Está en la página 1de 5

High Strain and Low Strain Dynamic Pile Testing

at the New Blanchetown Bridge, SA


Jonathan Cannon
Director, Independent Geoscience Pty Ltd, Australia

This paper describes the use of dynamic pile testing during the construction of a new bridge over the Murray River at Blanchetown,
SA. The bridge carries the Sturt Highway, which is the main road link between Adelaide and Sydney. The bridge foundations
comprise groups of driven precast prestressed concrete piles and dynamic pile testing was used to establish driving criteria and to
confirm pile resistance. At Pier 1, on the Adelaide side of the river, the first piles did not stop at the expected founding level and
dynamic testing was used to investigate their behaviour. The testing indicated all piles were broken. A program of high strain and low
strain dynamic testing was then used to assess the piles and to investigate alternative methods for successfully driving piles at the
required locations. The paper describes the testing methods and an interpretation of the cause of breakages and the remedial methods
adopted.
1

INTRODUCTION

A new $15 million bridge has been constructed over the Murray
River at Blanchetown, SA. It carries the Sturt Highway, which
is the main road link between Adelaide and Sydney, and
consequently it carries a high proportion of heavily loaded truck
traffic. The client was the State highway authority, Transport
SA, and the Main Contractor on the project was York Civil
Construction Pty Ltd. Geotechnical advisor to Transport SA
was Rust-PPK Pty Ltd.
There is an existing bridge roughly parallel to and at a similar
level to the new bridge but defects have been found in the
structure and it was not designed for current highway loadings.
A decision was made to construct a completely new bridge.
The new bridge comprises 7 main spans of 50m and 2 shorter
end spans. Foundations for all but the west abutment comprise
groups of driven precast prestressed concrete piles, which are
octagonal and 457mm across. Lengths varied from 12 to 19m
overall and they were prestressed to about 10MPa. The pile toes
were pointed with a steel pin about 75mm diameter and about
100mm in length at the tip of the point. The main contractor
drove the piles with a recently re-conditioned Kobe K35 diesel
hammer fitted to a fixed leader assembly. For work over water
the crane and leader were supported on a segmental floating
barge, which was kept in position using spud piles together
with winches attached to fixtures on the riverbank.
The foundation conditions across most of the site were
similar and comprised the valley of the Murray River. There
were variable depths of soft recent sediments and fill over the
local soft rock, which is described in the boreholes as Marl and
appears to be a weakly cemented calcareous sand or silty sand.
On the west side of the river between the Adelaide abutment and
Pier 1, there was a reasonably steep cliff, in which interbedded
layers of weakly cemented and strongly cemented layers could be
seen in the Marl.
2

DYNAMIC PILE TESTING

Dynamic pile testing has been in commercial use for over 25


years and has been in use in Australia since 1978. Both high
strain and low strain methods are recognised in the Australian
Piling Code as methods of assessing pile foundations for
resistance (high strain only) and integrity.
Other methods
presently available within Australia for assessing pile foundations
include conventional static testing, a new static method using

Osterberg Cells, where the jack is placed near the toe of the
pile and a quasi-dynamic method. Dynamic testing has been
shown to be by far the quickest and most cost effective of these
methods and is the most frequently used method for verification
of pile foundations.
The dynamic method relies on the transmission of stress
waves along an elastic rod. When an impact is applied to the
top of a long slender elastic rod, ie a pile, the whole rod does
not move as a single rigid object but rather the top section is
compressed. This section then compresses the adjacent section,
which in turn compresses the next adjacent section etc etc and a
compression wave travels along the rod.
When the wave
reaches the toe a reflection is generated, which is controlled by
the resistance and damping of the ground at the toe.
The
behaviour has been described mathematically by the wave
equation:

( 2u / t 2 )

E( 2u / x 2 )

(1)

where u is the displacement of a point at time t and at a location x


in an elastic rod, and is the density and E is Youngs Modulus
for the material of the rod. Changes in pile section, or resistance
along the pile return reflections to the top of the pile.
As noted above, well-proven equipment and software is
available to conduct testing and analysis using dynamic methods.
High strain testing equipment requires the pile to be struck with a
full sized piling hammer or a drop weight with sufficient energy
to set the pile. This strains the ground and thus mobilises
resistance from the ground around the pile. Pile force and
velocity are measured near the top of the pile. As both force and
velocity are measured the downward, or input wave can be
separated from the upward, or reflected wave. As resistance is
mobilised in the ground surrounding the pile a capacity can be
demonstrated. If the pile can be struck with sufficient energy
then plastic movement of the ground resistance can be reached
and the ultimate capacity of the pile can be determined. If less
energy is applied and a small set is achieved then the method
can demonstrate a certain resistance but it will probably be less
than the ultimate capacity. Typically plastic movement of skin
friction is achieved at a set of about 3mm/bl. End bearing
requires much greater displacement and general geotechnical
theory suggests the toe displacement needs to reach 10% or more
of the pile diameter to reach plunging failure. Consequently it is

unlikely that dynamic testing would be able to demonstrate the


ultimate resistance of a pile with a high proportion of end
bearing. However, in most cases more than enough resistance
can be demonstrated and most structures placed on piles will not
tolerate very high deflections and so plunging failure of an endbearing pile is of esoteric interest only.
A photo of a current model PAK Pile Driving Analyzer is
shown in Figure 1 below. Readers should see reference 1 for a
description of the theories used by the method and reference 2 for
a description of the State of the Art in high strain dynamic pile
testing.

determination of a pile top stiffness. However, it should be noted


that this is only applicable at very low strains and may not be
indicative of the pile behaviour at high strains (ie at real
deflections) and as noted previously the method cannot determine
the ultimate resistance of the pile.
2.1

2.1.1

Damage Detection Theory

High Strain

The Author previously described this theory in reference 3.


Consider a pile with a reduction in area or modulus at some point.
Let Z represent the impedance EA/c with Z1 being the
impedance of the top section and Z2 being the impedance of the
bottom section. When a downward traveling stress wave Fi
arrives at this point part of the wave is reflected upward Fu and
part is transmitted across the section change and continues
downward Fd such that both continuity and equilibrium are
satisfied.
It has been shown in reference 4 that:

Fd

Fi (2Z 2 /( Z 2

Fu

Fi (( Z 2

(2)

Z1 ))

Z1 ) /( Z 2

Z1 ))

(3)

Thus for a uniform pile where Z2=Z1, Fu=0 and Fd =Fi and the
wave passes unchanged. At a complete break in the pile where
Z2=0, Fd=0 and Fu=-FI and the wave is completely reflected but
an initial downward traveling compression wave is reflected
upwards as a tension wave. Likewise, any decrease in area or
modulus will cause an upward tension reflection and any increase
in area or modulus will cause an upward compression reflection.
If we now introduce an integrity factor, being the proportion
of section change,
Figure 1 Pile Driving Analyser
Low strain dynamic testing requires the top of the pile to be
struck only with a hand held hammer and a small wave of high
acceleration but very low strain is generated. Pile top velocity is
monitored by a very sensitive accelerometer attached to the top
of the pile. The impact wave and the reflection are measured and
displayed on a screen for interpretation by a trained and
experienced operator. As the strains are very small the technique
cannot mobilise and demonstrate any significant resistance.

(4)

Z 2 / Z1
then it can be shown that

Fu

Fi (1

) /(1

(5)

And if we define

Fu / Fi

(6)

then

(1

) /(1

(7)

and there can be some quantification of the change in impedance


from top measurements. An arbitrary classification of damage
has been used in reference 5 as follows:
Severity

Figure 2 Pile Integrity Tester


Some models, such as the PIT Collector from Pile Dynamics Inc.,
as shown in Figure 2, include an instrumented hammer to
measure the impact force. The impact force can be compared
with the measured velocity to provide a better indication of
possible defects at the top of the pile. It also allows for the

1.0
Undamaged
0.8-1.0
Slight damage
0.6-0.8
Damage
<0.6
Broken
If there are cracks in a pile or slacks at mechanical pile
joints high strain testing allows the width to be quantified in an
approximate manner.
The quantification is based on the
displacement required to close the gap so that the downward
wave can continue. This is the integration of the relative velocity
change
caused
by
the
reflector.

t2
t1

(v

(8)

R
I

) dt

where t1 and t2 are the beginning and end of the crack or slack
and R is the difference between force and proportional velocity
at time t1. I would strongly suggest that this quantification only
be conducted after you are confident that the reflector is a crack
or gap at a splice location and not more significant damage.
Generally damage gets worse with further driving whereas gaps
at splices or a crack do not. More significant damage can start
with a crack.
2.1.1.1.1

Low Strain

Low strain testing involves very small impact waves. After the
initial impact force will be zero at the pile top as this is a
definition of a free pile end. The upward traveling wave
produces a difference between top force and velocity

2Fu

Ftop

Z1 Vtop

(9)

If we now express the waves in terms of velocities from (6)


above

Fu / Fi

Z i Vtop / 2Fi

next day and all 3 were considered by the Author to be severely


damaged about 4m below ground level. In only one case was it
possible to detect any response from the pile toe.
Following consideration of the ground conditions by the
Contractor, his Client and the Clients advisors another 4 piles
were driven with dynamic testing from the first blow. Senior
representatives of all the above parties inspected the driving.
Three of these piles were also damaged. Damage appeared to
commence when each of the piles reached a penetration of about
7 to 8m. There was no detectable change in driving of the piles
when the dynamic testing indicated the piles to be damaged. The
Contractor was reluctant to believe the piles were damaged and
deliberately drove one of the piles until the PDA indicated very
severe damage ( =35%). One of the other piles was considered
significantly damaged ( =79%) and one pile was considered by
the Author to be starting to show damage but the calculated
section change remained 100%.
The contractor remained skeptical and so employed a very
large excavator to dig around the piles and a crane was used to
pull the piles out for visual inspection. The visual inspection
showed the piles were damaged at the locations and to the extent
suggested by the dynamic testing. The following figures and
photographs show the correlation between dynamic testing and
the visual inspection.

1 / 2(Vtop / Vi )
(10)

This suggests that we should adopt of the velocity increase


caused by the impedance change to calculate . This may be true
near the pile top but if there has been considerable damping of
the signal along the shaft, which is common in practical
applications, then the return velocity Vtop is often amplified to
about the magnitude of Vi in order to identify the response. At
the toe of the pile we would expect =0 hence =1 and if the toe
response is amplified to be the same magnitude as Vi then the
factor should not be applied. Hence at any significant distance
down the pile the application of the is questionable.
2.1.1.2
BLANCHTOWN TESTING
Initial testing for the project was at pier 8 at the east (Waikeri)
end of the bridge and this was uneventful with all piles remaining
intact and demonstrating the required resistance.
A driving
criterion was determined for driving of other piles at the pier.
Testing started at Pier 1 after 3 piles had been driven. Pier 1
is immediately adjacent to a 20m cliff at the west (Adelaide) end
of the bridge as described earlier. Piles had been expected to
found at 7 to 10m penetrations but two 12m piles and one 19m
pile had been driven approximately to ground level and none had
shown the required resistance. Testing was requested by the
Contractor to find out what resistance the piles would provide.
The experienced piling foreman suspected damage but it was not
considered to be a strong possibility. Testing was conducted the

Figure 3 PDA screen of badly damage pile


Note the very significant reflection from damage shown in
negative wave up ie Fu. There is no clear response from the
pile toe. BTA=35.

Figure 4 Photo of badly damaged pile after removal

The photo shows that the badly damaged pile was completely
broken with exposed and bent reinforcing being the only
connection between the 2 sections of pile.

There is a reflection from damage just starting to occur


above the toe. The toe is still clearly detected. BTA=100.

Figure 8 Photo of slightly damaged pile

Figure 5 PDA screen of damaged pile


Note there is still a significant reflection from damage above the
toe shown in negative wave up ie Fu. The reflection from the
toe is still visible. BTA=79.

Figure 6 Photo of damaged pile after removal


The photo shows the end of the pile is permanently deflected.
Even the 10MPa prestress did not pull the pile straight again.
However the toe section is still firmly attached to the upper
section hence the toe reflection detected by the PDA. The toe
detail is also shown clearly in this photo.

The photo shows cracks that had been closed by prestress after
the pile was removed from the ground. The cracks were only on
one side of the pile. A low strain test was conducted on this pile
after it was removed from the ground, as shown in the photo, and
the low strain testing, which is very sensitive to even small
cracks, did not detect any problem with the pile.
The damage to all piles appeared to be linked to lateral
deflection of the pile toe. As the piles were driven adjacent to a
steeply sloping rock surface it is the Authors opinion that the
sloping surface also continued below the ground and as the piles
met the sloping surface the pointed toe was deflected sideways,
despite the steel pin projecting from the pointed toe. It appears
the designers also agreed that the damage was related to
deflection of the pile toe. Driving at pier 1 continued with the
original piles after each location had been pre-drilled.
Interestingly the pre-drill was only 100mm diameter. However, it
would appear that in almost all cases this was sufficient for the
pile to follow the pre-drilled hole and toe into the soft rock.
Low strain integrity testing was subsequently conducted on
all Pier 1 piles after driving was completed together with most
piles at Piers 2 and 3. At Piers 2 and 3 the rock surface was not
steeply dipping and all of the piles that were tested appeared to be
undamaged. One further broken pile was detected at Pier 1 and
this was one of the extra piles installed to replace broken piles.
The low strain test data for this pile is shown below.

Figure 9 PIT data replacement pile C


The data shows a strong reflection at 4.5m, a second reflection
from the same defect at 2x4.5=9m and a third reflection at
3x4.5m=13.5m. No response from the toe can be detected at
14m, which suggests severe damage.
Figure 7 PDA screen for slightly damaged pile

CONCLUSIONS
The use of pointed piles should be questioned when driving
to sloping rock surfaces. If the slope is steeper than the angle of
the point then the tip will not touch the rock and the pile will
certainly be deflected. When such conditions are expected a flat
toe with a projecting steel pin ie a Balkan Point will probably
provide a better chance of avoiding deflection of the pile toe.
The Author believes a flat toe is better in all circumstances.
Driving behavior does not necessarily provide any indication
of pile damage. It is suggested that all piles driven to a sloping
rock surface are checked for integrity using some form of
dynamic testing.
Dynamic testing has proved useful at this project in
demonstrating the required pile resistance, determining required
driving criteria, and proving pile integrity.
The correlations presented in this paper confirm that
interpretation of dynamic test results by an experienced operator
can correctly identify the location and approximate extent of pile
damage.
REFERENCES
1.

2.

3.

4.
5.

J Cannon (1990). Dynamic Pile Testing for Marine


Construction. Proceedings of Third Australian Ports
and Harbours Conference, Melbourne Australia,
August Institn Engs Aust pp 66-72
G. Goble and G Likins (1996). On the Application of
PDA Dynamic Pile Testing. Proceedings of Fifth
International Conference on the Application of Stress
Wave Theory to Piles.
Orlando, Florida USA,
September Townsend, Hussein, McVay Eds pp 263273
J Cannon (1990). Integrity Testing of Piled
Foundations Proceedings of 2nd National Structural
Engineering Conference, Adelaide Australia, October
Institn of Engs Aust pp 450-455
F Rausche and GG Goble (1986). Determination of
Pile Damage by Top Measurements Special Technical
Publication 670 ASTM pp500-506
PDA-W Manual of Operation (2001) Pile Dynamics
Inc.

También podría gustarte