You are on page 1of 6

Geotechnical and Geophysical Site Characterization 4 Coutinho & Mayne (eds)

2013 Taylor & Francis Group, London, ISBN 978-0-415-62136-6

The use of Osterberg cell load tests to predict piles resistance


A. Pinto & X. Pita
JetSJ Geotecnia Lda, Lisbon, Portugal

ABSTRACT: The use of the Osterberg cells (O-cell) load testing technology has been applied in the last
years as a complement to the geological and geotechnical information, allowing the prediction of piles
and barrettes bearing and shaft resistance, as well as the verification of the construction procedures. The
aim of this paper is to present a case history where the O-cells were successfully used for the calibration of
the piles execution and design, for a high rise building, located at the Luanda bay, in Angola, emphasizing
its contribution for the optimization of the foundations design. At the end of the paper, the main advantages and limitations of the O-cells, comparing with conventional top load static tests, also performed in
the same site, are presented.

INTRODUCTION

The O-cell static bottom up load testing is being


used since 1984, where it was first adopted on
an experimental bored pile. In 1987 the O-cell
load test was used for the first time on a practical application on a 457 mm diameter steel pipe,
where a maximum load of 1.26MN was applied
(Schmertmann & Hayes (1997)).
The static and bi directional O-cell bottomup
load testing was developed as an alternative to the
conventional top load reaction testing, providing
a practical method for assessing the performance
of bored piles, as well as driven piles and barrettes,
allowing the confirmation of the main construction and design criteria, as well as the comparison
with the geological and geotechnical conditions
considered at the design stage (Fig. 1).
The O-cell testing method consists on placing a
sacrificial hydraulic jacking device at, or near, the
base of a bored pile or at the tip of a driven pile, and
expanding the device in order to reach the test load.
Appropriate instrumentation along the pile shaft
is very important in order to access the pile and
soil response, lateral and end bearing resistances.
This method allows the construction of two load
deflection curves: loadupward deflection, assessing upper skin friction, and loaddownward
defection, assessing base resistance and lower frictional capacity. The combination of the two curves
allows the calculation of an equivalent top-load
deflection curve with sufficient accuracy for most
engineering applications (Fig. 2). The test is considered complete after reaching ultimate capacity
above or below the O-cell, or upon reaching the
maximum capacity of the O-cell.

Figure 1. Comparison between conventional static top


load and bottom up O-cell test.

Figure 2.

Typical O-cell testing method results.

1773

COUNTINHO_Vol 2.indb 1773

8/13/2012 5:42:41 PM

2
2.1

2.3

CASE HISTORY
Introduction

The presented case history relates to the bored


pileraft foundation of a high rise building at the
Luanda bay, in Angola. The excavation works
comprised an area of about 1800 m2, with a rectangular shape of about 50 36 m2 and 10 m
of overall depth, necessary for the construction
of 3 underground floors and 26 upper floors
(Fig. 3).

Downloaded by [Engineers Australia ] at 09:51 26 June 2015

2.2

Geologic and geotechnical settings

The geological and geotechnical conditions comprised high permeability sandy soils with the
ground water table located about 1.5 m below the
ground surface, and oscillating due to the influence
of the tides at the Luanda bay. The sands resistance increased with depth, as assessed through
several bore holes and in-situ SPT tests, as well as
through laboratorial identification tests from the
collected samples (Fig. 4).

Figure 3.

Building location and perspective.

Figure 4. Solution cross section and geological


conditions.

Main design and execution criteria

As main design criteria it was considered the need


to avoid the executions of ground anchors below
the ground water table, as well as the control of the
ground water inflow to the excavation pit. This criteria led to the development of a solution of peripheral 0.60 m diaphragm walls, with 8 m of embedment
length. The walls were braced during the excavation
works by a 3 m thickness horizontal jet grouting
sealing slab, located below the excavation final level.
Additional support of the diaphragm wall was provided by three levels of reinforced concrete bracing
slabs, acting as stiff diaphragms, located at the level
of the basement floors, and cast against the ground
during the excavation works. Those slabs will become
part of the basement structure. The jet grouting sealing slab would also be used for limiting the ground
water inflow through the excavation base. For the
building structure foundations, 1 m diameter bored
piles with 20 to 30 m length were adopted, being
capped by a reinforced concrete raft with 0.6 to 2.0 m
thickness (Figs. 46). The piles were built before the
excavation works (Pinto & Pita (2011)).

Figure 5.

Plan of the adopted bracing solution.

Figure 6.

Plan of the foundations solution.

1774

COUNTINHO_Vol 2.indb 1774

8/13/2012 5:42:45 PM

The design of both, earth retaining and foundations solutions, was performed using 2D FEM
programs (Plaxis V8 and SAP2000-V14).

Downloaded by [Engineers Australia ] at 09:51 26 June 2015

2.4

Pile testing

In order to assess the piles main design and execution criteria, four O-cell tests were carried out
prior to the excavation works. The main aim of
the tests was to evaluate the bored piles shaft and
base resistance, as a complement to the geological and geotechnical information, as well as the
consequences of the use of bentonite slurry and
temporary casing to stabilize the piles hole. Piles
were drilled and cleaned using telescopic Kelly
bar with auger and bucket.
Two tests were performed on experimental piles
(P1 and P2, with 20 m and 25 m respectively) and
two tests were performed on working piles (W3
and W4, with 20 m and 25 m respectively). The
O-cells were installed at approximately 10 m above
the piles toe. The first 10 m, correspondent to the
excavation depth were not casted.
Complementary to the four O-cell tests, a conventional topbottom static load test was also
performed. This test was performed on a 25 m length
bored pile. The main aim of this test was to evaluate
the bored piles resistance, including the creep effect.
In this paper, are presented and compared the
results of the 25 m length W4 bored pile (tested with
e O-cell) and the 25 m length bored pile tested through
the conventional topbottom static load test.

3
3.1

Figure 7.

O-cell bottom up static W4 test layout.

Figure 8. O-cell bottom up static W4 measuring system.

BOTTOM UP O-CELL STATIC LOAD


TESTS
W4 O-cell test

The W4 pile test was performed using two 330 mm


O-cells, located 9.1 m above the pile toe. Two and
three levels of twin sister bar vibrating wire strain
gauges, attached diametrically opposed on the
reinforced cage, were installed below and above the
base of the O-cell assembly (Fig. 7).
The complete measurement system was divided
into two independent ones: the main system, twin
sister bar vibrating wire strain gauges, and the
reference measurement system, steel beams. The
measurement systems were fixed on the pile head
as well as on the measure beams (Figs. 79).
The maximum sustained bottom up bi-directional load applied to the pile was 2.9MN
(3.7 MPa). At the maximum load, the displacements above and below the O-cell, were 4.1 mm
and 42.8 mm, respectively. The gathered data
enabled the estimation of the top load curve of the
tested pile.

Figure 9. O-cell bottom up static W4: Strain gauge load


distribution.

For a top load of 3.5MN (4.5 MPa), which corresponded approximately to the pile working load,
the test data indicated a settlement of approximately 6.2 mm. The shaft resistance ranged from 20
to 70 KPa (sands with NSPT blows lesser than 38),

1775

COUNTINHO_Vol 2.indb 1775

8/13/2012 5:42:51 PM

at the upward direction, and from 100 to 160 kPa


at the downward direction (sands with NSPT blows
bigger than 60).
The tested pile behavior was calculated for a
combined shaft and end bearing load of approximately 5.9MN (7.5 MPa). For a top loading of
3.5MN, the adjusted test data indicated that the
pile would settle approximately 6.2 mm, of which
2.8 mm was the estimated to be due to the additional elastic compression (Figs. 2, 10 and 11).

Downloaded by [Engineers Australia ] at 09:51 26 June 2015

3.2

W4 O-cell test remarks

The obtained results allowed for the confirmation of both the suitability of the construction
method, as well as the assumptions for the design
shaft and end bearing resistance. It was possible to
verify that the values of the shaft resistance were
in accordance with the geological information
regarding the sandy materials, with bigger values,
about 160 kPa, at the more compact sands.

CONVENTIONAL STATIC TOP


BOTTOM LOAD TEST

4.1 Conventional static topbottom load test


The test device comprised one working pile with 1 m
of diameter and a reaction system with two reaction
piles, with also 1 m of diameter, connected to a steel
beam, bridging a span of 8 m. At the top of each
reaction pile four steel strands, sealed on the reaction
piles allowed the direct connection to the reaction
beams. The reaction system was designed to accommodate a load of about 4.8MN (Figs. 12 and 13).
Taking into account the previous four O-cell
tests, only the vertical head displacements of the
pile were monitored. The load was applied by a
hydraulic jack and controlled by a load cell. The
complete measurement system was divided in two
independent ones: the main system and the reference measurement system. The measurement systems were fixed on the pile head as well as on the
measure beams.

O-cell test: Mobilized downward shaft

Figure 12. Conventional topbottom static test layout.

Figure 11. O-cell test: Mobilized upward shaft resistance.

Figure 13. Conventional topbottom static reaction


structure.

Figure 10.
resistance.

1776

COUNTINHO_Vol 2.indb 1776

8/13/2012 5:42:56 PM

Downloaded by [Engineers Australia ] at 09:51 26 June 2015

A maximum load of approximately 4.8MN


(6.1 MPa) was achieved, corresponding to approximately 1.5 times de working load. The main
results fulfilled the test requirements, confirming
also both the geological information and the O-cell
tests results.
For the service load and for the maximum load
of 4775 kN, the vertical displacement was about
6.1 and 12.6 mm, respectively, very similar to the
one obtained through the W4 O-cell test (Fig. 14).
The test also confirmed that the creep was very
low for all the load stages, including the maximum
load at the 4th cycle. At this cycle the maximum
deformation was 0.47 mm in 7 hours, leading to a
rate of 0.067 mm/hour, much lesser than the creep
criteria of 0.25 mm/hour (Fig. 15).
4.2

Figure 16. Comparison of results of the conventional


test and the O-cell static load test.

Conventional load test remarks

The obtained results allowed the confirmation


of the very good correlation between the conventional load test and the W4 O-cell test, mainly the

maximum deformation value of 6.1 mm and 12.6 mm


for the service load (3.5MN) and the 1.5 times the
service load, respectively (Figs. 1416).

Figure 14. Conventional topbottom static test:


loadhead displacement.

An analytic model was used to predict the pile


behavior. For this purpose an axisymmetric FEM
model was used (Plaxis V8), with hardening soil
failure criteria. The soil parameters were calibrated
trough a back analysis (Fig. 17) in order to match
the load tests results.
The obtained displacement for the service load,
about 6 mm, is compatible with the good performance of the building structure under service
conditions.
It should also be pointed out the good correlation between the FEM analytic model and both the
conventional top load test and the O-cell load test
(Fig. 18).

Figure 15. Conventional topbottom static test: Creep.

ANALYTIC FEM MODEL

MAIN CONCLUSIONS

It was presented in this paper a case history where the


use O-cell load testing technology has been applied
successfully as a complement to the geological and
geotechnical information, allowing the prediction of
the piles bearing and shaft resistance, as well as the
verification of the piles construction procedures.
This situation was also validated by the execution
of one conventional static load test, as well as through
an analytical axisymmetric FEM back analysis
(Fig. 18), confirming the importance of the pile load
tests for the design and execution criteria of piles,
as already stated in some codes of practice, as for
instance the Eurocode n7Geotechnical Structures.

1777

COUNTINHO_Vol 2.indb 1777

8/13/2012 5:43:04 PM

Downloaded by [Engineers Australia ] at 09:51 26 June 2015

Figure 17.
parameters.

FEM model and main geotechnical

Figure 18. Displacementload curves: Analytical


model, conventional top load test and O-cell test.

In this code the values of the partial safety factors on


the resistance side could be decreased in function of
the overall number of full scale load tests, leading to
an optimization of the piles design.
It should be stressed that the validation of the
pile design, enabled by the performed full scale
load tests, allowed the measurement and validation
of the piles end bearing and shaft resistance, confirming the prediction of the building foundations
good performance under service loads.
Finally it should be point out the versatility of the
O-cell load tests, with both economical and schedule
advantages, comparing with the conventional static
load tests (Fellenius (2001)). However, when necessary in more complex geological scenarios, as the
one at the present case history, the O-cell load test
should be complemented and its results calibrated
by the execution of conventional topbottom load
tests (Fig. 19), as well as through analytical FEM
back analysis, allowing the optimization of the piles
design and execution methods.
In June 2012 all the foundations works, piles
and jet grouting, as well as the diaphragm walls

Figure 19. Comparison between the conventional and the


O-cell static load tests: Main advantages and limitations.

Figure 20.

View of the initial excavation works.

works, were concluded, allowing the beginning of


the excavation works (Fig. 20).
ACKNOWLEDGMENT
The authors thank to the owners of the described
case history their permission for the presentation
of this paper.

REFERENCES
Fellenius, B.H. 2001. The O-cellAn Innovative Engineering Tool. Geotechnical News Magazine. Vol. 19.
N2. 3233.
Pinto, A. & Pita, X. 2011. Deep Excavations in Luanda
City Centre. 15th African Regional Conference Conference on Soil Mechanics and Geotechnical Engineering,
Session 5Lateral Support and Retaining Structures:
269274. Maputo, Mozambique.
Schmertmann, J.H. & Hayes, J.A. 1997. The Osterberg
Cell and Bored Piled Testing. The Third Geotechnical
Engineering Conference, Cairo University, Egypt.

1778

COUNTINHO_Vol 2.indb 1778

8/13/2012 5:43:08 PM

You might also like