void ratios. By testing at these ex-
to compact the soil, to install and test the footings, and tremes, the effect of void ratios on the ultimate over-
to sample and test the sandy clay specimens. turning loads was indicated.
The results of unconsolidated-undrained, quick tri- In the preparation of these tests, the precast con-
axial compression tests are shown in Table 2: crete footings were suspended in place and the sand was
placed around them. To achieve the loosest void ratio
TABLE 2 condition (e = 0.62), a "raining" technique was used
to place the sand around the footing in the bin. The
Angle of Internal sand was placed in a bucket held approximately 3 in.
Batch Cohesion Friction above the surface and poured in place to a depth of 18
Number (degrees)
(psf)
inches. The densest condition ( e = 0.51) was obtained
L-1 749 3.5 by placing the sand in the bin in 6-in. lifts and using a
L-2 1152 5.0 combination of rodding and vibration t<Y densify it. A
L-3 1411 12.0 portable concrete vibratm was used to vibrate both the
sand and the bin.
Cylindrical specimens were prepared with dry Otta-
Ottawa Sand wa sand at each of the two test void ratios, 0.51 and
A series of c = 0 case tests was conducted in the 0.62, and tested using a consolidated-drained triaxial
laboratory using 20-30 dry Ottawa sand. This material compression test. The weight of sand was computed
was selected as the test media because of its commercial to fill a right circular cylinder 2.80 in. in diameter and
availability and wide use as a standard research soil. 6.0 in. tall so as to achieve a desired void ratio. Mohr-
The test apparatus consisted of a shallow round bin Coulomb failure envelopes were prepared from the re-
rigidly attached to the testing frame. A series of tests sults of triaxial compression tests on dry sand specimens
wa§ conducted to establish the range of void ratios that and the angles of shear resistance were determined. The
could be achieved (loose to highly compacted) by dif- sand specimens prepared at initial void ratios of 0.51
ferent methods of placing the sand in the bin. These and 0.62 possessed angles of shear resistance o.f 37"
extreme conditions were represented by void ratios of and 32°, respectively. The in situ unit weights were
0.51 and 0.62. It was found feasible to achieve either 109 pcf for the void ratio of 0.51 and 102 pcf for the
of these extremes but very difficult to consistently pro- void ratio of 0.62.

Testing Procedure
The testing procedure used on all tests is sum-
marized in the following paragraph.
The cable from the loading machine was secured
to the pipe at the specific height of pull that had been
predetermined fm the particular test. The slack in the
loading cable was taken up in the pulley system and
the loading frame was adjusted to insure a horizontal
pull. Care was taken to apply only a minimum load to
the pipe. The ultraviolet lamps were then turned on
briefly to mark the initial position of the footing prior
to the test. The loading machine and visicorder were
turned on and load was applied to the footing. The
application of load to the footing- was recorded on the
visicorder and, as the pole rotated, the lamps inside the
pipe were turned on at intervals to mark its position.
This procedure was continued throughout the test. Most
tests were terminated after about 20 degrees of rotation.
The intersection of consecutive position lines 11;ave an
estimate of the footing ro•tation point as shown in
Figure 6. · ·

INITIAL FOOTING POSITION

FINAL FOOTING ROTATION POINT

POSITION

Figure 6. Displacement traces.

PAGE TEN
 ·Test Results
The results of the model footing are presented in 40 40

Table 3. Both the maximum load resisted and the load

resisted after a rotation of 5o are given. In general, the 5 5
maximum load occurred at a ro,tation slightly greater
than 5o, but there were a few tests in the Ottawa sand
30
where the maximum load occurred slightly before the
5o rotation was achieved. /r·
30
/-""
Graphs of load and rotation point versus rotation
(angle of deflection) are given for each test in the
25

I """ '· I'·.._

·- -'-·-
25
i \"
"·------ -~

appendix. Typical test graphs for the footing in dif- 0

«
20 0
«
20
0 0
ferent materials are shown in Figures 7 through 10. .J .J

In the tests in sands and clays, Figures 7 and 8, a well 15 15

defined peak load is reached. The footings in sand
continue to support about 70% of the peak load up to 10 10
rotations of about 20o while the footings in the clay
with no angle of shear resistance lose l()ad pmgressively
after the peak is reached. The footings tested in soils 5

with both cohesi()n and an angle of shear resistance had

no well defined point of maximum resistance up to rota- 0 TEST NO. 3 0 TEST NO. 4
tions of about 20°, but were marked by a very rapid TYPE OF SOIL 5 AND
VOID RATIO : 0.513
TYPE OF SOIL : SAND
VOID RATIO : 0.613
increase in load with in.creased rotations up to a ro-ta- SIZE OF FOOTING
HE I GHT OF PULL ; 24
4A x 12"
6
SIZE OF FOOTING : 4,~
HEIGHT OF PULL : 24
lt 12"
0.0 0.0
tion of about 5o, with a rapid decrease in the slope of
the load versus rotation curve for rotations between 5o
and 20°. Photographs of the surrounding soil surface
for typical tests are shown in Figures 11 through 13.
~ 0.5
\
' o-"--o -~ /"-----
~ 0.5
I\-.,/ -·-·--·--...../ r-----

On the basis of these model tests the load corresponding

1.0
0 10 15 20
1.0
0
I
10 15 20
to a footing rotation of 5o was chosen t() use in corre-
ANGLE OF OEFLECTION, deorees
lating the theory with test data.
Figure 7. Typical tests in Ottawa Sand.

TABLE 3. RESULTS OF MODEL FOOTING TESTS

Description of Tests Results of Tests Soil Parameters

Size of Max.
Footing, Height Load, De fl. Load, P Pt. of c </> 'Y
Tes.t d X D, of Pull, Pmax @ Pmax, @ 5o, Rotation
Number Type of Soil Inches H, Inches Lbs. De g. Lbs. @ 5°, a/D PSF De g. PCF

1 Ottawa Sand 4 X 12* 24 28.3 6° 30' 28.1 0.74 37 109

2 Ottawa Sand 4 X 12 24 34.0 2° 30' 29.0 0.60 37 109
3 Ottawa Sand 4 X 12 24 28.1 4° 30' 28.0 0.55 37 109
4 Ottawa Sand 4 X 12 24 29.7 3° 10' 28.3 0.62 37 109
5 Ottawa Sand 4 X 10 20 20.0 37 109
6 Ottawa Sand 4 X 12 24 33.0 4° 00' 31.0 0.55 37 109
7 Ottawa Sand 4 X 12 24 26.1 3° 40' 25.6 0.63 37 109
8 Ottawa Sand 4 X 12 24 9.3 0.64 32 102
9 Ottawa Sand 4 X 12 24 14.3 0.65 32 102
10 Ottawa Sand 4 X 10 20 22.0 3° 30' 20.8 0.63 37 109
11 Ottawa Sand 4 X 10 20 18.5 2° 50' 16.6 0.62 37 109
12 Ottawa Sand 4 X 10 10 33.5 4° 20' 33.3 0.60 37 109
13 Ottawa Sand 3 X 12 24 22.0 4° 50' 22.0 0.67 37 109
14 Ottawa Sand 3 X 12 12 48.0 4° 40' 48.0 0.58 37 109
15 Ottawa Sand 2 X 12 24 22.3 0.70 37 109
16 Ottawa Sand 4 X 12 12 52.0 4° 10' 51.3 0.63 37 109
17 Ottawa Sand 4 X 12 0 115.0 4° 10' 113.0 0.69 37 109
E-1 Easterwood Clay 4 X 10 20 498.0 7° 40' 488.0 0.46 2810 9 133
E-2 Easterwood Clay 4 X 12 42 410.0 9° 00' 385.0 0.53 2810 9 133
E-3 Easterwood Clay 4 X 12 24 500.0 0..54 2810 9 133
E-4 Easterwood Clay 4 X 12 12 615.0 0.64 2'810 9 133
E-5 Easterwood Clay 3 X 12 24 352.0 7° 30' 340.0 0.63 2810 9 133
C-1 Trinity Clay 4 X 10 20 246.0 4° 30' 245.0 0.56 3790 126
C-2 Trinity Clay 4 X 12 24 357.0 6° 20' 348.0 0.68 3790 126
C-3 Trinity Clay 3 X 12 24 290.0 3° 20' 273.0 0.66 3790 126
L-1 33% Trinity Clay 4 X 12 24 120.0 0.73 794 3.5 123
67% Concrete Sand
L-2 67% Trinity Clay 4 X 12 24 164.0 0.59 1152 5 126
33% Concrete Sand
L-3 33% Trinity Clay 4 X 12 24 238.0 0.63 1411 12 132
67% Concrete Sand

*4-inch diameter X 12-inch depth.

PAGE ELEVEN
 40 0 40 0 700 700

!~·- i - - ·
350 - - 350
I
..'\ _}· i

II
I
60 0
II
60 or-- .
i
30 o~
0
"'--~
,
.. .---· L-.-+-·-~
• i ( I
I

'I
500 50
1/ I

r·~.""'
25 250

0
I " I
I

II
!
i
I I 400

I
40 0 '

150
0

I '·""' -·, "-...

0

150
I
i
I
!
100 100
I I
a· 30 0
""-'
0
!
i
I
I
Q300
""0
-'
I
!
I

0
i 0
200 200

I I
I I
0 TEST NO. C 1 0 TEST NO. C ~ 10 0 100
TYPE OF SOIL : TRINITY CLAY TYPE OF SOIL - TRINITY CLAY
MOISTURE CONTENT : 18% MOISTURF CONTENT : 18%
51 ZE OF FOOTING : x 10" f SIZE OF FOOTING . 4: x 12.•
HEIGHT OF PULL 20 HEIGHT OF PULL 24
0.0 0.0
I I
~ 0.5

1.0
~/r
I
··--- ~ 0.5

1.0
.____,_ ,_./ ··- 0

0.0
TEST
TYPE
NO.
OF SOIL
MOISTURE CONTENT
SIZE OF FOOTING
HEIGHT 0 F PULL
E 3
: EASTERWOOD CLAY
17.5•1. - 19•1.
4• 11. 12"
24 '
0.0
0 TEST_ NO. E -4
TYPE OF SOIL
MOISTURE CONTENT
SIZE OF FOOTING
liEIGH T OF PilL
: EASTERWOOD CLAY
17.!5% - 19'"1.
4" x 12M
12'

0 10 15 20 0 10 15 20
/'-.
ANGLE OF DEFLECTION , degrees
~ 0.5

IV
-....- "
~ o.
5v~ 1~.-~--··
Figure 8. Typical tests in Trinity Clay. to
0 10 15 zo
1.0
0 10 15 20

ANGLE OF DEFLECTION , degrees

Figure 9. Typic(J)l tests in Easterwood Clay.

300 300

I i I
I
i
!
I
I
I I
250
I ---
I
250
i
I !

II
! v-·---
I ! ~!',.·:/
I I ! ~'

}l'
200 20 0

.... I
,,
. ..,
0 150 0 150

~r·I T /!I
<(
<3 0
-' -'
~1~ I
0

JO 0
I 100
I
/;· I
I
I
I
I
I

I I
50
01/ I

0 T ST L-1
I I 0 TEST NO. L 2
I
TYPE OF SOIL :TRINITY CLAY 67-J.
SAND 33%
SIZE OF FOOTING 4: x 12M
HEIGHT OF PULL 24
0. 0

~ 0.51---1----->-----l-------j
D
ci 0.5
·-·- f_--.--
I.OoL.--.L.--IJ,0---1,':5---:-20 10 I
0 10 15 20

ANGLE OF DEFLECTION, deg<ees

Figure 10. Typical tests in clay-sand mixtures.

PAGE TWELVE
 Figure 11. Test of footing m Ottawa Sand. Figure 13. Test of footing m Trinity Clay.

Figure 12. Test of footing m Easterwood Clay. Figure 13a. Test se·t up at Easterwood.

Comparison of Test Results With Theory

In developing a correlation between the new theo·ry
and the model tests reported in this paper, it was neces-
sary to determine the influence of the various coeffi-
cients introduced in Research Report 105-l and to
determine the optimum value of these coefficients.
As discussed in 105-l, it was necessary to intro-
duce a modifying factor for the Rankine coefficients of
passive and active earth pressure. This factor was
designated B., and was applied as follows.
An additional coefficient was dictated by analysis
of the model tests in Ottawa sands. It was found that
predicted values of the footing rotation point were con-
siderably lower than those values observed in the tests.
Study of this problem indicated that the soil surround-
ing the footing was developing higher passive stresses
in the direction of the applied load (above the rotation
point) and lower soil stresses opposite in direction to
the applied load (below the rotation point) than the
unmodified theory indicated.

K1 Kv B tan 2 ( 45° + t) It was surmised that the shearing stresses, acting

downward on the soil in front of the advancing face of
the footing above the rotation point, produced a slightly
Kz = 2B tan ( 45° + f) greater confining pressure on this soil than would be
indicated by the weight of the soil alone. Similarly, the
KA = B tan 2 ( 45° t) upward shearing stresses on the soil behind the footing
below the rotation point act to reduce the effective
PAGE THIRTEEN
confining pressure. This change in confining pressure a high influence on the theo-retical solution. The unit
was accounted fo·r by a mod.ification of the effective weight coefficient, k, has a high degree of influence on
unit weight of the soil in the following way. the predicted position of the point of rotation. The earth
pressure coefficient modifier, B, has a great ·influence
Y1 y (1 + k tan cf>) on the predicted load.
Y2 y (1 k tan cf>)
In Figures 15 and 16, which represent soils with co-
Where: hesion only and soils with both cohesion and an angle
Y1 Modified unit weight of soil in the direc- of shear resistance, respectively, the great influence of
tion of the applied load. B is again shown, while k is no longer very influential.
Y2 Modified unit weight of soil in the direc- A study of coefficient sensitivity charts and compari-
tion opposite the applied load. son of predicted and observed loads indicated that all
k Unit weight coefficient. coefficients except B could be held constant for the
range of soils tested in this program. The values set
Other coefficients introduced or used in Research for these coefficients are:
Report 105-1 were: Jl J2 = 0.7
J1 Coefficient of shear stress, vertical foot- Ko =
0.5
ing surfaces. k = 0.5
J2 Coefficient of shear stress, footing bottom. It was found that the value o.f B necessary to suc-
K0 Coefficient of earth pressure at rest. cessfully predict ultimate loads varied with the soil
parameters of cohesion (c) and angle of internal friction
The influence of these coefficients on the predicted ( cf>) . The average values of B indicated by the tests
load and the point o.f rotation was determined by allow- on Ottawa sand, Easterwood clay, laboratory clay-sand
ing each to vary while the others were held constant. mixtures, and Trinity clay were used to develop a pre-
Typical sets of curves developed in this manner are diction equation for B. This equation was determined
shown in Figures 14, 15, and 16. from the function of cf> used in the theoretical solution
Figure 14, representing a cohesionless material such and a linear model in terms of c. This prediction equa-
as the dry Ottawa sand, shows that two coefficients have tion is given and presented in graphical form in Figure
17.

0.6
0.7
-
,: A'-... >-'
z"
;~
(5 0.7 'h 5Q.
'" :e.~ z
0
0.6
r-o~ 1!:1 ~ -I
~- o - "'~- _,.__
4!
() r:: _,._
,......
~=
., 1-
()
~
""'---- .._""'-.
"'~~
>-,_;

c: b oc
0.6 >-
()
>·
~
:r:
1-
0.5
.
,.o.
0:
-~
Q.
w
0
OJ

" 0.5
LEGEND• B -X 0.4
K0 - 0
LEGEND:
k -A
J, - ·
h-0
60
1000

/" X

/
-·-
40
X/ 600

/
-·- -·- ~==
0
,;
·<
-fi rl----' !=O=-
-/
c s
..J
X

"'
.J
·~ (/)
WID
1-..J
20
/
l< 0(/)
wm
1-..J
!<, 600
/
:;]a.
!,!I
OQ.
w
a:
Cl.
/
X:
a:
Cl.
X
/
/
-·
-
0
400 -~
=:=;u- ·u ·u ·u

~
X

200
B 0 4 6 8 B 0 4
0.3 0.5 0.7 0.9
k 1 Ko 0.1 0.3 0.5 0.7 0.9 0.2 0.4 0.6 0.8 1.0

Figure 14. Coefficient sensitivity, c=O psf, cf>=37", Figure 15. Coefficient sensitivity, c=3900 psf, cp=Oo,
H=2 ft., D=1 ft., d=0.33 ft. H=2 ft., D=1 ft., d=0.33 ft.
PAGE FOURTEEN
 H=2, 0=1 1 d = 0.33 TABLE 4. COMPARISON OF TEST RESULTS WITH
THEORY, COHESIVE SOILS
0.7
(1) (2) (3) (4) (5)
Test No. Load at Predicted Loads
Rotation Conven- Test Load to
New
..
z
z
5'
0.6
~ :--. d l f l t - fl. ---::-t:Si
of 5°
(1)
tiona! Theoryl
Theory7 (3)
Predicted
Load Ratios
0
j:: ~ -x.-x '<-. lbs. (2) lbs. (1)/(2) (1)/(3)
«,_:
bU. lbs.
"'~
0 0.5
f- E-1 488 200 459 2.44 1.06
:I:
f-
a.
E-2 385 179 337 2.15 1.14
"'0 E-3 500 241 535 2.07 .935
E-4 615 337 871 1.82 .706
800 E-5 340 181 384 1.88 .885
LEGEND: 8 -X K0 - O C-1 245 230 300 1.07 .817
C-2 348 276 344 1.26 1.01
Jl - · k -ll. C-3 273 207 241 1.32 1.13
J2 - 0 L-1 120 63 92 1.90 1.30
600 L-2 164 93 154 1.76 1.06
L-3 238 128 324 1.86 .735

0
<(
400 Using these coefficients in the theo·retical solution
0
__) X/ r:esulted in the comparison of predicted loads and test
o<n
wro
f-_J
/ loads shown in Tables 4 and 5. Also included in these

.1
Ul
o"- /X/ tables are values of loads predicted using the old theory.
"'"'a.
200
0
• ~-== f----~71l
X
8 "£j
In column ( 4) o.f Tables 4 and 5 the ratio of each
test load to the corresponding load predicted by the
conventional theory 7 is tabulated. The test loads vary
00
8
from an average o-f 20% higher than predicted loads
0 2 3 4
for the Trinity clay, to about 500% higher for the
0.1 0.3 0.5 0.7 0.9 Ottawa sand. For all 28 tests (seven different soil con-
0.1 0.3 0.5 0.7 ditions) the test values average 4.09 times greater than
the loads predicted by the conventional theory. Thus,
Figure 16. Coefficient sensitivity, c=ll40 psf, cp=l2o, the test loads are about 300% higher than the predicted
H=2 ft., D=l ft., d=0.33 ft. loads.
Tables 4 and 5 also present a comparison between
the test loads and the new theory in column ( 5) . The
·ratio of test loads to predicted loads shown in column
5.0
,j,: 40
( 5) are about the same for each different type of soil.
lj>=30 Two rather low values of this ratio ( .410 and .630)
are given by tests S8 and S9. These weTe tests in
4.0 -25 Ottaw~ sand which was in a very loose condition. Ap-
"'
,; parently, the new theory gives a rather poor estimate of
"'~ 4>=20 the ultimate load for footings in a very loose sandy soil.
0
0
::E Fortunately, this is not often a condition of practical
3.0
f-
z <1>=15 importance.
"';;::
0
Another test which correlated poorly was SIS. This
... was an extremely slender footing, with a ratio of footing
"'u0 ,j,: 10

2.0 width to depth of 0.167. The test load was 50% higher
"'"'
::J than the predicted load, an error on the conservative
"'"'
UJ
,. =5
side.
"'a.
I
f-
. The average value of the test load to· predicted load
1.0 lj>=O
"'
<(
bJ
ratio was 0.912 for all footings tested. The new theory
thus predicted loads that averaged 10% higher than the
test loads.
0.0 0 1000 2000 3000 4000 5000 The reliability of the theory with variation in the
COHESION, c, lbs/ft
2 1/:eometry of the footing tests (i.e. variation in H/D and
d/D) is shown by Figures 18 through 20. Of the 20
test points compared to theoretical curves in these fig-
8 EQUATION ures, only three show considerable divergence from the
B =0.0000673 c + 10.25 ton</>+ 2.686 ton (45 + </>12) theoretical curves. The test point at a value o.f H/D
- 2.141 ton 2 ·(45 + .p/2)
of 1.0 in Figure 18 is about 30% low and the first and
third points in Figure 20 are 18% high and 13% low,
Figure 17. So.Zution for earth pressure coefficient respectively. With these possible exceptions, the test
modifier. data points closely followed the theoretical curves.
PAGE FIFTEEN
 TABLE 5. COMPARISON OF TEST RESULTS WITH THEORY, OTTAWA SAND

(1) (2) (3) (4) (5)

Test No. Load at Predicted Loads
Rotation Conventional Test Load to Predicted Load Ratios
of 5° New
Theory7 Theoryl
(1) (2) (3) (1)/(2) (1)/(3)
lbs. lbs. lbs.

S1 28.1 5.47 32.5 5.14 .865

S2 29.0 5.47 32.5 5.30 .892
S3 28.0 5.47 32.5 5.12 .862
S4 28.3 5.47 32.5 5.17 .871
S5 20.0 3.79 23.9 5.28 .837
S6 31.0 5.47 32.5 5.67 .954
S7 25.6 3.47 32.5 4.68 .788
S8 9.3 3.83 22.7 2.43 .410
S9 14.3 3.83 22.7 3.73 .630
S10 20.8 3.79 23.9 5.49 .870
S11 16.6 3.79 23.9 4.38 .695
S12 33.3 5.30 38.2 6.28 .872
S13 22.0 4.11 23.3 5.35 .944
S14 48.0 .5.74 50.8 8.36 .945
S15 22.3 2.75 14.9 8.11 1.50
S16 51.3 7.64 52.5 6.71 .977
S17 113.0 14.7 134.0 7.69 .843

Average 4.09 .912.

Range 1.07-8.36 .410-1.50
c.v. 55% 23%

900

/'
150

600 /
,/ 100

HID= 2.
. o''· Rl'i!>l'· ~
?,;)?
300
v2. ~
HID=2

--
0/
.~
50
,; 1-o--
~
CD

-
~ i--9.--
~'hr:!.:.--
...J
a:
a:-
.,
0
0 /
0.0 0.1 0.2. 0.3 0.4 0.5
:;,
"-
0
0
0.0
-_o- ~
0.1
D.

0.2. 0.3
~·
0.4 0. 5
"- diD
0 z diD
0
z FOOTING DIAMETER TO DEPTH RATIO
;:: FOOTING DIAMETER TO DEPTH RATIO
0
;:: ~THEORETICAL ~ TESTS
,__
4
~ ~
g 0
a:
THEORETICAL TESTS

a: 900 1- 150

~
!;( "'
~
0
0

"'9 "'
0
...J

600 100
"-.,

f ' ~-----~--
--. ~
.
diD =0.33

300 50
~.
0
~---- t---o-.
0 2. 3. 4 5 0
0 I 2 3 4 5
HID HID
HEIGHT OF LOAD TO FOOTING DEPTH RATIO HEIGHT OF LOAD TO FOOTING DEPTH RATIO

Figure 18. Easterwood Clay tests. Figure 19. Ottawa Sand tests.
PAGE SIXTEEN
 500
;'
,; 400
~THEORETICAL

iTESTS
-
I
1/
m

{:;/
/
...1
cC
" HID= 2
0

"'
lL
0 300
I/ .
,.~~
z

/
0
o'< '<
i= .
i'!
0
a:
1- 200
?0 '

0
"'
"'0 0/
/'/~
...1

100

••~/
0
~ 0.1 0.2 0.3 0.4 0.5
d/0
FOOTING DIAMETER TO DEP.TH RATIO

Figure 20. Trinity Clwy tests.

Summary
The purpose of the model tests reported in this
paper was to determine if the theory developed in Re-
search Report 105-1 could be successfully used to predict
the ultimate lateral loads which could be resisted by
drilled shaft footings.
The various coefficients involved in the the()retical
solution have been evaluated by correlation of the theory
with the model tests. Based ()n this correlation the
theory can predict, with reasonable accuracy, the ulti-
mate loads on footing models in soils ranging from
cohesionless sand to clays.
The indication given by these model tests is that
the conventional design techniques for service structure
footings are extremely conservative, and that existing
footings are overdesigned by a large margin.
The next step in determining the value of the new
theory can predict, with reasonable accuracy, the ulti-
ings are presently under construction as part of Research
Study 2-5-66-105.

Selected References
l. lvey, Don L., "The()ry, Resistance of a Drill~d Shaft
Footing to Overturning Loads," Research Report No.
105-1, Texas Transportation Institute, August, 1967.
2. Lambe, William T., Soil Testing for Engineers, John
Wiley & Sons, Inc., 1951.
3. Taylor, Donald W., Fundumentals of Soil Mechanics,
John Wiley & Sons, Inc., 1948.
4. Seed, H. B. and Chan, C. K., "Thixotropic Charac-
teristics of Compacted Clay," Proceedings, ASCE,
Vol. 83, S.M. 4, November, 1957.
5. Mitchell, J. K., "Fundamental Aspects of Thixotropy
in Soils," Proceedings, ASCE, Vol. 86, S.M. 3, June,
1960.
6. Moretto, 0., "Effects o-f Natural Hardening on the
Unconfined Compression Strength o·f Remolded
Clays," Second International Conference on Soil Me-
chanics and Foundation Engineering, Rotterdam,
Vol. 1, 1948.
7. lvey, Don L., and Hawkins, Leon, "Signboard Fo()t-
ings to Resist Wind Loads,'' Civil Engineering, De-
cember, 1966, p. 34.
PAGE SEVENTEEN
 Appendix
40

5
4

:u (\i.I
40

('\
30 0
i
/' f--·" ....._
30

5 /
5
1\"
.....
--·- ·-- 25
/ 1\
'·'-...,
0
<[
2.0 0 20 AN ATTIMrT WAS MADe TO r-- . r--.-.~
_,
0 _,<3 '"OTOtfU.I'H THI PtitOtftl:ll Of
c 20
5 5
THII
THE
TIST,
Pl.ASH
LIIHT
ATTACHMENT CAUSED
HODUCED 1Y
<1:
0 .
. OYift IXPOIUIII
PAH:It,
0'
OUI
DIFLICT\ONI
OF THI ltECORDIHI
TO THIS THE ltECOIID
WAt DUTI!:OYID,
..J

15
0 0

5 5
I 10

( MAXIMUM LOAD • 20 LBS.)

0 0
TEST NO. ·1 TEST NO. 2 5
TYPE OF SOIL : SAND TYPE OF SOIL : SAND
VOID RATIO : 0.537 VOID RATIO 0.513
SIZE OF FOOTING 4~ x 12.~ SIZE OF FOOTING 4~ k 12•
HEIGHT OF PULL 24'
0.0
HEIGHT OF PU L L 24"
0.0 '
TEST NO. 5 0
~\
TEST NO. 6
TYPE OF SOIL • SAND TYPE OF SOIL : SAND
VOID RATIO 0.511 VOID RATIO : 0.515
~ 0. 5
-·-
0
';;: 0. 5
. \.-·-, SI·ZE OF FOOTING • 4M x 10" SIZE OF FOOTING : 4': x 12''
___ .........-·
r-" ..-----·- 1--· 1.0
HEIGHT OF PULL • 20"
0.0
HEIGHT OF PULL 24'

A.~ v\
0
0 10 15 20 0 10 15 20

ANGLE OF DEFLECTION, degrees

~ 0.5
if "'" ~ ~---·- -- /'
1.0
0 5 10 15 20

ANGLE OF DEFLECT! 0 N , degrees

0 0 o~~l-~,-~ 40

5 5 35
i
-r--·f-----1----l : 5
I

0 0,~---~---4-----+----- 30
0

/'~
5 5,~----f-----4-----+----- 5

5 I '~.
0 ~"'- o 2ol---~+-----+-----+---~ 0 20
./~.
2
I 0 20
"'_,
f
<(
0_, 0! <(
_,
0 0 <3..J ''\
15 5 51
'"'--.
__.,..f/ li
i
v·
.--1--j--1-- . 10 t--·.
...--· 0

I-I··
0

·~j_I
/'
/'
5 •
! 5

0 TEST NO. 7
i i 0 TEST
I NO. 8
oL-~TE~S~T~Nn~~.~----~----__J 0 TEST NO 10
TYPE OF SOIL SAND TYPE OF SOIL SAND TYPE OF SOIL SAND TYPE OF SOIL SAND
VOID RATIO 0.517 VOID RATIO 0.615 VOID RATIO ~ 0.508 .
SIZE OF FOOTING 12."
VOID
SIZE
RATIO
OF FOOTING
0.616
4" ll 12."
SIZE OF FOOTING -'!" x 12~
~I~H~F0 ;op~~I~G 4: 11 10"
HEIGHT OF PULL 24 20
O.O>r---"H"'EI,_GHTT'-"0"-F_,P;,U';'LL'-'-"!....,--~ O.O r---"H"'EI,GHTT__.,OF"-.!:P:><UL'f'L~_...24,_",--~ 0.0 0.0

0 0 1\
'ri 0.5
l--·- .~
';; 0.
I [/'·--·
~-
5
/-·

0
·-·
1.0
I.OO_L---J,---IJ.O---IJ.5_ _...J20 10 15 20 0 10 15 20

ANGLE OF DEFLECTION, degrees ANGLE OF DEFLECTION, degrees

PAGE EIGHTEEN
 40 40 40 0

5 5 5 0

( 1\
0
0 I \ 0

I
0

5 5 5 0
I
(.---· ·\.,
~"-·- .,-.,-o-
/ ......
0
<l
0

r . ., 0
<l
20 0
g
20
"\ 0
<(
40

I i ~'--·~·-
0
...J

5
• 1\
\
...... _o-_,..,_
0
...J

5
...J
I . ["'-_
0
...J

0
~·-

5 /
t-
0 0
011
I
0 I J
5 5 5 ol:----· j.
I
0 T EST NO. II 0 0 TEST NO. 13 0
I
TEST NO. 12
TYPE OF SOIL : SAND TYPE OF SOIL SAND TYPE OF SOIL SAND SAND
VOID RATIO 0.503 VOID RATIO 0.508 VOID RATIO · 0.503
SIZE OF FOOTING . 4" 1t 10~ SIZE OF FOOTING : 4" x 10" SIZE OF FO"OTING 3" r: 12"
HEIG HT 0 F PULL 20" HEIGHT OF PULL 10 H~"IGHT OF PULL 24"
0.0 ·o .0 0.0

~
I i\ ~
\
\ . !
0.5
\.. +-·-.-·, / II '---. / ·-·- .-~-
5\\
0.
__./ ' - ·-~·: ·-r----.
-- /::0.......--+-----+------l

1.0
0
i '
10 15 20
0
0 10 15 20
1.0
0
i
10 15 20
10
! I
oL-----------1~0-----1L5----~20

ANGLE OF DEFLECTION, degrees ANGLE OF DEFLECTION , degrees

40 80 12 0

/
35 70 0
/ I\

30 60 100
i \ \_-
·- r \
25 50
I .,......... 0
/'/
.
c 20
I
,.<

c 40
I
I
t--
c 0
\
<l
0
...J

15
I
t
<(
0
...J

0
<l
0
...J

0
\
l .
0 0 6 0 II
'\
~

0 TEST NO. 15
TYPE OF SOIL SAND
VOID RATIO : 0.507
SIZE OF FOOTING : 2"
HEIGHT OF PULL 24"
x 12"
0

0 TEST
TYPE
NO. 16
OF SOIL : SAND
VOID RATIO : O.GI7
SIZE OF FOOTING : 4 "
HEIGHT OF PULL 12 11
x 12 11
fllfi
O TEST NO. 17
TYPE OF SOIL : SAND
VOID RATIO : 0.508
SIZE OF FOOTING : 4" x 1211
o.o ' 0. 0 O .Oor----"H"'EI!!!.GH'\-TL.>O!!.F...JPt:>U'-'jLJ.L--'-'_,_o~'·,------.

I'
~ 0.
5"-y r--.-1 ..... 0
';:; 0.
5V\ 1/-...- I\_.
1.0 1.0
0 10 15 20 0 5 10 15 20

ANGLE OF DEFLECTION , "-grees ANGLE OF DEFLECTION, degrees

PAGE NINETEEN
 700 700

60 0
I 60 0

500 50 0

/" ~-.'\
400
/ 400
v' .
\ /"
/

..
a3o 0 o I
/
0
.J

20ol----- 20 0

'

10 0 10 0

0 TEST NO. E I 0 TEST NO. E •2

TYPE OF SOIL : EASTERWOOD CLAY TYPE OF SOIL : EASTERWOOD CLAY
MOISTURE CONTENT : 17.5•!.- 19°/. MOISTURE CONTENT : 17.5% - 19•!.
SIZE OF FOOTING : 4,." x 10" SIZE OF FOOTING : 4• x 12."
HEIGHT OF PULL 20 HEIGHT 0 F PULL 42."
0.0 0.0

~ 0.5
f\ .. ,
---.......v
/-..
~ 0.5
~.
t---:7
v'.
1.0 1.0
.
0 10 15 20 0 10 15 20

ANGLE OF DEFLECTION , degroos

70 0 400 350

350
300
,.-;:/

50
30 0
,r-•,
,. 1\
25 0
~I
"'
<

250

-\ I
400

0' -
.,. -·\
o-zo
~
.J
of ~- ~
20

.I
cS 15

\ 1-- r-. 150

")
g
.J
01/
200 10 0
100

10 0 50 0

0
0 TEST NO. -
C 3
TEST NO. E !5 TYPE OF SOIL : TRINITY CLAY 0 TEST NO. L 3
TYPE OF SOIL : EASTERWOOD CLAY MOISTURE CONTENT : 18%
r
...
TYPE OF SOIL: TRINITY CLAY 33"!.
MOISTURE CONTENT • 17.5%-19% SIZE OF FOOTING : x 12" SAND 67%
SIZE OF FOOTING • 3" x 12• HEIGHT OF PULL 24 SIZE OF FOOTING : 4• 1 12.•
HEIGHT OF Dl'L
' 0.0 '
0.0

-~
0 .
-·
.... 0.5
0.
51'7"--
I/"'"'['\.
0

1.0 1.0
0 10 15 20
0 5 10 15 20
ANGLE OF DEFLECTION, dooreoo ANGLE OF DEFLECTION, deQrua
ANGLE OF DEFLECTJON,deor-

PAGE TWENTY