The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

THE OSTERBERG CELL AND BORED PILE TESTING - A SYMBIOSIS By:

John H. Schmertmann1 and John A. Hayes2

Abstract

This paper describes the O-cell 'bottom-up' method for the static load testing of
bored piles and driven piles, presents its history and lists some of its special
advantages and limitations. This method has almost completely replaced the
previously conventional top-load-reaction testing of bored piles in the United
States. A review of LOADTEST, Inc. files shows two, perhaps previously
underappreciated, aspects of bored pile design and construction which the use of
O-cell testing demonstrates and counteracts: 1) The (measured/estimated) load
capacity ratio tends to increase dramatically with the increasing strength of the
surrounding soil or rock. 2) Several commonly used construction techniques can
dramatically reduce one or both components of shaft capacity.

1. Introduction
Thanks to the new Osterberg Cell (O-cell) testing method^ we can
now statically test full sized bored piles (also known as drilled shafts,
drilled piers, barrettes and caissons) to near their ultimate capacity.
Furthermore, we can do it more conveniently, economically arid safely
than ever before. We also obtain more information about shaft
performance. All this has led, and should lead further, to reducing over-
conservative design and improving construction methods, which in turn,
leads to increasing the use of bored pile foundations, resulting in more
testing, etc... O-cell testing and bored piles have a mutually beneficial,
cooperative relationship - or a "symbiosis."

1
Professor Emeritus University of Florida, Principal, Schmertmann and Crapps, Inc.,
Gainesville, Florida, USA
2
President, LOADTEST, Inc., Gainesville, Florida, USA
The Third International Geotechnical Engineering Conference • Cairo University - Egypt _________________ 5-8 January, 1997

The use of bored pile foundations has increased as an alternative

to driven piles for other reasons as well, including:
■ Higher capacity, with potentially better economics
■ Relatively noiseless and vibrationless installation
■ Deeper foundations possible to overcome scour problems
■ Versatility in, or adaptability to, a variety of subsurface
conditions.

The designer however, must face both the problems of predicting

subsurface soil and/or rock strength and compressibility characteristics
and the difficulty of estimating the impact of construction technique on
the completed shaft. Neither model testing nor laboratory analysis helps
much in dealing with complex intermediate geomechanical materials (i.e.
glacial tills, weathered rock, residual soils). Neither technique lends itself
to assessing the effects of construction methods. Therefore only insitu
prototype testing provides a practical method for assessing the
performance of a bored pile foundation.
Generally the high capacity of bored piles, in combination with the
high cost of top load systems providing over 10 MN reaction, make
conventional load testing too costly or otherwise impractical for routine
testing. The O-cell static load test method, providing high capacities at
affordable cost, has therefore become an attractive alternative method
for testing bored piles.
One can fairly say that in the United States the conventional top-
load testing of bored piles has become nearly extinct. A mid-1994 survey
by the U.S. Federal Highway Administration (FHWA) found that
engineers and contractors considered the O-cell method "the method of
choice" and that its use had risen rapidly to about 65% of all bored pile
testing. (Baker, 1994) This trend has continued and the usage probably
now exceeds 90% in the USA.
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 199

O-CELL TEST ON A ROCK-SOCKETED BORED PILE

 U.S. 231 Highway Bridge Over the Ohio River
Owensboro, Kentucky

Purpose: To determine the bearing capacity of the drilled shafts supporting

a bridge pier.
Method: Shafts drilled with polymer drilling slurry, 1.9 m O.D. casing set on
top of shale and 1.8 m hole drilled 5.8 m into shale. 864 mm load
cell with 1.7 m diameter plate welded to bottom lowered into the
hole. Pumped concrete placed to design top elevation 6f shaft.
Results: Load reached capacity of device at 34.56 mm downward and
7.62 mm upward movement. The ultimate capacity was greater
than 54 MN.
Osterberg Cell Load-Movement Curves
Schematic Section of test Shaft

River
level •

Hydraulic Supply
& ^ -i

30
10 15 20 25
Load(MN>

35.7m
Owner:
Kentucky Department of Highways

Contractor:
National Engineering and Contracting
Company.

Testing:
LOADTEST, Inc.

Rock consists of weathered shale with

average compressive strength of 3.45
MPa.
FIGURE 1
The Third International Ceotechnical Engineering Conference - Cairo Uriiversity - Egypt 5-8 January, 1997

O-CELL TEST ON A DRIVEN PILE

Pines River Bridge
Massachusetts
Hydraulic- supply &
iell tale casing

PILE DATA

460mmO.D. 13 mm.
wall concrete filled
pipe pile driven with a
Delmag D36-13
hammer to 10 blows
last 13 mm.
rJ^VVV Glacial Till j^HH^—Osterberg Load Cell

Weathered
Argilite
TEST RESULTS
Peak load = 1.9 MN (7.62 mm movement at failure)
Skin friction = 36.9 kPa. End bearing = 11.7
MPa (2.03 mm movement) Ultimate pile load
capacity > 3.80 MN
FIGURE 2
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5 - 8 January, 1997

2. Static Load Testing Using the Osterberq Ceil Method

To provide the reader an overview of the test method and to
prepare for the subsequent discussion, we have included real examples
of O-cell test results in the form of 1 page summaries. Figure 1 shows a
driven pile in sand and clay, and Figure 2 a bored pile socketed into
shale. They show the position of the O-cell, the upward and downward
movements recorded, and some comments about the test conditions.
Simpfy put, the O-cell is a sacrificial jack-like device which the
Engineer can have installed at the tip of a driven pile or on the
reinforcement cage of a bored pile. It provides the static loading and
requires no overhead frame or other external reaction system.
Figure 3 illustrates schematically the difference between a
conventional static load test and an O-cell test. A conventional test loads
the bored pile in compression, at its top, using an overhead reaction
system or dead load. Side shear F and end bearing Q combine to resist
the top load P and the Engineer can only separate these components
approximately by analysis of strain or compression measurements
together with modulus estimates.
An O-cell test also loads the bored pile in compression, but from its
bottom. As the O-cell expands, the end bearing Q provides reaction for
the side shear F, and vice versa, until reaching the capacity of one of the
two components or until the O-cell reaches its capacity. Static tests using
the O-cell automatically separate the end bearing and side shear
components. When one of the components reaches ultimate capacity at
an O-cell load Q (see Figure 3). the required conventional top load P, to
reach both side shear and end bearing capacity, would have to exceed
2Q. Thus, an O-cell test load placed at, or near, the bottom of a bored
pffe has twice the testing effectiveness of that same toad placed at the
top.
Tests performed using the O-cell usually follow the ASTM Quick
Test Method D1143, although other methods are not precluded.
The Third International Ceotechnical Engineering Conference - Cairo University • Egypt 5-8 January, 1997

COMPARISON OF 0-CELL AND CONVENTIONAL TESTS

REACTION SYSTEM

n
P=F+Q
1
' "
t
i

F F. Q
T A

EXPANDING
OSTERBERG
I^§§&$
JL CELL

T
Q Q
CONVENTIONAL nr.^nn^n ~ TOP LOAD
TEST OSTERBERG CELL TEST

FIGURE 3
37 The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

Instrumentation used to measure load and deflection is similar to that

used for conventional load tests. Figure 4 shows the typical
arrangement in a bored pile set up for an O-cell test.
After construction of the test shaft, the test operator connects an
automatic pump (electric or air driven) to the hydraulic lines that provide
a pressure conduit to the O-cell. The load applied by an O-cell is
calibrated versus hydraulic pressure before installation and the pressure
applied to the cell is measured using a Bourdon gage or pressure
transducer. A special calibration test to 27 MN of an 0.86 m diameter O-
cell showed that the special design of the O-cell seals typically limits
internal friction to less than 2% of the applied load.
We measure the movements during an O-cell test in the United
States by electronic gages connected to a computerized data acquisition
system. Figure 5 shows the basic instrumentation schematic for an O-
cell test on a bored pile. The total opening, or extension, of the O-cell is
measured by a pair of linear vibrating wire displacement transducers
(LVWDTs) the lower ends of which are attached to the bottom plate of
the O-cell. The upward movement of the top of the O-cell is measured
directly from a pair of steel telltales which extend to the top of the O-cell
('C and 'D' in Figures 4 and 5). These telltales also allow the
measurement of the compression of the test pile. Subtracting the
upward movement of the top of the O-cell from the total extension of the
O-cell (as determined by the LVWDTs) provides the downward
movement of the bottom plate.
The upward movement of the top of the test pile is measured with
digital gages mounted on a reference beam and set over the top of the
test pile ('A' and 'B' in Figure 5). A conventional survey level is used to
check both the stability of the reference beam and the top-of-pile
movements.
Although to date the Osterberg Cell has been used primarily in
bored piles, it is also being used with driven piles. In fact, two of the first
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

TYPICAL BORED PILE INSTALLATION

 son. PROFILE telltale rods
WITH BLOW COUNTS (N)
A B
ELEVATION ELEVATION
(METER) (METER)

------ 0.0 0.0

River
WAWAWA pwwwwx -3.3
'Alluvium
very soft -5.0
silty clay
Alluvium NOTE:
vary loon DRILLED PRODUCTION SHAFT
SHAFT DIAMETER 1.83 METER
fin* sand
PERMANENT CASINO SET TO
-10.0 ELEVATION -17.0 METER TIP
ELEVATION -29.0 METER
31 fins ^grained NOTATIONS:
SC: ONE PAIR OF
Sandstone
SISTER BAR
27 VMRATINC WIRE
coarse grained STRAIN GAGES LVWDT:
Sandstone ONE PAIR OF LINEAR

m
57 -15.0
VIBRATING WIRE
DISPLACEMENT
fin* grained
36 TRANSDUCERS

m
Sandstone

a
25 coarse groined
Sandstone -17.0

50 /10

50 /" -19.0

62

49
-
moldic 22.0
65 Limestone

50 /'2 21" Osterberg Cell

83 -25.0
LVWOT

l!
-27.5
-28.0

-20.0 -
29.0

-25.0

bftok in shofl '

------ -30.0
FIGURE 4
The third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

TYPICAL TEST INSTRUMENTATION

 REfERENCE BEAM

AIR DRIVEN
HYDRAULIC
PUMP WITH
PRESSURE
GUAGE

ntRITAL INOICATOR NOTFS:

A&B - 0.0025 mm division indicators
measuring upward movement of the top of
shaft:
C&D - 0.0025 mm division indicators
measuring shaft compression between the
top 61 the 0-Cell and top of shaft.
LVWDT - Vibrating wire displacement
transducer measuring expansion of the 0-
Cell.
DIGITAL
- SHAFT SIDE SHEAR

-Telltale Casing

-Reinforcing
Steel

-Hydraulic Supply Line

LVWDT -
INDICATOR
ttiiitmiiiiitti
SHAFT END BEARING

FIGURE 5
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

three tests performed with the O-cell were on steel pipe piles. The next
sections describe some of the history of the Osterberg Cell and its
development, discuss some of its advantages and limitations and finally
present some informative case histories.

3. History of the O-cell

Dr. Jorj Osterberg, Professor Emeritus at Northwestern University,
developed and patented the test which now carries his name. He and
Case Foundation, Inc. first used it in an experimental bored pile in 1984.
(Osterberg, 1984) Following this successful prototype test Dr. Osterberg
worked closely with Mr. Charles Guild of American Equipment and
Fabrication Corp. (AEFC) to refine the cell design and promote its use.
Through their collaboration the O-cell evolved from a bellows type
expansion cell to the current design, very similar to the piston type jack
commonly used for conventional tests. However, the piston of the O-cell
extends downward instead of upward.
Engineers at Haley and Aldrich, Inc. (H&A) were the first to use the
O-cell in a practical application. In 1987 they welded a 457 mm diameter
O-cell to the tip of a 457 mm diameter steel pipe pile at a bridge over the
Saugus River in Saugus, Massachusetts. This pile was driven to 10 blows
over the last 13 mm using a Delmag D62-22 diesel hammer. This O-cell
test reached an ultimate side shear of 1.26MN. Later that same year, on a
610 mm diameter steel pipe pile in Rochester, NY, H&A used another 457
mm O-cell fitted with 560 mm top and bottom plates to obtain an ultimate
side shear of 4.0 MN. H&A then recovered the cell, removed the plates
and used it on a second 457 mm steel pipe pile at a bridge over the Pines
River near Saugus, Massachusetts. (See Figure 2^ In 1988 Schmertmann
& Crapps, Inc. performed two more tests, the first on bored piles at" a
bridge in Port Orange, FL. The first pile, 914 mm diameter and 30 m long,
failed in bearing at 2.28 MN. The second, 29 m long with a 1.37 m
diameter, failed in bearing at 3.10 MN. The cell in the
The Third International Geotechnical Engineering Conference - Cairo University - Egypt_____________________________5 -8 January, 1997

second test was then filled with cement grout and this pile was used in the
permanent foundation.
The O-cell test method has steadily gained popularity and, as of
September 1996, approximately 200 tests have been performed on bored
piles in the United States and Southeast Asia. LOADTEST, Inc. (LTI) in
Gainesville, Florida now distributes the O-cell and provides installation
and test support services. Dr. Osterberg continues to promote the O-cell
and provide consulting support.

4. Advantages of the O-cell Test

The O-cell test method offers a number of potential advantages
versus the conventional testing of bored piles. These include:

4.1 Economy : The O-cell test is usually less expensive to perfoim than
a conventional static test despite sacrificing the O-cell. Savings are
realized through reduced construction time and capital outlay for a
test, no top-of-pile reaction equipment requirements and less test
design effort. O-cell tests are typically 1/3 to 2/3 the cost of
conventional tests. The comparative cost reduces as the load
increases.

4.2 High Load Capacity: Bored piles have been tested in Kentucky,
Massachusetts and Georgia to equivalent conventional test loads
of 54 MN, 56 MN and 66 MN respectively. A program is underway
in Florida to carry out a bored pile test with a group of O-cells that
have the capacity to reach a loading of 160 MN. Very high capacity
loading is also possible for large driven piles

4.3 Shear/Bearing Components: The O-cell test automatically

separates the side shear and end bearing components. (Osterberg,
1989) It also helps determine if construction
77»e Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5 - 8 January, \^

techniques have adversely affected each component (see Sect.

8).

4.4 Improved Safetv: The test energy lies deeply buried and there is
no overhead load.

4.5 Rock Sockets: Conventional load tests often have difficulty

adequately testing rock sockets because of limited reaction
capacity and load shedding in the soils above the socket.
Instrumentation interpretation problems often preclude any
accurate separation of socket shear and bearing. The O-cell
places its large test load capability directly at the bottom of the
socket, and also gives an automatic separation of components.

4.6 Reduced Work Area: The work area required to perform an O-cell
test, both overhead and laterally, is much smaller than the area
required by a conventional load system. For example, a 56 MN
O-cell test, conducted in a 3-meter wide median strip of a busy
Interstate Highway, would have been impossible with any other
method.

4.7 Over-water and Battered Shafts/Piles: Although often impractical

to test conventionally, testing over water or on a batter pose no
special problems for O-cell testing.

4.8 Static Creep and Setup (Aging) Effects: Because the O-cell test
is static, and the test load can be held for any desired length of
time, the Engineer also obtains separate data about the creep
behavior of the side shear and end bearing components. Creep
load limits may be obtained which are similar to those from
pressuremeter tests as described in ASTM D4719. As explained
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5 - 8 January, 1997

in 4.9, the engineer can conveniently measure important aging

effects at any time after installation.

4.9 Sequential Testing: Taking advantage of the O-cell's capability of

sequential testing, researchers at the University of Florida have,
since 1994, driven five 460 mm square prestressed concrete piles
with O-cells cast into their tips. The purpose of this ongoing
research program is to assess setup or "aging" effects on heavily
instrumented piles driven in a variety of soil types. A recent (June,
1996) test program instituted by the Louisiana Department of
Transportation will examine similar aging effects in clay soils on a
760 mm square prestressed concrete pile over a two year period.
Both programs illustrate one of the unique advantages of the O-cell
method, namely the ability to carry out long term stage testing with
minimal effort and equipment.

5. Limitations of the O-cell Test

The O-cell method also has some limitations compared to
conventional top load testing. These include:

5.1 Advance Installation Required: With bored piles and most driven
piles, the O-cell must be installed prior to construction or driving.

5.2 Balanced Component Requirement: An O-cell test usually reaches

the ultimate load in only one of the two resistance components. The
test shaft capacity demonstrated by the O-cell test is limited to two
times the capacity of the component reaching ultimate. Also, once
installed the O-cell capacity cannot be increased if inadequate. To
use the O-cell efficiently the Engineer should first analyze the
expected side shear and end bearing components and either
attempt to balance the two to get the most
. The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5 - 8 January, 19^

information from both or unbalance them to ensure the preferred

component reaches ultimate first.

The introduction of multi-level O-cell testing mitigates this

limitation, allowing the Engineer to obtain both ultimate end
bearing and ultimate side shear values in cases where the end
bearing is less than the side shear.

5.3 Equivalent Top Load Curve: Although the equivalent static top
load-deflection curve can be estimated with conservatism, it
remains an estimate. See Section 6 for more details.

5.4 Sacrificial O-cell: The O-cell is normally considered expendable

and not recovered after the test is completed. However, grouting
the cell after completion of the test allows using the tested bored
pile or driven pile as a load carrying part of the foundation.

5.5 Not suitable for certain types of piles: The O-cell cannot be used
to test sheet piles or H-piles. It will also not fully develop the side
shear of a tapered pile when loaded in compression. Installation
of an O-cell on a tapered wooden pile would be difficult.

Interpretation of Test Data

The Osterberg Cell loads the test pile in compression similar to a
conventional top load test, and hence the data from an Osterberg test
are analyzed in much the same way as data from a conventional test.
The only significant difference is that the O-cell provides two load versus
movement curves, one for side shear and one for end bearing. Figures 1.
6. 8 and 11 provide examples. The ultimate load for each component
may be determined from these curves using the criteria recommended
for conventional load tests. To determine the side shear capacity,
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

subtract the buoyant weight of the test pile from the upward O-cell load.
Note that the movement curve for end bearing does not include elastic
pile compression because the load is applied directly.
The Engineer may further utilize the two component curves to
construct an equivalent top load-deflection curve and thereby investigate
the combined two-component pile capacity. Construction of the
equivalent top load curve begins by determining the side shear at an
arbitrary deflection point on the side shear-deflection curve. If the bored
pile is assumed rigid, its top and bottom move together and have the
same deflection at this load. By adding the side shear to the mobilized
end bearing at the chosen deflection, one determines a single point on
the equivalent top load curve. Additional points may then be calculated
to develop the curve up to the maximum deflection (or maximum
extrapolated deflection) of the component that did not reach ultimate
value. Figure 7 presents such an equivalent top-load curve obtained
from the test curves in Rqure 6. In this case both components
fortuitously reached near-ultimate simultaneously, which does not
normally occur.
Points beyond the maximum deflection of the component that
does not reach an ultimate may also be obtained by conservatively
assuming that at greater deflections it remains constant at the maximum
applied load. An example result using this method is shown on Figure 9.
based on the O-cell curves in Rqure 8. We also sometimes, but less
conservatively, extrapolate the non-ultimate component by using
hyperbolic curve fitting and then use this extrapolated curve in the
calculations to produce the equivalent curve, as also shown on Figure 9
for comparison.
As noted by Osterberg (1994), the above construction makes the
following three basic assumptions.
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 199*

25 Osterberg Cell Load-Movement Curves

20 Reference test 127
= ? 7E
15
-S#
I*!
3Q.S 10
.
O>

K|
,

0 -

5
e
?
?S§ -10 j

OOc ■0 = »
log
1 = -20
0

.0 2.5 5.0 75 10.0 125 15.0 175 20.0 225 25.0 275 30.0
Load(MN)
Figure 6
i -20.0 Equivalent Top Load-Settlement

I
Curve
Reference test 127
60

Equivalent Top Load (MN)

20 25 30 35
40
-22.5 •
-25.0
Figure 7
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

Osterberg Cell Load-Movement Curves

Reference test 145

ill »
in •
h
I o
T
-5 -10

III:
■ -- :

•
Assumed compression of
disturbed loose material

| 1 -30
-35 -40

JL.
i^v ;
]
.0 0.5 .0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
1 Load(MN)

Figure 8

Equivalent Top Load-Settlement Curve

Reference test 145
Equivalent Top Load (MN)
4 6 8
-45
Figure 9
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 199?

1. The side shear load-deflection curve resulting from the upward

movement of the top of the O-cell equals the downward top-of-pile
movement in a conventional compression load test.

2. The end bearing load-deflection curve resulting from the

downward movement of the bottom of the O-cell equals downward
bottom-of-pile movement in a conventional top load test.

3. The compression of the pile is considered negligible, Le^ a rigid

pile. Typical bored piles compress 1 to 3 mm.

The first of these assumptions highlights a significant difference

between the O-cell test and a conventional compression load test,
namely the change in direction of the mobilized side shear from
downward to upward. Engineers at LOADTEST, Inc. and researchers at
the University of Florida and elsewhere have investigated the effect of
this direction reversal using the finite element method and also via a
search of the literature. Their results indicate that the O-cell usually
produces slightly Jower side shear than a top load test, but that in
general the effect is small and may be ignored (conservative approach).
A few full scale field tests tend to confirm these findings. Note that the
side shear direction in an O-cell test matches that in a conventional
tension test.
We can also comment on the expected accuracy from the
equivalent top-loaded curve. We know of four series of tests that provide
data needed to make a direct comparison between actual, full-scale, top-
loaded shaft and pile movement behavior and the equivalent behavior
obtained from an O-cell test by the methods described herein. These
involve three sites in a variety of soils, all in Japan, with two compression
tests on bored piles, one compression test on a driven pile and one
tension test on a bored pile. The largest bored pile had a 1.2 m diameter
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

and 37 m length. The driven pile had a 1-m increment modular

construction and a 9 m length. Kisida et. al. (1992) and Ogura et. al.
(1995) detail the aforementioned tests and the results therefrom.
We compared the predicted equivalent and measured movement
behavior at three top movements in each of the above four comparisons,
ranging from 6 to 40 mm movement depending on the data available.
The (equiv./meas.) ratios averaged 1.03 in the 12 comparisons with a
coefficient of variation 9.4%. We believe that these available
comparisons help support the practical validity of the equivalent top load
method described herein.

7. Observations from the O-cell Test Data

A review of the load test results from LOADTEST, Inc. files
indicates that, in most cases, the measured.capacity of the test shaft
exceeds the designer's estimated ultimate capacity. Ironically, the ratio
of measured to estimated capacity (M/E) tends to increase as the
strength of the supporting strata improves. Figure 10 provides a
summary of the M/E ratio data from 25 projects where we have
information regarding the engineer's estimated capacities. Tables 7.1.
7.2 and 7.3 provide more detail about some of these projects in soils,
intermediate materials and hard rocks, respectively.
We based the horizontal positioning on the graph within the
material classifications shown on a subjective assessment of material
characteristics. The data points with an upward arrow indicate tests in
which neither component reached an ultimate capacity and, therefore,
the test would plot at some unknown higher M/E ratio had the O-cell had
enough capacity to reach an ultimate. (Note also that most of the M/E
ratios would be somewhat higher than shown since the O-cell.test
usually determines the ultimate capacity of only one component.)
The Third International Ceotechnical Engineering Conference • Cairo University - Egypt 5-8 January, 1997

RATIO OF MEASURED TO ESTIMATED

ULTIMATE LOADS (M/E Ratio)

m 0 m T-
'f-
JL

31 ^

<D

much higher than show

Hard Roc
- -
© N.
1 \

1 1 \ .. ,r i.e.references.
2. M/E
il XI
'~*~"^««^ f*i\ /^o\ /sS \
^^^^Xr^V \TZS XL^/ \ (0 a.
2
Inc. file
antechnique.

^^ ca
1
ultimate,
Intermediate Mater<

rs are
denoted numbe due mponent
to poor LOADTEST
construction
reached

\ .2
cates neither co
cates low results

Y~0Q
fttoHardSoi

finYco
lo>I v

/cc\ ^j CO
2 =5 -5
(0,
£ £ £

O®
Figure 10
fpe Third International Geotechnical Engineering Conference - Cairo University - Egypt 5 -8 January, 1997

TABLE 7.1 - CASE HISTORIES IN SOILS

FILE I 38 46 106
REFERENCE
LOCATION Truth or Los Angeles Coliseum, Owensboro, Kentucky
Consequences, New Los Angeles, California
Mexico
PURPOSE Test of 2.2 m bored Tested 28 (1.1m Isolate shear and end
pile for remedial diameter) bored piles bearing characteristics
foundation system to allow reduction of
Factor of Safety to 1.5
METHOD Cased to 3.0 m then Drilled uncased and dry 1.5 m diameter bored
drilled under polymer to 19.8 m. O-cell placed pile under water to 23.5
slurry to 22.3 m. O-cell at bottom m. One O-cell placed
placed at 4.0 m above at 17.7 m and another
bottom. Concrete to one at the bottom
5.8 m from surface (multi-level)'
CONDITIONS Interbedded cobbly Interbedded sand and Clay to 3.0 meter, then
sand/sandy grave! gravel, compact to fine to coarse sand
N = 20 - 86 dense with silt layers. N = 7
GWT @ 13.4 m to 30 along shaft,
itermediate material N = 30 at base
RESULTS Tested to 37.7 MN Tested to 10.7 MN Tested to 11.3 MN
Estimated Ultimate M/E = 1.2 End Bearing = 3.1 MN+
Load = 10.3 MN Side Shear = 8.2 MN
M/E = 3.7 M/E = 1.3
GWT denotes ground water table
N denotes standard penetration test blowcount

The O-cell testing shows that in "normal" soil conditions (sand, silt
and clay strata) the engineering estimates are usually reasonably good,
with M/E ranging from about 0.7 to 2.
In intermediate geomechanical materials (soft rocks, weathered
rocks, coarse dense soils, glacial tills) however, the measured/estimated
(M/E) ratio generally increases into the 2 to 5 range. This no doubt reflects
the difficulty in measuring and estimating the strength parameters of these
materials, coupled with natural engineering conservatism. In unweathered
competent rock formations the M/E ratios in Figure 10 are even higher,
from 7+ to 25+. We suspect that to some extent such high M/E ratios in
competent rock have resulted from the inability (prior to the O-cell) to test
the high capacity of these foundations.
The Third International Geotechnical Engineering Conference - Cairo University - Egypt_______________ 5-8 January, 1997

TABLE 7.2 - CASE HISTORIES IN INTERMEDIATE

MATERIALS
RLE 19 28 29 127
REFERENCE
LOCATION Forest Glen, Owensboro, Owensboro, East Milton,
Maryland Kentucky Kentucky Massachusetts
PURPOSE Determine end Determine end Determine end . Determine end
bearing and side bearing and side bearing and side bearing and side
shear components shear to confirm shear to confirm shear to allow
for design. Tested design parameters design parameters shorter shafts if
a 0.9 meter ' Tested at 1.8 Tested at 1.8 possible
diameter bored pile meter diameter meter diameter
rock socket rock socket

METHOD Cased to 5.5 m Cased to top of Cased to top of Cased to a depth

then drilled "dry" to rock at 27.4 m, rock at 27.4 m of 2.8 m then
15.5 m. O-cellat then drilled under then drilled under drilled "dry* to
bottom, concrete polymer slurry to polymer slurry to 9.5m. O-cellat
placed in 2 stages 45.7m. O-cellat 33.5 m, O-cell at bottom of rock
to get ultimate bottom of shaft bottom of shaft socket
shear and end
bearing.
First stage with
1.5 m socket
Second stage with
9.4 m socket

CONDITIONS Clay to 4.6 m, Water to 18.9 m, Water to 18.9 m, Earthfillto1.5m

then weathered then sand and then sand and then shale to
schist to 19.8 m. •gravel to 27.4 m. gravel to 27.4 m. 15m+
Unconfined Then shale and Then shale and Unconfined
compression limestone. limestone with Compressive
strength 500 to coal seams. Strength = 17.2
2500 psi. MPa to 20.7 MPa
RQD = 17

RESULTS Side shear Side shear 24.5 Side shear 25.8 Side shear 27.6
Stage 1, averaged MN (7 mm) MN (7 mm) MN (13 mm)
720 kPa End bearing End bearing End bearing 28.0
End Bearing = = 26.7MN = 27.6MN MN (20 mm)
8.6MN (81 mm) (34 mm) M/E = 3.3
Total shaft M/E>2 M/E>2
capacity exceeds Ultimate capacity Ultimate capacity
17.8 MN not reached not reached
M/E>2

( ) denotes cell movement in mm, downward for EB and upward for SS

Thus it would appear that geotechnical engineers have tended to

increase the "factors of safety" applied to bored pile foundations as the
competency of the founding formation improves. In fact, however, the
increasing M/E ratios reflect increasing "factors of uncertainty" which can
jtte Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

TABLE 7.3 - CASE HISTORIES IN HARD ROCK

FILE 35 31 129
REFERENCE
LOCATION Decatur, Alabama Burgin, Kentucky Stillwater, Minnesota
PURPOSE Verify design parameters, Proof test on a 0.9 meter Pre-design test to
leading to adjustments production shaft. (1.8 MN determine shear
where feasible 0.9 meter design load) parameters for axial test
diameter shaft on 1.2 m diameter shaft.
(Deep-seated lateral load
test also carried out).
METHOD Cased to top of rock Cased to top of rock at Cased to 15.0 m then
socket at 17.7 m then 3.0 m then cored *dry" to drilled to 55.0 m. O-cell
cored to 22.6 m under 6.1 m. O-cell set on s.et at 4.5 m from bottom
water. 26.7 MN O-cell bottom. Concreted to (i.e. 3.5 m socket above
placed at 22.1 m. form 2.0 m socket. O-cell)
Concrete filled rock
socket to 17.7 m level
(i.e. 4.4 m socket)
CONDITIONS Fill sand and clay to Earth fill and sand to Overburden to 15.0 m
8.2 m, then limestone 3.0 m then limestone. then dolostone over
with some voids and clay Unconfined Compressive sandstone.
seams. Unconfined Strength = Unconfined Compressive
Compressive Strength of' 100-140MPa Strength =
limestone = 50-58MPa
100-150MPa
[email protected]
RESULTS Shear Load 22.2 MN Shear Load 8.0 MN Shear Load 21.8 MN
(18 mm) (2 mm) (33 mm)
End Bearing Load End bearing Load Shear + End Bearing
22.7 MN (36 mm) 8.0 MN (4 mm) Load (lower socket)
M/E>5.5 M/E > 9.3 21.8 MN (15 mm)

GWT denotes ground water table

( ) denotes cell movement in mm, downward for EB and upward for SS

be reduced only by better understanding following the use of better

testing methods.

8. The Influence of Construction Technique on Bored pile Capacity

Although geotechnical engineers have made, and continue to
make, great strides in understanding soil and rock mass behavior under
load, our understanding of the interaction between foundations and the
surrounding soil or rock mass continues to be uncertain. This is
especially true for bored pile foundations, which are constructed within
the soil or rock mass. The insitu construction process introduces many
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt_________________________ 5-8 January, 1997

TABLE 8.1 - IMPACT OF CONSTRUCTION TECHNIQUE

IMPACT ON SIDE SHEAR

File Reference SS23 SS27 SS108 SS93 SS93

Diameter 990 mm 910 mm 910 mm 1520 mm 1520 mm

Length of Test Shaft 5.8 m 18.3 m 3.7 m 15.2 m 15.2 m
Subsurface Sand Residual 12.2 m Saprolitic Saprolitic
Conditions soils overburden Clay Clay
(silt/sand) over shale
(rock socket)
G/WHead 2.1m 10.7 m 15.2 m 15.8 m 0.6 m
Construction Heavy sand Drilled "dry*. Drilled "dry" Drilled with Drilled with
Method slurry plus with no fluid to with casing corebarrel- corebarrel.
improper counteract extended to type casing Sidewalls
concrete hydro-static bottom of (no rifling) rifled
placement pressure rock socket
at 15.8 m
First Test Result 0.3 MN (SS) 0.4 MN (SS) 0.4'MN 7.1 MN (SS) -
(SS) (84 mm)
Remedial Test 1.6MN(SS) 6.1 MN (SS) 10.7 MN " 10.7 MN
Result (6 mm) (SS) (SS) (2
mm)

IMPACT ON END BEARING

File Reference EB128 EB141 EB142

Diameter 610 mm 860 mm 1220 mm

Length of Test Shaft 6.7 m 13.1m 13.7 m
Subsurface Decomposed granite Dense to very dense Weathered rock, silty
Conditions siltysand(N = 30to85) sand (N = 30'to 100+)

G/WHead - 0.6 m 6.1m

Construction Drilled dry. Cleaned out Drilled dry. Cleaned out Drilled dry. Cleaned
Method only with auger only with auger with clean-out bucket.
(Estimated 150-200 (Estimated 50 - 80 mm Seepage at base
mm loose material at loose material at base)
base)
Test Result 8.0 MN (33 mm, SS) 4.0 MN (SS) (2.5 mm), 2.2 MN (EB) (127
(127 mm, EB) 102 mm 2.2 MN (EB) (38 mm) mm) Estimated 115 mm
compression of loose 32 mm compression of compression of loose
base material loose material material
Notes: SS = Side Shear Component EB
= End Bearing Component G/W =
Ground water
The Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5 -8 January, 1997

disturbing forces, and sometimes*materials, which can have a profound

effect on the subsequent behavior of the bored pile under load. In some
cases the impact of the construction process can reduce actual bored
pile capacities to 5% to 50% of the capacity achieved with better
construction methods. Table 8.1 provides examples.
One of the characteristics of the O-cell test method is that it
readily provides evidence of any abnormality related either to side shear
or to end bearing capacity. This is an important distinction as compared
to top load testing methods. A top load test may indicate apparently
reasonable test capacity when, in fact, either the side shear component
or the end bearing component could be seriously compromised by
construction technique. The O-cell test by contrast relies on, and
measures, both end bearing and side shear capacities and it is soon
obvious if one of them is deficient. The examples in Table 8.1 illustrate
how the O-cell test method has provided the comparative data we need
to help our understanding of the impact of construction technique on
bored pile capacity.
In example SS 23 the bored pile was constructed by "thinning" the
soil with water and bentonite as the shaft was excavated. The drilling
auger was used to chum the sand-water-bentonite mixture. Concrete
was placed by lowering concrete through the "thinned" soil slurry using a
cleanout bucket. The process resulted in highly disturbed soils both
along the shaft side walls and at the shaft base. The construction
procedure was changed to lower the sand content in the slurry to about
8% and the concrete was placed by tremmie (pumped) methods. The
testing showed that the side shear increased from 15 kPa to 70 kPa as a
result of the improved construction procedure.
In example SS 27 the contractor attempted to drill the shaft "dry,"
with no slurry to counteract hydrostatic pressure from groundwater. The
residual soils were impermeable enough that seepage became evident
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 195.!

only during drilling near the shaft base. Disturbance to the surrounding
soils' was massive, however, resulting in a failure load in side shear of
0.44 MN. A second test shaft constructed using water as a stabilizing fluid
resulted in an ultimate load of 6.1 MN in side shear.
In example SS 108 a rock socket test shaft was drilled "dry" with
casing extending to the bottom of the shaft but with no water inside the
casing. The hydrostatic head outside the casing was 12.2 m. The 3.7 m
rock socket was then filled with concrete followed by pulling the casing
back about 4.0 m. Virtually no side shear bond developed between the
shale and concrete. The test shaft was reconstructed with casing
extending only to the top of \be rock socket and with water inside the
casing to counteract hydrostatic pressures. Concrete was placed by
tremmie methods. The side shear capacity of the reconstructed shaft
exceeded the capacity of the O-cell (10 MN).
In example SS 93 the test shaft was drilled through 30.5 m of
saprolite clay using a rotating corebarrel type casing technique. The
material inside the casing was removed with a clam type digging tool. The
casing was pulled in sections as the concrete was placed. The O-cell was
set at a depth of 15 m. A special "rifling" tool was used to roughen the
side walls of the 15 m. section of shaft above the O-cell. The test results
showed that the lower smooth-walled section of the shaft failed at a load
of 7.1 MN and a movement of 84 mm. The upper rifled section, however,
sustained a shear load of 10 MN with a movement of less than 2 mm.
Examples EB 128, EB 141 and EB 142 illustrate the impact on end
bearing capacity of loose material at the base of the shaft. In all three
cases the test shafts were drilled "dry" (i.e. no stabilizing fluid) making it
possible to see the base of the shaft. The rough estimates of the depth of
loose material at the base were made by "sounding" with a weighted tape
and by visual observations.
fte Third International Ceotechnical Engineering Conference - Cairo University - Egypt 5-8 January, 1997

Figure 11 shows a typical end bearing load/movement curve from

EB 128. The testing shows that for all practical purposes the end bearing
capacity was lost (i.e. only 0.25 MN of end bearing at a movement of 25
mm). It should be noted that in all of these cases a typical top load test
would not have revealed a significant problem with the test shaft capacity
since the side shear capacity would have masked the end bearing
deficiency.
Such experiences lead one to reflect on the potential bottom-
condition problems wherein water or drilling mud prevents a simple visual
inspection. As we have often seen when we inspect such shaft bottoms
with a special video camera system designed for this purpose,
considerable effort is often required to provide a "clean" bottom under
those conditions. These experiences also reinforce the importance of load
testing to determine the effect of construction technique as well as to
assess side shear and end bearing parameters.

Osterberg Cell Load-Movement Curves

Reference test 128
 S

j|f
O

; S^ \
2b

-
25 s * i

. .
- - Assumed compression of . ..
50 disturbed loose material
:

-
' L '.''.*

75
- ;
10
12 .... ——_-... .-•-...
HT^ ^-—^
5
1C
A 4 5 8 9 10
6
Load (MN)
Figure 11
The Third International Geotechnical Engineering Conference - Cairo University - Egypt 5-8 January, lggj

9. Summary and Conclusions

9.1 The O-cell testing method consists of placing a hydraulic jacking

device at or near the base of a bored pile or the tip of a driven pile
and expanding the device to apply to the test load. Appropriate
instrumentation measures the response.

9.2 The O-cell testing method has many significant advantages over
conventional top-load-reaction testing, and often makes practical
an otherwise impractical testing situation.

9.3 The O-cell testing method has found its greatest use in the testing
of bored piles. Approximately 90% of such tests in the United
States now utilize this method.

9.4 The O-cell testing method obtains two load-deflection curves.

Combining the curves can create an equivalent top-load deflection
curve with sufficient accuracy for most engineering applications.

9.5 Our experience shows that generally the stronger the supporting
material, the larger the ratio between the measured shaft capacity
and the prior engineering estimate of capacity. This ratio often
exceeds 10 in rock sockets!

9.6 The techniques of shaft construction play an important part in

subsequent shaft capacity. Improper hydrostatic balance, not
desanding a slurry, poor bottom cleaning technique, failure to
roughen side walls, dropping concrete through water and
premature casing withdrawal all can seriously reduce capacity.
The effects of such defects may be masked in a conventional top
load test, but become obvious in an O-cell test.
flie Third International Geotechnical Engineering Conference - Cairo University - Egypt 5 -8 January, 1997

List of References

Baker, C. N., Jr., 1994, "Current U.S. Design and

Construction Practices for Drilled Piers", Proceedings: 1st
International Conference on Design and Construction of
Deep Foundations, U.W. Federal Highway Administration,
Vol. 1,pp.305-323.

Kishida H. et al.. 1992, "Pile Loading Tests at Osaka Amenity Park

Project," Paper by Mitsubishi Co., also briefly described in
Schmertmann (1993).

Ogura, H. et al.. 1995, "Application of Pile Toe Load Test to

Cast-in-place Concrete Pile and Precast Pile," special
volume Tsuchi-to-Kiso' on Pile Loading Test, Japanese
Geotechnical Society, Vol. 3, No. 5, Ser. No. 448. Original in
Japanese, Translated by M. B. Karkee, GEOTOP
Corporation.

Osterberg, J. O., 1984, "A New Simplified Method for Load

Testing Drilled Shafts", FOUNDATION DRILLING. Vol.
XXIII, No. 6 (July/August, 1984), ADSC, p.9.

Osterberg, J. O., 1989, "New Device for Load Testing

Driven Piles and Drilled Shafts Separates Friction and End
Bearing", Proceedings: International Conference on Piling
and Deep Foundations, London, A. A. Balkema, p.421.

Osterberg, J. O., 1994, "Recent Advances in Load Testing

Driven Piles and Drilled Shafts using the Osterberg Load
Cell Method", Presented at the 1994 Geotechnical Lecture
Series Sponsored by the Geotechnical Division of the Illinois
Section, ASCE, Chicago, Illinois.

Schmertmann, J. H., 1993, The Bottom-Up, Osterberg

Method for Static Testing of Shafts and Piles", Proceedings:
Progress in Geotechnical Engineering Practice, 13 Central
Pennsylvania Geotechnical Seminar, Central PA ASCE &
PennDOT, Hershey, Pennsylvania, 7 pp..