University
stringclasses 19
values | Text
stringlengths 458
20.7k
|
|---|---|
ADE | Chapter 3
Pressure response
MOC (Wylie and Streeter 1993; Chaudhry 2014) is used to simulate the
transient response of the simple pipe network system. Steady friction is
considered to evaluate its impact on the leak detection. The time step used is
0.0002 s. Two transient tests are simulated using the generator G and G ,
1 2
respectively. Considering the complexity of the network, the excitation signal
used in both tests is a special type of pseudo-random binary signal (PRBS) –
the inverse repeat signal (IRS) instead of discrete pulse or step signals. The IRS
is a periodic signal that is suitable for extracting the pipeline frequency response
(Gong et al. 2013b), and it can be generated by continuously altering the
opening area of a side-discharge valve between two levels (Gong et al. 2016b).
The IRS signal used in this study is the same as that described in Gong et al.
(2013b) (simulating 10 shift registers with a clock frequency of 100 Hz), and
has a period of 20.46 s. Each numerical test has a simulated time duration of 20
mins, which is over 58 periods of the IRS. Spectrum analysis confirms that the
pipe system reaches the steady-oscillatory condition after 200 s (about 10
periods). A section of the pressure traces at T as obtained from the first test is
A
shown in Figure 3.8. Due to the pseudo-random nature of the excitation signal,
the pressure response of the pipe system is complex and difficult to analysis
directly in the time domain.
96 |
ADE | Chapter 3
Leak detection
Leak detection is conducted by analysing the numerically obtained
ImU U using the technique outline in Equations (3.25) and (3.26), and
22 11
the results are shown in Figure 3.10. The distinctive pike indicates that there is
one leak in the pipe section of interest. The normalized locations are determined
as x = 0.20 from the x-axis, and the value of the impedance ratio is Z Z =
L c L
0.0442 according to the size of the spike. The results are highly consistent with
the theoretical values as shown in Table 3.2. The successful detection has once
again validated the effectiveness of the proposed targeted leak detection
technique.
3.5 Discussion
3.5.1 Effect of friction
Friction is neglected in the proposed leak detection algorithm. The effect of
friction on the frequency response of pipeline systems has been investigated in
detail by Lee et al. (2005b) for leak detection in R-P-V systems. It has been
found that the effect of steady-friction is minor and approximately uniform
across all the frequencies, therefore it should not affect the period of the
sinusoidal waves in the T signal or the localization of the leak. The impact of
N
steady friction on the amplitude of the sinusoidal waves is very limited for real
water transmission pipelines, therefore the impact on the leak impedance/size
determination is limited. The above has been confirmed by the two numerical
case studies conducted in the current research, in which the locations of the
99 |
ADE | Chapter 3
leaks are accurately determined despite that steady friction is included in the
numerical simulations.
The unsteady friction, however, will induce a non-uniform dampening for the
frequency responses, and a correction technique has been proposed in Lee et al.
(2006). Recent research on the unsteady friction in water pipelines concludes
that the effect of unsteady friction in large diameter water transmission
pipelines is limited (Vardy et al. 2015), and it has been generally neglected in
practice (Shucksmith et al. 2012; Meniconi et al. 2013; Stephens et al. 2013). If
the pipe section of interest (the section in bracket of the two pairs of transducers)
is relatively long such that the fundamental frequency is low, the excitation and
the analysis only need to focus on the low frequencies [e.g. in Case 1, the
periodic nature of ImU U is already clear in the range of 0 to 30 Hz
22 11
Figure(3.6)]. In the low frequency range, the effect of unsteady friction is
limited and less non-uniform.
3.5.2 Challenges in field application
Challenges are expected in real application of the proposed leak detection
technique. Although the two-source-four-sensor testing configuration for water
pipe transfer matrix extraction has been validated in the laboratory (Yamamoto
et al. 2015), the implementation of this testing configuration in real pipe
systems can be difficult. Transient generators (source) can be installed on
existing access points such as fire hydrants or air valves. The transducers need
to be installed in pairs and the distance between the two sensors in a pair needs
to be short to enable the analysis. This is challenging since in real water
100 |
ADE | Chapter 3
pipelines it is uncommon to have two accessible points in close proximity.
Although live tapping can be done to create new fittings for the transducers, it
is undesirable or expensive especially when the pipe is buried. Resent research
on fibre optic pressure sensor arrays (Gong et al. 2018b) may provide a solution
in the future. It is envisaged that a fibre optic pressure sensor array, as in the
form of a flexible cable, can be inserted into a pipeline through a single access
point. The same access point can also be used for transient wave generation.
The fibre optic pressure sensors measure the transient response of the pipe
system. The same configuration can be repeated at another access point to
achieve the two-source-four-sensor testing configuration. Preliminary success
has been achieved in the laboratory (Gong et al. 2018b), in which leak
reflections are identifiable from the pressure measurement; however, several
design challenges need to be resolved to enhance the accuracy and robustness
of the measurements.
Another challenge is the effectiveness (e.g. bandwidth and tolerance to noise)
of the excitation transient wave. Pressure measurements in real pipeline systems
will suffer from noise and transient interference (e.g. generated from water
users). It is expected that conventional discrete transient excitation (e.g. pulse
and step waves) will not be effective, and the PRBS is needed to achieve
accurate extraction of the transfer matrix. Averaging the results from multiple
periods of the steady-oscillatory response will reduce the effect of noise. A side-
discharge valve based PRBS transient generator has been developed and tested
in the laboratory (Gong et al. 2016b); however, field applications may require
a larger and more powerful transient wave generator. The challenge is how to
101 |
ADE | Chapter 3
maintain the fast response of the valve for a consistent and wide bandwidth of
the excitation.
The structure complexity of ageing pipelines can be another challenge. The
proposed technique is a significant step forward to tackle complex pipe systems,
and it enables a targeted pipe section to be visually isolated for independent
analysis in any complex network. However, within the targeted pipe section,
the condition of the pipe can still be complex, with the presence of not only
leaks but non uniform pipe wall deterioration. Duan et al. (2011) demonstrated
that FRF-based leak detection is applicable to complex series pipelines. Further
research is needed to investigate the impact of pipe wall deterioration or other
defects (e.g. blockages) in the targeted pipe section on leak detection.
3.6 Conclusions
A new pipeline leak detection technique has been proposed in this research. The
technique enables leak detection for a targeted pipe section independent from
the complexities of the pipe system where the targeted section is embedded in.
This is achieved by extracting the transfer matrix of the targeted pipe section
using a two-source-four-sensor hydraulic transient testing strategy, and
analysing the resultant transfer matrix by a newly developed algorithm. The
proposed technique has been validated by two numerical case studies. In the
first study (Case 1), the two leaks in a pipe section embedded in a reservoir-
pipeline-valve system have been successfully determined using pulse excitation
waves. In the second study (Case 2), the location and impedance of the leak in
a pipe section embedded in a simple pipe network are successfully determined
using pseudo-random binary signals as the excitation.
102 |
ADE | Chapter 4
Abstract
The use of two pressure transducers in close proximity can enable the separation
of the directional travelling pressure waves in pipelines. However, the
implementation of this measurement strategy in real water pipes is difficult due
to the lack of closely located access points. This paper reports the use of a
customised in-pipe fibre optic pressure sensor array for hydraulic transient
wave separation and pipeline condition assessment. The fibre optic pressure
sensor array can be inserted into a pressurised pipeline through a single access
point. The array consists of multiple fibre Bragg grating (FBG)-based pressure
sensors in close proximity (~0.5 m apart). A previously developed wave
separation algorithm is adapted to analyse the transient pressure measurement
from the FBG sensors. The resultant directional pressure waves are then used
to detect pipe sections with a thinner wall thickness. A challenge is the influence
of the in-pipe fibre optic sensing cable on the transient pressure measurement.
The impact is analysed and adjustments to the pipeline condition assessment
algorithm are undertaken to resolve the issue. The successful experimental
application verifies the usefulness of the in-pipe fibre optic sensor array, which
can facilitate transient-based pipeline condition assessment for buried water
pipes with limited access points.
109 |
ADE | Chapter 4
4.1 Introduction
Water utilities globally are facing the problem of ageing water distribution
systems (WDSs), and the cost of maintenance and replacement is predicted to
explode under current practice. For instance, it is estimated that more than US$1
trillion will be required between 2011 to 2035 to replace ageing water mains and
address projected growth (American Water Works Association 2012). Pipeline
condition assessment has becoming increasingly important, because the actual
condition of pipelines can help strategically prioritise investment and extend
asset life.
Among many pipeline condition assessment techniques available, hydraulic
transient-based methods are particularly attractive because they can achieve
continuous pipe wall condition assessment for hundreds of metres up to
kilometres of pipe in a single test (Stephens et al. 2013). The approach uses
small controlled hydraulic transient pressure waves, which travel at about 1200
m/s in pressurised metallic water pipes, will induce wave reflections at pipe
cross-sections with physical changes (e.g. leaks, blockages and wall thinning
due to corrosion), and the reflections can be interpreted by appropriate
algorithms to reveal the nature of the anomaly (Chaudhry 2014). In addition to
pipe wall condition assessment (Zeng et al. 2018b; Zhang et al. 2018a), many
transient-based techniques have been developed for the detection of leaks
(Brunone and Ferrante 2001; Covas et al. 2005; Shamloo and Haghighi 2009;
Soares et al. 2010; Gong et al. 2013a; Duan 2016a), blockages (Sattar et al.
2008; Meniconi et al. 2013; Massari et al. 2014), illegal branches (Meniconi et
111 |
ADE | Chapter 4
al. 2011a), general anomaly screening (Meniconi et al. 2015) or system
parameter identification (Zecchin et al. 2014a).
In operational pipe systems, pressure sensors are installed at existing access
points (Ghazali et al. 2012; Gong et al. 2015), such as air valves or fire hydrants,
which are typically hundreds of meters apart from each other, to avoid
excavation or tapping. For any interior point along a pipe, the measurement
from a single pressure sensor is always the superimposed amplitude of the
pressure waves travelling upstream and downstream. As a result, it is difficult
to tell whether a measured pressure reflection is from the upstream or
downstream side of the transducer, or actually a combination of waves from
both sides. The use of measurements from multiple access points (hundreds of
metres apart) and a time-shifting technique (Gong et al. 2016c) is helpful in
providing the directional information, but only when the reflected waves are
simple in wave form and limited in number (i.e. the pipeline configuration and
condition are not complex).
A wave separation technique recently developed for hydraulic transient
pressure waves in pipelines has provided a robust solution to obtain the
directional information of wave reflections (Shi et al. 2017). The wave
separation technique uses two pressure sensors in close proximity along a pipe
(in the scale of metres), and they are referred to as a “dual-sensor”. It was
adapted to water pipe systems from the original technique for separating
directional acoustic waves using multiple microphones (Chung and Blaser
1980). In the wave separation technique, the time-domain pressure reflection
signals as measured by the dual-sensor are transformed into the frequency
112 |
ADE | Chapter 4
domain, processed based on the fact that a directional wave arrives at the two
sensors at different times but with a specific time delay, and then transformed
back into the time domain. The results are two directional pressure waves with
significantly reduced complexity, in which any major wave reflection can be
attributed to its source by the principle of time-domain reflectometry. One
practical challenge of this wave separation technique in buried water pipelines,
however, is the difficulty in achieving the dual-sensor measurement
configuration.
To address this challenge, the authors have developed in-pipe fibre optic
transient pressure sensors, with the first generation tested in the laboratory for
proof-of-concept (Shi et al. 2015) and the second generation tested and reported
in this paper for wave separation and pipe condition assessment. The optic
pressure sensors are based on fibre Bragg gratings (FBGs), and multiple sensors
are placed in close proximity (~0.5 m) in a protective cable. The sensor cable,
with a diameter of ~4 mm, can be inserted into a pipeline through a single access
point. Laboratory experiments are conducted in a single copper pipeline with
two short sections in thinner wall thicknesses, and the wave separation
technique is applied to the measured transient pressure data for pipe wall
condition assessment purpose. One particular challenge is the impact of the
sensor cable on the transient response. The sensor design, the impact of the
sensor cable, the wave separation and the application to pipe wall condition
assessment are discussed in the following sections.
113 |
ADE | Chapter 4
4.2 In-pipe fibre optic transient pressure sensor
array
The in-pipe fibre optic sensor array used in this research includes five FBG-
based pressure sensors (FBG1 to FBG5) in a 5.37 m long cable, the schematic
of which is given in Figure 4.1. It is a further development based on the FBG
manometry catheter developed by Arkwright et al. (2012) for measuring
muscular activity in the human gut. The distance between FBG1 and FBG2 is
0.725 m and that for the rest is 0.5 m. The cable that protects the optical fibre
is made from plastic material, and has a diameter of approximately 4 mm. At
each FBG pressure sensor, a 10 mm window is open in the protective cable, and
a flexible elastomeric sleeve is used to cover the FBG, as illustrated in Figure
4.2. The FBG is designed to have a downward arc under atmosphere pressure
and the flexible sleeve is in close contact with the sensor. As the pressure
increases from atmosphere pressure, the sleeve presses the FBG further
downwards, which causes a change in the strain and in turn a shift in the
reflected wavelength of the FBG. This configuration, in particular the arc-
shaped pre-load, is a new design compared with its predecessor reported in Shi
et al. (2015), and it enables high sensitivity to pressure variations under high
background pressure condition (as is the case in pressurised water pipes). This
in-pipe fibre optic sensor array has recently been applied to measure leak-
induced hydraulic noise in the steady state and wave reflections under transient
events (Gong et al. 2018b).
114 |
ADE | Chapter 4
Table 4.1 Physical details of the pipeline system used in the laboratory
experiments.
Wave speed Impedance
Pipe Internal Wall thickness
class diameter (mm) (mm) (m/s) (s/m2)
A D e a B
= 22.14 =1.63 = 1,319 = 349,000
A A A A
B D e a B
= 22.96 =1.22 = 1,273 = 314,000
B B B B
C D e a B
= 23.58 = 0.91 = 1,217 = 284,000
C C C C
Two conventional pressure transducers (T1 and T2 in Figure 3, M5HB, Keller
AG, Switzerland) were flush mounted on the pipe wall through small brass
blocks encasing the pipe. A solenoid-controlled side-discharge valve was used
as the transient wave generator (G), and it was installed at the same cross-
section of the pipe where T1 was located. The solenoid valve was installed on
the top of the pipe for water discharge and the release of any trapped air. The
transducers were installed on the side and with an upward angle to prevent any
air from being trapped at the sensor head. The fibre optic sensor cable was
inserted into the pipeline through an angled tapping point and sealed with an O-
ring gland. The insertion point is 1.05 m away from the transient generator G
and transducer T1. The optical fibre was illuminated using a super-luminescent
light emitting diode (DL-BP1-1501A, DenseLight, Singapore) and the reflected
wavelengths from the sensor array were monitored using a solid-state
spectrometer (I-MON 512 HS, Ibsen Photonics, Denmark). Transducer T2 and
FBG3 were placed at the same location, and as a result, the overall length of the
117 |
ADE | Chapter 4
fibre optic cable inside the pipeline is 3.1 m. The sampling rate for the FBG
sensors was 12.376 kHz and that for the conventional sensors was 20 kHz.
4.3.2 Pressure measurements and simulations
In the initial steady-state condition, the pipeline system was pressurised at 3 bar
by the pressurised tank, and the solenoid valve was fully open to discharge
water. The solenoid valve was then abruptly closed, which resulted in two
identical incident step waves propagating along the pipe in two directions.
The pressure measurements from T2 and FBG3 are shown in Figure 4.4,
together with the measurement from T2 when no fibre optic cable was present,
and with the numerical pressure response at T2 obtained by the method of
characteristics (MOC). In the MOC simulation, a pipe model is established
based on the information in Figure 4.3 and Table 4.1 for the scenario that the
fibre optic sensor cable is present. The wave speed in the pipe section where
the fibre cable is enclosed is 1,230 m/s as determined from the laboratory
measurements, and this is used in the numerical model. The time step used in
the simulation is 0.5 ms. Steady friction with a Darcy-Weisbach frictional factor
of 0.02 is considered.
The incident wave and the major wave reflections from key features are
highlighted in Figure 4.4. According to the principle of time-domain
reflectometry (TDR), once a wave front encounters a physical discontinuity (e.g.
a wall thickness change), reflections occur and propagate backwards.
Reflections from features closer to the generator will arrive at the transducer (at
the same or close location of the generator) sooner. The short duration of the
118 |
ADE | Chapter 4
small change in the pipe cross-sectional area (the cross-sectional area of cable
is about 3% that of the pipe, but the cable area may be compressed to be even
smaller under pressure), the impedance of the pipe section with the in-pipe fibre
optic cable is calculated as 337,000 s/m2, which is smaller than the impedance
of normal Class A pipe. The pressure signal seems complex in the original
measurement as shown in Figure 3.4 but will become clearer after the wave
separation as discussed later.
Combined pressure wave reflections from the Class B and C sections are also
recorded in the traces in Figure 4.4. The laboratory measurements from both T2
and FBG3 when the in-pipe cable was present are slightly smoother than that
from T2 when no cable was in the pipe. The effect of signal smoothness is most
likely due to the viscoelasticity of the plastic material.
Overall, those three experimental traces are generally consistent, and they are
also consistent with the numerical results. Since the length of the cable inside
the pipe is short (3.1 m), the pressure oscillations induced by the cable is
confined in a very short time period (~5 ms) after the generation of the incident
wave and the effect of signal smoothness is insignificant.
4.4 Wave Separation
4.4.1 Directional pressure waves
The pressure measurements from FBG2 and FBG3, as shown in Figure 4.5, are
used for wave separation. They are close to the transient generator and the
results can be compared with those obtained from the conventional sensors (T1
121 |
ADE | Chapter 4
with the numerical results. The wave separation decomposes the superimposed
raw pressure measurement into directional pressure waves, simplifies the
complexity of the signal, and enables better understanding of the source of
reflections. Discrepancies exist and are discussed in the flowing sub-section.
4.4.2 Discussion
The experimentally determined directional waves using the FRB sensors and
the conventional sensors are highly consistent in the wave form. For the wave
reflections from the upstream side of the dual-sensor and propagating towards
the dead-end [Figure 4.6(a)], the experimental results are also highly consistent
with the numerical results. The first step rise is the reflection from the upstream
boundary of the in-pipe cable, and it is because that the pipe section (Class A)
with the cable has a lower impedance than that in normal Class A pipe sections.
The cable is made from plastic material, which is much lower in strength
compared to the material of the pipe wall (copper), and it results in a lower
wave speed and therefore a lower impedance for the pipe section hosting the
cable. When a positive wave propagating from a lower impedance pipe section
to a higher impedance one, a positive pressure wave reflection will occur (Wylie
1983). The following step drop and then step rise are reflections from the Class
C pipe section, which has an impedance lower than the Class A pipe. Detailed
explanation of the wave reflection mechanism from a short pipe section with a
different impedance can be found in Gong et al. (2013c). The experimental
results have some small oscillations in addition to the major reflections. The
oscillations can be resulted from the transient interference induced by the
vibration of the solenoid valve at the sudden closure, and the uncertainty in the
125 |
ADE | Chapter 4
empirical determination of the transfer function. The numerical results are
smooth and clearly show the expected reflections, which confirms the
effectiveness and accuracy of the wave separation algorithm itself.
For the wave reflections from the downstream side of the dual-sensor and
propagating towards the tank [Figure 4.6(b)], the reflections from the Class B
section (as highlighted in the figure) are generally consistent for the three sets
of results; however, discrepancies are observed between the experimental and
the numerical results for the reflections from the downstream boundary of the
in-pipe cable (which was the cable insertion point). The experimental results
show a negative reflection followed by a positive reflection and some wave
oscillations, while the numerical results show a step response similar to the
upstream-boundary reflection as seen in Figure 4.6(a). The insertion tapping
point (an angled conduit in a brass block) was sealed by an O-ring gland with a
“finger-tight” condition. The conduit and the O-ring seal are likely to respond
to pressure transients like a small accumulator, thus producing the signature of
a negative reflection followed by a positive reflection as observed in the
experimental results. The numerical model did not include this complexity,
therefore the numerical result only shows the step reflection as induced by an
impedance change at the downstream-boundary of the in-pipe cable.
4.5 Pipe wall condition assessment
4.5.1 Methodology
The original direct-reflection-analysis-based condition assessment technique
(Gong et al. 2013c) needs to be further developed for the directional waves
126 |
ADE | Chapter 4
obtained from the in-pipe fibre optic sensor array. The original condition
assessment algorithm assumes that the incident wave is generated and the
pressure responses are measured in an intact pipe section, and the deteriorated
sections are limited in number and mild in deterioration. In the case of using
the in-pipe fibre optic sensor cable, the generation of the incident wave and the
pressure measurement are undertaken in the section that encloses the cable, and
the impedance of this section is considerably lower than that in normal intact
pipe sections. However, the same principle still applies. That is, when a pressure
wave propagates from the ith pipe section to the (i+1)th pipe section where the
impedance changes, the sign-sensitive amplitude of the normalised wave
reflection (equivalent to the reflection coefficient) and that of the normalised
transmitted wave (equivalent to the transmission coefficient) are related to the
impedance of the two sections (Wylie 1983; Gong et al. 2013c), as given in
Equations (4.2) and (4.3), respectively.
B B
R i1 i (4.2)
i,i1 B B
i1 i
2B
T i1 1R (4.3)
i,i1 B B i,i1
i1 i
where R and T are the reflection and transmission coefficient,
i,i1 i,i1
respectively, for a wave propagating from the ith to the (i+1)th pipe section.
The relationships shown in Equations (4.2) and (4.3) are the same as the
reflection and transmission coefficients defined in acoustic reflectometry
(Sharp 1996), where typically acoustic waves propagating in air and in a wave
127 |
ADE | Chapter 4
guide are considered. Considering an incident wave generated at section 1 and
only propagating towards one direction (as in the case of directional waves as
previously discussed), the normalised initial reflection from the nth section as
would be measured in the 1st section can be derived as
n2
R R (1R2 ) (4.4)
1,n n1,n i,i1
i1
4.5.2 Application and verification
To verify the modified condition assessment algorithm as discussed in the
previous section, Equation (4.4) is applied to the directional pressure wave
coming from the upstream side of the dual-sensor and propagating towards the
dead-end [Figure 4.6(a)]. Rearranging Equation (4.2), the impedance of the
(i+1)th section can be calculated by
1R
B B i,i1 (4.5)
i1 i1R
i,i1
The pipe section with the FBG sensor cable inside can be considered as the 1st
section, the normal Class A section on the upstream side is the 2nd section, and
R R
the Class C section is the 3rd section. The reflection coefficients and
1,2 1,3
can be determined from the sign-sensitive amplitude of the major wave
reflections in the directional wave shown in Figure 4.6(a). The reflection
R
coefficient can then be calculated by Equation (4.4). Finally Equation (4.5)
2,3
B
can be used to calculate , which, in this case, is the impedance of the Class
3
128 |
ADE | Chapter 4
B
C section . Table 4.2 summarises the results. Note that, although three
C
significant figures are used for the values, the reflection coefficients (i.e. the
amplitude of the normalised reflections) are determined manually and
uncertainties are involved.
Table 4.2 Pipe impedance determined from the directional pressure wave
coming from the upstream side of the dual-sensor and propagating towards the
dead-end [Figure 4.6(a)].
Reflection Reflection Reflection Determined
coefficient coefficient coefficient impedance
Cases
R
1,2
R
1,3
R
2,3
B
C
(s/m2)
Numerical 0.040 –0.10 –0.10 285,000
FBG 0.0588 –0.0962 –0.0965 287,000
Conventional 0.0497 –0.110 –0.110 280,000
It can be seen from Table 4.2 that the determined impedance values for the
B
Class C section are consistent with the calculated theoretical value of as
C
given in Table 4.1 (284,000 s/m2), which verifies that pipeline condition
assessment can be conducted using the in-pipe fibre optic sensors with the
R R
methodology presented. The values of are very close to those of ,
2,3 1,3
which demonstrates that the local impedance change induced by the in-pipe
sensor cable has an insignificant impact on the pipeline condition assessment.
The discrepancies between the determined impedance values and the theoretical
value are mainly due to the error associated with the wave separation,
129 |
ADE | Chapter 4
uncertainties in the determined amplitude of wave reflections and uncertainties
in the pipeline parameters used to calculate the theoretical value.
4.6 Conclusions
A customised in-pipe fibre optic pressure sensor array has been used in the
laboratory for hydraulic transient wave separation and pipeline condition
assessment. The in-pipe fibre optic sensor array consists of five FBG-based
pressure sensors in close proximity. The optic fibre is protected by a plastic
cable with a diameter ~4 mm and the cable can be inserted into a pressurised
pipeline through a single tapping point. With empirical calibration of the
transfer function of the short pipe section between two sensors, a previously
developed wave separation technique is successfully implemented on the
transient pressure data measured from the fibre optic sensors, and the resultant
directional waves are consistent with those obtained from conventional pressure
sensors.
The impact of the in-pipe fibre optic sensor cable to the transient pressure
response of the pipeline system has been assessed and discussed based on the
directional pressure waves. The in-pipe sensor cable, as made from a plastic
material, reduces the local pipe impedance and therefore introduces wave
reflections. It also slightly smoothen the transient pressure signal. The entrance
point of the cable acts like a small accumulator and introduces pressure
oscillations. However, overall the impact of the in-pipe cable is moderate
because of its short length, and it does not impede the application of transient
pressure measurement for pipeline condition assessment.
130 |
ADE | Chapter 5
(2) A new transient pressure generation and measurement strategy has been
proposed for the purpose of condition assessment of targeted pipe sections
embedded in complex pipe systems (Chapter 3). The strategy, termed as the
two-generator-four-sensor strategy, uses two dual-sensor units to bracket a pipe
section to be analysed, and the two transient pressure generators to bracket the
two dual-sensor units (and the pipe section in-between). The configuration
enables an extraction of the transfer matrix of the in-bracket pipe section
through the virtual isolation. The characterisation of the transfer matrix
properties of this section (upon which the following leak detection methods are
based) would remain unresolvable without the two-generator-four-sensor
strategy.
(3) A new pipeline leak detection algorithm has been developed based on the
analysis of the transfer function of a pipe section “virtually” isolated by the two-
generator-four-sensor configuration, as explained above. It has been found that
the imaginary part of the difference between two elements in the transfer matrix
is sensitive to leaks. The result should be zero if no leak is present, while a leak
will introduce a sinusoidal pattern on this imaginary part of the transfer function.
The period and the magnitude of the pattern are related to the location and
impedance of the leak, respectively. The algorithm determines the location and
size of the leak based on this information, and is applicable to the analysis of
multiple leaks.
(4) A technique has been developed for realising distributed pressure
measurement along a pipe through only a single access point (Chapter 4). This
is achieved by a customised in-pipe fibre optic pressure sensor array. The fibre
134 |
ADE | Chapter 5
optic array is encased in a flexible cable with a diameter of 4 mm. The sensor
cable can be inserted into a pipeline through a small opening on the pipe wall.
The optical fibre based pressure sensors can measure transient pressure
fluctuations at a high sampling rate (up to 20 kHz) and a high background
pressure (up to 10 bar). The in-pipe fibre optic sensor array has been tested in
the laboratory for pipe wall condition assessment. It has found that the presence
of the in-pipe sensor cable, as made from a plastic material, reduces the local
pipe impedance and therefore introduces wave reflections. It also slightly
smooths the transient pressure signal, as the high frequency components
dissipate at a faster rate. However, the overall the impact of the in-pipe cable is
moderate because of its short length, and it does not impede the application of
transient pressure measurement for pipeline condition assessment. A TDR-
based pipeline condition assessment algorithm has been further developed to
incorporate the impact of the local impedance change induced by the in-pipe
cable.
5.2 Research contributions
The key contributions of the aforementioned research outcomes have been
summarised as follows:
(1) The wave separation reduces the complexities associated with wave
superposition and provides the directional information of the measured wave
reflections. This unprecedented directional information creates opportunities to
develop more advanced pipeline leak detection and condition assessment
techniques. The proposed technique provides a model free wave separation
135 |
ADE | Chapter 5
approach, thereby overcoming past limitations associated with parametric
uncertainties in model-based approaches.
(2) The two-generator-four-sensor configuration, combined with custom
developed signal processing algorithms, can “virtually” break any complex
pipeline systems down to its simplest form – a single pipe section. The extracted
transfer matrix is a full representation of the characteristics of the “virtually”
isolated pipe section, and is independent from any complexities of the rest of
the pipe system (e.g. boundary conditions and other network connectivity). As
a result, the extracted transfer matrix is much simpler than the transfer matrix
of the overall pipe system, and the analysis is more straightforward, and
ameanable to analytic investigation. By focusing on the isolation of the
dynamics of single pipes through the use of the two-generator-four-sensor
configuration, the proposed approach represents an significant departure from
the conventional research approach of gradually adapting the transient-based
techniques developed for single pipeline systems (e.g. reservoir-pipeline-valve
or reservoir-pipeline-reservoir systems) to more complex pipe systems and
networks.
(3) The new leak detection technique, as combined with the two-generator-
four-sensor configuration, enables the detection and location of multiple leaks
in targeted pipe sections embedded in complex pipe systems (including pipe
networks). This research represents a significant step towards the application
of hydraulic transient-based leak detection techniques in real water distribution
systems. It is envisaged that other defects, such as blockages and extended pipe
136 |
ADE | Chapter 5
wall deterioration, can also be detected using the extracted transfer matrix of
the “virtually” isolated pipe section.
(4) The in-pipe sensor cable provides the ability to have multiple pressure
measurements through one access point. This is important since most water
pipes in the field are buried underground, and access is only obtainable through
sparsely available existing hydraulic devices such as air valves and fire hydrants.
Successful laboratory verification has proven the concept and provided useful
information for future developments.
5.3 Future work
Specific topics for future work have been identified based on the findings of
this PhD research, these include:
(1) To further develop the wave separation technique for persistent transient
excitation.
Recent research has shown that persistent transient excitation, such as pseudo
random binary sequences, can enhance the robustness and accuracy of system
identification, either in the time domain (Nguyen et al. 2018) or in the frequency
domain (Gong et al. 2016b). The current generation wave separation technique
has been developed for and validated by discrete transient excitation only,
including pulse and step waves.
137 |
ADE | Chapter 5
(2) To conduct experimental verification of the two-generator-four-sensor
configuration and the associated leak detection technique for targeted pipe
sections in complex pipe systems.
Extensive numerical simulations have been conducted to validate the two-
generator-four-sensor-based transfer matrix extraction and the associated leak
detection for the “virtually” isolated pipe sections. Experimental studies are
needed to further enhance the practicality of the techniques.
(3) To develop next generation in-pipe fibre optic pressure sensor cables for
field applications.
This research validated the in-pipe fibre optic pressure sensor cable in a single
copper pipeline with 25 mm diameter in the laboratory only. It is envisaged that
the distributed pressure measurement will be equally effective in larger field
pipes, but challenges exist in the insertion of the sensor cable. Field water pipes,
as buried underground, are typically only accessible through a stand pipe with
a length about one metre and perpendicular to the main pipe. To avoid
disruptions to service, the insertion needs to be conducted in the normal system
operating condition and against the back pressure, which is typically in the
range of 3 to 8 bars.
138 |
ADE | The Impact of Rolling Dynamic Compaction
Abstract
Rolling dynamic compaction (RDC) consists of a non-circular module of 3, 4 or 5 sides,
that rotates as it is towed, causing it to fall to the ground and compact it dynamically.
There is currently little guidance available for geotechnical practitioners regarding the
depths of improvement that are possible in varying soil conditions. Current practice
dictates that practitioners rely on personal experiences or available published project
case studies that are limited in scope and applicability as they are typically aimed at
achieving a project specification. There is a reluctance to adopt RDC as a ground
improvement technique as there is uncertainty regarding its limitations and capabilities.
The underlying objective of this research is to quantify the ground response of the
8-tonne 4-sided impact roller. This research has used full-scale field trials and bespoke
instrumentation to capture the ground response due to dynamic loading in homogeneous
soil conditions. It was found that towing speed quantifiably influenced the energy
imparted into the ground, with towing speeds of 10-12 km/h found to be optimal.
Targeted full-scale field trials were undertaken to quantify the depth of improvement
that can be achieved using RDC. Field results were compared to a number of published
case studies that have used the 8-tonne 4-sided roller. Significantly, separate equations
have been developed to allow practitioners to predict the depths that can be improved
for the two major applications of RDC: improving ground in situ and compacting soil in
thick layers.
Finally, the in-ground response of RDC was measured using buried earth pressure cells
(EPCs) and accelerometers. Force was determined from the measured change in stress
recorded by EPCs whereas displacement was inferred from the double integration of
acceleration-time data to give real-time load-displacement behaviour resulting from a
single impact. The energy delivered to the soil by RDC is quantified in terms of the
work done, defined as the area under the force versus displacement curve.
Quantifying the energy imparted into the ground in terms of the work done is a key
difference from past studies. Previous estimates of the energy delivered by impact roller
at the ground surface has traditionally been predicted based on either gravitational
potential energy (12 kJ) or kinetic energy (30 kJ to 54 kJ for typical towing speeds of
9 to 12 km/h). The two different values have caused confusion amongst practitioners.
vii |
ADE | The Impact of Rolling Dynamic Compaction
Statement of Originality
I certify that this work contains no material which has been accepted for the award of
any other degree or diploma in my name, in any university or other tertiary institution
and, to the best of my knowledge and belief, contains no material previously published
or written by another person, except where due reference has been made in the text. In
addition, I certify that no part of this work will, in the future, be used in a submission in
my name, for any other degree or diploma in any university or other tertiary institution
without the prior approval of the University of Adelaide and where applicable, any
partner institution responsible for the joint-award of this degree.
I acknowledge that copyright of published works contained within this thesis resides
with the copyright holder(s) of those works.
I also give permission for the digital version of my thesis to be made available on the
web, via the University’s digital research repository, the Library Search and also
through web search engines, unless permission has been granted by the University to
restrict access for a period of time.
I acknowledge the support that I have received for my research through the provision of
an Australian Government Research Training Program Scholarship.
ix |
ADE | The Impact of Rolling Dynamic Compaction
Acknowledgements
My sincerest thanks go to my primary supervisor, Professor Mark Jaksa from The
University of Adelaide for his ongoing support, advice, friendship, guidance and
encouragement throughout my research. I would like to thank my co-supervisor, Dr.
Peter Mitchell from Aurecon Australia for his advice and support throughout my
candidature. The assistance of my former supervisors, Dr. William Kaggwa (now
retired but formerly of The University of Adelaide) and Mr. Derek Avalle (now
working at Broons after formerly working at Keller Ground Engineering and Broons) is
also greatly appreciated.
I wish to thank the School of Civil, Environmental and Mining Engineering for their
support and patience during my candidature. I wish to acknowledge the support given to
me by the School’s laboratory and instrumentation staff (past and present), in particular:
Mr. Gary Bowman, Mr. Ian Cates; Mr. Terry Cox; Mr. Simon Golding; Mr. Tom
Stanef; and Mr. Stan Woithe, who each contributed to this research through their
dedication, persistence and hard work in developing equipment aimed at capturing the
ground response of the impact roller. It took many trials, failures and refinements to
develop the bespoke software program and data acquisition system, earth pressure cell
configuration and accelerometers that were used in this research.
The generous support from Broons, in particular company Director, Mr. Stuart Bowes is
greatly appreciated. The time, patience and support (financial and in-kind), that Broons
provided me, has given me many opportunities to conduct research in different soil
conditions and at various sites across Australia. The costs and logistics of providing
access to the impact roller, personnel to operate it, hire of earthmoving equipment,
flights, meals and accommodation in many parts of Australia are significant; the work
undertaken in this thesis would not have been possible but for their contribution. The
support of Broons staff who have assisted with and helped organise dedicated research-
intensive field trials is greatly appreciated, in particular, Mr. Bruce Constable, Mr. Guy
Bowden, Mr. Angus Bowes, Mr. John Drogemuller and Mr. Ron Hanson.
I wish to acknowledge the assistance of the following organisations for who have
allowed me to work on impact rolling field trials and compaction projects at sites that
were under their control: Ascon, BAM Clough, Bardavcol, Bechtel, Collins Transport,
xi |
ADE | The Impact of Rolling Dynamic Compaction
Ertech, Evansbuilt, Flinders Ports, Hansen Yuncken, HWE Mining, Leighton
Contractors, Monarto Quarries, Puma Roseneath, Terra Forma Civil, Territory Iron and
Windsor Earthmoving. Having access to the 4-sided impact roller in a range of site
conditions across Australia has significantly helped my understanding and provided
invaluable practical experience for which I am very grateful.
I wish to thank my colleague Dr. Yien Lik Kuo for his friendship, suggestions and
encouragement. In addition, I would like to thank Ms. Cathy Cates for her assistance
with typing and formatting this document.
Since 2010, a number of final year undergraduate students have assisted with field work
related to this research. I would like to thank the following students for their enthusiasm
and hard work in the field: Tom Bierbaum, Gianfranco Canala, Stefan Chenoweth,
Jordan Colbert, Chris Gauro, Rob Lane, Dapeng Liu, Jackson March, Nicole Mentha,
Jessica Piotto, Simon Pointon, Julianne Saw, Chris Smith, Richard Strapps,
Aidan Symons, Tom Treloar, Ross Vince and Penelope Wrightson.
I wish to thank my wife, Brooke, for her love, sacrifice and unconditional support; this
research would not have been possible without her help, patience and understanding.
The encouragement and support from my sisters Lydia and Naomi, and Brooke’s
parents, Dean and Jennifer, is also greatly appreciated. Finally, I wish to thank my
parents, Richard and Margaret, for the many sacrifices they have made on my behalf
over the years and for always supporting my academic pursuits.
xii |
ADE | The Impact of Rolling Dynamic Compaction
Firstly, there is a need to quantify what, if any, effect that towing speed has on the
effectiveness of RDC. RDC imparts energy to the ground via the use of a heavy non-
circular module that impacts the ground. At project sites involving mixed soils, isolating
and quantifying the effects of a single constraint such as towing speed is difficult due to
soil heterogeneity. To ensure that the effects of towing speed are not concealed by other
variables, there is a need to vary towing speed in homogeneous soil conditions in
dedicated research field trials.
Secondly, there is a limited depth to which ground improvement using RDC is effective.
There is little published information quantifying depths to which ground can be
improved, and how that may vary depending upon the soil type. Additionally, there is
confusion between two common applications of RDC: (1) compacting soil in thick lifts
and (2) improving soil in situ; these are two distinctly different applications of RDC that
must be treated separately. This research will distinguish between the two and will
provide recommended improvement depths for both applications using the results of
research field trials and published case studies over the past four decades.
Thirdly, this thesis introduces a new approach to quantify the ground response to RDC
via measuring stress (which can be converted to load) using buried earth pressure cells,
and measuring the acceleration response due to RDC via accelerometers placed in three
orthogonal directions. Double integration of the acceleration-time response allows
displacement to be inferred. This research will quantify the in-ground load-displacement
response of RDC in real-time that traditional pre- and post-compaction testing is unable
to do.
1.4 Research Objectives
The underlying objective of this research is to quantify the impact of RDC. This will
enable greater understating of RDC theory so that its application and validation in the
compaction and improvement of poor quality ground can be achieved more
appropriately and with greater confidence.
The objectives of this research are to:
1. Determine the effects of towing speed for the 4-sided impact roller. Trials will
be undertaken using buried earth pressure cells to quantify the differences in
4 |
ADE | The Impact of Rolling Dynamic Compaction
Influence of towing speed on the effectiveness of Rolling Dynamic
Compaction (Paper 1)
Abstract
The influence of towing speed on the effectiveness of the 4-sided impact roller using
earth pressure cells (EPCs) is investigated. Two field trials were undertaken; the first
trial used three EPCs placed at varying depths between 0.5 m and 1.5 m with towing
speeds of 9-12 km/h. The second used three EPCs placed at a uniform depth of 0.8 m,
with towing speeds of 5-15 km/h. The findings from the two trials confirmed that
towing speed influences the pressure imparted to the ground and hence compactive
effort. This paper proposes that the energy imparted to the ground is best described in
terms of work done, which is the sum of the change in both potential and kinetic
energies. Current practice of using either kinetic energy or gravitational potential energy
should be avoided as neither can accurately quantify rolling dynamic compaction
(RDC) when towing speed is varied.
2.1 Introduction
Improving the ground is a fundamental and essential part of civil construction.
Compaction is a prevalent ground improvement technique that involves increasing the
density of soil by means of mechanically applied energy to increase shear strength and
stiffness or reduce permeability. This paper is concerned with rolling dynamic
compaction (RDC) which involves traversing the ground with a non-circular roller.
Typical module designs have 3, 4 or 5 sides. As the module rotates, it imparts energy to
the soil as it falls and impacts the ground. More introductory information pertaining to
RDC is included in Scott and Jaksa (2015) and Ranasinghe et al. (2017).
At filled sites containing significant soil variability, it can be difficult to quantify the
effect of a single variable. Similarly, the inherent soil heterogeneity of natural ground
can also influence results, often making it hard to quantify the effect of towing speed
alone. To overcome this limitation, two compaction trials that used homogeneous soil
conditions are described in this paper. Both trials used buried earth pressure cells
(EPCs) and were undertaken at a dedicated research site. Whilst replacing natural soil
with fill material and conducting full-scale trials are expensive exercises, particularly
where the trial is not part of a client funded project, having full control over a site
15 |
ADE | The Impact of Rolling Dynamic Compaction
enabled variables other than towing speed to be held constant. The aim of this paper is
to determine the influence, if any, of towing speed on the energy imparted to the
ground.
The impact roller was originally developed in South Africa with the intention of
improving the properties of granular soils, in particular to identify and improve
collapsing sands within 3 m below the ground surface in southern Africa (Clifford,
1978). Wolmarans and Clifford (1975) described a case study of compacting Kalahari
(collapsing) sand in Rhodesia where at least 25 passes were required; layers were able
to be compacted in thicknesses of up to 1.5 m and still achieve the target density.
Clifford (1975) stated that the impact roller is not a finishing roller, as it over-compacts
the near-surface soils, often requiring the upper 0.1-0.2m to be compacted by rollers
used for surfacing works. Ellis (1979) described that one of the main advantages of
RDC was to compact cohesionless soils in thick layers; however, he cited a
disadvantage that in loose soils, the near-surface soil is disturbed by RDC and must be
compacted by other machines, agreeing with the results of Clifford (1975).
The typical operating speed range of the 4-sided impact roller, as shown in Fig. 2.1, is
9-12 km/h. Clifford (1980) stated that one of the difficulties encountered with RDC is
the need for rollers to be operated at their optimum speed to ensure that sufficient
energy is generated for each impact blow. In cases where the towing speed is slower
than the typical range, or the module slides across the surface, Clifford (1980) found
that adding a capping layer of material containing a granular/cohesive mixture could
reduce lateral shearing effects and aided traction of the module for typical towing
speeds. Clifford (1978) described a case study where an insufficiently thick capping
layer was adopted which resulted in individual impact blows punching through to the
underlying dredged fill; the site was also divided into a series of small working areas in
which the roller was unable to maintain a towing speed within the typical range.
According to Clifford (1978), both factors cause a reduction in speed and are the key
reasons that better results could not be obtained.
16 |
ADE | The Impact of Rolling Dynamic Compaction
Figure 2.1: 4-sided RDC module (Broons)
Clifford (1980) discussed that there is an upper speed limit beyond which an impact
blow is not delivered by the face of the module. At towing speeds greater than the
typical range, Clifford (1980) stated that the roller can spin as a circular mass and only
contact the ground with its corners, a condition that should be avoided. Avsar et al.
(2006) described the compaction of a 22-km2 reclamation area for the new Doha
International Airport Project. They identified towing speed as one of the most important
indicators that directly influenced the in situ dry density that could be achieved; an
optimum towing speed of the 4-sided roller for that project was found to be 11 km/h.
Chen et al. (2014) conducted a laboratory investigation on a scale model impact roller
device in loose dry sand, by examining the effect of module weight, size and towing
speed. They used a Chinese cone penetration test to confirm that towing speed was one
of the most important factors contributing to the effectiveness of the impact roller. The
aforementioned cases generally support the concept that towing speed influenced the
effectiveness, as did the findings of Scott and Suto (2007), who stated that ground near
the perimeter of a fenced site could not be improved as successfully as the rest of site
due to access-related issues that reduced the towing speed of the module. This paper
presents the findings of two full-scale field trials that were undertaken to quantify the
effect of towing speed for the 4-sided impact roller.
17 |
ADE | The Impact of Rolling Dynamic Compaction
2.2 Testing Methodology
Each time the module of an impact roller strikes the ground, a pressure wave is created
that travels through the soil from the surface. A key aim of the trial is to measure the
loading-induced stresses below the ground due to RDC. EPCs allow real-time
measurements of stresses imparted to the ground. Rinehart and Mooney (2009)
successfully used Geokon Model 3500 semiconductor type EPCs in a field trial to
measure dynamic loading induced from vibratory circular drum rollers. They used
100 mm diameter cells that were 10 mm thick with normal stress measurement ranges
of 250 kPa, 400 kPa and 1,000 kPa. The same type of cells were selected to measure the
pressure imparted into the soil due to RDC, albeit 230 mm diameter cells of 6 mm
thickness with a normal stress measurement range of 6,000 kPa to capture the expected
higher loads from the impact roller.
It has been well documented by researchers (e.g. Weiler and Kulhawy, 1982; Rinehart
and Mooney, 2009) that a buried cell can influence localised stress fields and therefore
any measurements may not be representative of the true loading-induced stresses. They
discussed that errors can be minimised via the choice of pressure cell design, by
undertaking calibration and by the use of correct field placement techniques. Given the
challenges associated with measuring in situ stress accurately, it was important to
characterise the uncertainty in the measurement techniques adopted. A whole system
calibration was performed both pre- and post-testing, whereby the worst-case scenario
was a difference of 8.5%. This magnitude of error is generally consistent with that
reported by Dave and Dasaka (2011) who compared different calibration techniques for
EPCs and stated that pressure cell output could be considered reliable within an error of
approximately 10%. The dynamic frequency response (peak capture) was affected by
the data acquisition rate and any internal filtering used in the signal path. The data
acquisition rate selected was 2,000 samples per second, and the filter used was set at
800 Hz. Fast Fourier transform analysis of the data indicated that the fundamental
frequency of impulses due to RDC was less than 800 Hz, confirming that the peak
values were not attenuated by the adopted filter.
2.2.1 Trial A
A field trial was undertaken at Monarto Quarries, located approximately 60 km
southeast of Adelaide, South Australia. The test site was primarily chosen because there
18 |
ADE | The Impact of Rolling Dynamic Compaction
was access to earthmoving equipment, and importantly, homogeneous quarry material
was used for the field trial. An area within the quarry where the ground was flat, close
to material stockpiles, yet away from quarry operations was chosen for the trial. Natural
soil was removed to a depth of 1.75 m, over a plan area that was 10 m long and 5.5 m
wide. Three Geokon Model 3500 EPCs were buried at nominal depths of 0.5 m, 1 m
and 1.5 m within the quarry fill material that was placed in seven lifts of 250 mm
thickness. Bedding sand was placed immediately below and above each pressure cell to
ensure horizontal placement and to prevent gravel sized particles of the fill material
from damaging the cells. Each lift was wheel-rolled using a Volvo L150E loader; a
vibrating plate compactor was used to compact soil within 250 mm from each EPC to
prevent possible damage.
2.2.1.1 Material classification
The fill material placed for the trial was a crushed rock with a maximum particle size of
20 mm that was readily available and locally produced. A summary of the particle size
distribution and Proctor compaction test results for Trial A is given in Table 2.1. For
Trial A, particle size distribution (ASTM D6913-04(2009), 2009) results are the average
of nine tests, and the standard (ASTMD698-12, 2012) and modified (ASTMD1557-12,
2012) Proctor compaction results are the average of three curves. The field moisture
content (ASTM D2216-10, 2010) reported is the average of nine tests undertaken.
Atterberg limit testing (ASTM D4318-10, 2010) confirmed that the fines consisted of
clay of low plasticity. According to the Unified Soil Classification System (USCS), the
fill material used for this compaction trial could be described as well-graded gravel
(GW).
Table 2.1: Particle size distribution, compaction and field moisture test results of 20 mm crushed
rock fill material for Trials A and B
d50 G sr ia zv ee l Sand Fines OS Mtd C MS Dtd D FMC OM Mod C MM Dod D
Trial (mm) (%) size (%) (%) (%) (kN/m3) (%) (%) (kN/m3)
A 4.0 57 40 3 7.9 17.9 8.6 7.2 18.9
B 3.5 58 38 4 12.6 19.2 9.6 10.0 19.8
Note:d = particle size at percent finer of 50%; OMC = optimum moisture content; MDD = maximum
50
dry density; FMC = field moisture content.
19 |
ADE | The Impact of Rolling Dynamic Compaction
The aim of Trial A undertaken in August 2012 was to measure the loading-induced
stress at three different depths for 40 passes in total; 10 passes of the roller were
conducted at each of the towing speeds of 9, 10, 11 and 12 km/h. Towing speed was
controlled via the control panel in the towing unit (i.e. tractor) but was subsequently
validated by dividing the distance between EPCs by the time interval between the peak
pressures that were measured. Three EPCs were used to measure the pressure imparted
to the ground, each offset by one-half of one revolution of the module (2.9 m) in the
forward direction of travel. Avalle et al. (2009) used buried instrumentation to capture
the ground response of the 4-sided impact roller and their work found that the time
during which the impulse load occurred was less than 0.1 s. They found that a sampling
frequency of 2 kHz was sufficient to capture the rapid increase in pressure caused by
impact from RDC and this same sampling frequency is adopted for the field trial
presented in this paper. The selection of thin EPCs used in the present trial provides a
much more reliable measurement of in situ soil stress than the bulky load cell used by
Avalle et al. (2009), which is significantly stiffer than the surrounding soil.
2.2.1.2 Assessment of EPC Results
Fig. 2.2 presents example results of the measured pressures versus time for a single pass
of the impact roller travelling across the test site. The order in which the three traces
were recorded is a function of the physical placement of the EPCs in the ground; 1.5 m
depth located farthest left, 1 m depth in the middle and 0.5 m farthest right. The largest
peak pressure was observed for the EPC buried at 0.5 m depth, whereas the deeper
pressure cells at 1 m and 1.5 m depths recorded smaller impulses, indicating that the
pressure imparted into the soil reduces in magnitude and increases in area with greater
depth, as expected. Fig. 2.3 highlights a single impact blow measured by an EPC, where
a loading-induced peak pressure of 648 kPa was recorded at 0.5 m depth. Fig. 2.3
demonstrates the dynamic nature of RDC and the importance of adopting a 2 kHz
sampling frequency is evident from the individual data points shown, given that the
loading and unloading phases occur over a time period of approximately 0.045 s.
20 |
ADE | The Impact of Rolling Dynamic Compaction
Fig. 2.4 presents the relationship between the measured peak pressures versus depth for
each of the towing speeds examined, with an increasing trend between the peak pressure
and towing speed evident for all depths measured, and a decrease in pressure with
depth, as one would expect. As can be observed from these results, a clear relationship
exists between measured pressure and towing speed, with the slowest speed of 9 km/h
yielding the lowest pressures, and progressively increasing with greater speed. Fig. 2.5
presents the results of the measured peak pressure plotted against offset distance for all
depths, whereby the offset distance is defined as the distance between the centre of the
module and the centre of the buried EPC. From this figure, it can be observed that, at
shallow depths, offset distance has a large influence on the peak pressure recorded.
However, with increasing depth, the effects of offset distance are less pronounced,
suggesting a greater radial effect away from the centre of impact as depth increases. For
an EPC depth of 0.5 m, offset distances between -100 mm and 400 mm generated the
greatest pressures, apart from an anomalous result at an offset of -275 mm, and two
other offsets that coincide with the corners of the module (-650 mm and 650 mm). This
finding is generally consistent with Avalle et al. (2009), who found that the zone of
maximum impact was located from 0 mm to 400 mm from the centre of the module. In
order to further examine the effects of towing speed, an additional field trial was
undertaken.
22 |
ADE | The Impact of Rolling Dynamic Compaction
1000 1000
1.5 m depth
900 900
1.0 m depth
800 800
0.5 m depth
700 700
600 600
)
a
P 500 500
k
(
e
r
u 400 400
s
s
e
r
P 300 300
200 200
100 100
0 0
-800 -600 -400 -200 0 200 400 600 800
Offset between centre of pressure cell and centre of module (mm)
Figure 2.5: Non-uniform pressure distribution measured at 0.5 m, 1.0 m and 1.5 m depths.
2.2.2 Trial B
Field Trial B was undertaken at Monarto Quarries during August 2014, albeit at a
different location from Trial A. Natural soil was removed to a depth of 1.2 m, over a
plan area 12 m long and 3 m wide. Three Geokon Model 3500 EPCs were placed at a
constant depth of 0.8 m. Quarry fill material was placed in six equal lifts of 200 mm
thickness, with each lift again being wheel-rolled using a Volvo L150E loader and a
vibrating plate compactor used to compact soil within 200 mm from each EPC. The aim
of the field trial was to measure the loading-induced stress at a single depth for 100
passes in total; 35 passes of the roller were conducted at a towing speed of 12 km/h
prior to comparative EPC measurements being undertaken to achieve effective refusal.
Five passes were conducted at each of the following towing speeds and in the following
order: 12, 10, 8, 6, 9, 7, 5, 11, 14, 13 and 15 km/h, respectively. Due to time constraints,
no EPC measurements were recorded between passes 90 and 100.
24 |
ADE | The Impact of Rolling Dynamic Compaction
2.2.2.1 Material classification
The fill material placed for the trial was a crushed rock with a maximum particle size of
20 mm that was readily available on site. A summary of the particle size distribution
(ASTM D6913-04(2009), 2009) and standard (ASTM D698-12, 2012) and modified
(ASTM D1557-12, 2012) Proctor compaction test results for Trial B is given in
Table 2.1. The test results indicate that the material is similar to that used in Trial A;
however, there are differences which can be attributed to the two-year interval between
trials, different weather conditions at the time of testing, and the material being sourced
from different parts of the quarry. For Trial B, the particle size distribution results are
the average of seven tests, and the standard and modified Proctor compaction curves
were generated using a minimum of five data points each; both laboratory compaction
curves were generated five times. The field moisture content reported is the average of
30 tests undertaken. According to the USCS, the fill material is again classified as well-
graded gravel (GW). Atterberg limit testing confirmed that the fines consisted of clay of
low plasticity.
Density measurements and other in situ tests were not undertaken during either field
trial presented in this paper. However, the authors carried out in situ tests from pre- and
post-compaction in very similar soil conditions as this study during a separate field trial
that was also conducted at Monarto Quarries. The results have been published in Scott
et al. (2016). It is acknowledged that only undertaking pre- and post-compaction testing
provides limited information regarding changes in soil state with increasing compactive
effort; however, such testing regimes are common as they are effective at determining
whether a project specification has been met, or otherwise. A recently published paper
by Scott et al. (2019) captured the ground response of a single module impact in real-
time using buried EPCs and accelerometers.
2.2.2.2 Assessment of EPC Results
Fig. 2.6 presents the minimum, maximum and average peak pressures that were
recorded at varying towing speeds. As mentioned above, five passes were conducted at
each target towing speed, with each pass traversing over three EPCs at a uniform depth
of 0.8 m, resulting in 15 data points per towing speed. It can be observed that at towing
speeds lower than 9 km/h, significantly lower pressure is imparted to the soil. The
maximum pressure (1,220 kPa) was recorded at a towing speed of 14 km/h and the
25 |
ADE | The Impact of Rolling Dynamic Compaction
highest average peak pressure (646 kPa) at a towing speed of 11 km/h. Large pressure
variations were measured for the same towing speed due to limitations of using EPCs
that are buried at fixed locations. The location of the centre of the module landing on
the ground surface relative to the centre of a buried EPC is variable. As discussed by
Avalle et al. (2009), this variability is something unable to be controlled (despite some
attempts at trying to do so). As discussed by Scott et al. (2016), whilst the module is
nominally a “square”, the sides have curved features, and this results in a non-uniform
pressure distribution and is a key contributing factor why some passes yielded much
larger peak pressures for the same towing speed than others.
1400
Max Peak Pressure
1200 Mean Peak Pressure
Min Peak Pressure
1000
800
)
a
P
k
(
e 600
r
u
s
s
e
r
P 400
200
0
5.0 6.0 7.0 8.0 9.0 10.0 11.0 11.9 13.0 14.0 15.2
Towing Speed (km/h)
Figure 2.6: Minimum, maximum and average peak pressures for varying towing speeds.
Fig. 2.7 presents the same data set, plotted instead with peak pressure versus offset
distance. Adjacent speeds have been combined to yield 30 data points for each line. It
can be observed that, for increasing towing speed, greater pressure is imparted to the
ground up to 11-12 km/h. For speeds of 13-14 km/h, the shape of the pressure versus
offset relationship is in contrast to the other towing speeds, indicating that the corners of
the module impart the greatest pressure. This suggests that the behaviour of the module
changes with increasing towing speed, which is consistent with the findings of Clifford
26 |
ADE | The Impact of Rolling Dynamic Compaction
1400
13 km/h
1200
11
1000 9
15
) 800
a
P
k
(
e
r
u
s
s
e
r P 600 7
400
200
0
0 0.02 0.04 0.06 0.08 0.1
Time (seconds)
Figure 2.8: Duration of pressure impulse not greatly influenced by towing speed.
To confirm the observations from the pressure cell data, a number of qualitative
behaviours were observed; at lower towing speeds, the blows were delivered by the face
of the module, which maintained a regular contact pattern with the ground. At faster
speeds, the blows were delivered towards the corners, and the module was observed to
skip along the surface from corner to corner, which is again consistent with the findings
from Fig. 2.7 and Clifford (1980). The spacing between successive blows of the roller
module was also monitored and physically measured on site. The module imprint length
was measured to be significantly larger than the physical face length (1,450 mm) of the
module for towing speeds greater than 13 km/h as indicated in Fig. 2.9, implying non-
uniform rotation and skipping behaviour. Bradley et al. (2019) used high-speed
photography that captured the kinematics of the 4-sided module at 1000 frames per
28 |
ADE | The Impact of Rolling Dynamic Compaction
second. The field work undertaken by Bradley et al. (2019) is highly relevant to the
field work of this study even though the two field trials had different aims and
motivations and were undertaken on separate (adjacent) test areas within the Monarto
Quarries site. There are strong similarities between the two; both field trials were held
concurrently, allowing the same 4-sided impact roller to be used and fill material from
the same stockpile to be used. The study by Bradley et al. (2019) captured the motion
and estimated the kinematic profile of the module during impact to estimate the energy
imparted to the ground (23 kJ ± 4 kJ) for a constant towing speed of 10 km/h that was
adopted during the trial.
2.4
Too Less than
Optimum Toofast
2.2 slow optimum
)m
(
e
c
a 2.0
f
r
u
s
d
n
u 1.8
o
r
g
n
o
h 1.6
t
g
n
e l
t
1.45 m
n 1.4
ir
p
m
i
e
lu 1.2
d
o
M
1.0
4 5 6 7 8 9 10 11 12 13 14 15 16
Towing speed (km/h)
Figure 2.9: Inconsistent module imprint length on ground surface with increasing towing speed.
2.3 Discussion
In this paper, towing speed refers to the horizontal motion of the towing unit, whereas
rotational velocity refers to the angular velocity of the module. To quantify the
difference between the two, Clifford and Bowes (1995) presented theoretical analyses
from independent mathematicians who predicted the change in rotational velocity of the
module as it falls to impact the ground. They claimed that towing speed was more
29 |
ADE | The Impact of Rolling Dynamic Compaction
significant than other factors such as module mass or lift height. Whilst the use of load
cells is referenced in their paper, no experimental results were included to confirm their
findings. Clifford and Bowes (1995) used high-speed photography to support their
calculations regarding the change in angular velocity of the module during the lifting
and falling phases of each impact for a constant towing speed. They explained that a
key reason why the angular velocity of the module is not constant (unlike the towing
speed) is due to the double-spring-linkage system on the 4-sided impact roller. Clifford
and Bowes (1995) explained that the module velocity is slowed during the lifting phase
as the springs of the double-linkage system are compressed. This causes the module to
lag a little behind the towing frame that is travelling at a constant speed. During the
impact phase, the springs are then discharged which cause the module to move faster
than the towing frame as the spring energy is released. Whilst no results of the high-
speed photography were presented in their paper, they claimed that the spring energy
resulted in a decrease in rotational velocity during lifting, and an increase in module
velocity during the falling phase. They found that the magnitude of change in module
rotational velocity was inconsistent and was dependent upon soil surface irregularities.
Their calculations proposed that the energy delivered by the 4-sided roller during a
single impact can be described by kinetic energy, estimated to be up to 50 kJ, depending
upon their assumptions made regarding the velocity of the module upon impact with the
ground, v.
f
McCann (2015) used 3- and 5-sided modules and presented an alternative viewpoint,
stating that the magnitude of the gravitational potential energy provides a reasonable
estimate of the energy delivered by the 3-sided roller. McCann (2015) cited the work of
Heyns (1998) who undertook both theoretical and empirical analyses. Heyns (1998)
placed an accelerometer on the axle of a 3-sided impact roller to measure the magnitude
of the peak deceleration of the module as it impacted the ground. Heyns (1998) used
dynamic compaction theory from Mayne and Jones (1983) to infer the energy imparted
to the ground based on the measured peak deceleration. Whilst good agreement between
estimated and measured accelerations was noted by Heyns (1998), both are
fundamentally based on dynamic compaction theory. The use of this theory without
modification for RDC applications is questionable and requires further research. Heyns
(1998), cited by Berry (2001), observed that an increase in towing speed resulted in an
increase in energy imparted to the ground, but it was not the major component of the
energy for towing speeds tested between 9 km/h and 14 km/h. After losses were taken
30 |
ADE | The Impact of Rolling Dynamic Compaction
into account, Heyns (1998) concluded that the magnitude of the gravitational potential
energy, PE (Eq. (2.1)), was a reasonable estimate for the energy delivered by the
g
3-sided roller to the ground. If this theory is applied to a 4-sided impact roller with a
module mass, m, of 8-tonne and a maximum module drop height, h, of 0.15 m, the
estimated energy imparted to the ground would be approximately 12 kJ.
𝑃𝐸 = 𝑚𝑔ℎ [2.1]
𝑔
where g is the gravitational acceleration.
Clearly, there is a need for further research as this finding is in stark contrast with that
of Clifford and Bowes (1995) who estimated the energy for a single impact using total
kinetic energy, KE (Eq. (2.2)), based on an 8-tonne module mass, m, and a module
landing velocity, v , that was assumed to be greater than the towing speed.
f
𝐾𝐸 = 1 𝑚𝑣2 [2.2]
𝑓
2
The fact that Clifford and Bowes (1995) analysed a 4-sided roller and Heyns (1998)
analysed a 3-sided roller may, to some extent, explains the disparity in results. The
standard 4-sided impact roller, as shown in Fig. 2.1, consists of a single 8-tonne module
that is 1,300 mm wide, 1,450 mm high and rotates with the aid of a double-spring-
linkage system. The standard 3-sided impact roller, as shown in Fig. 2.10, consists of
twin 6-tonne modules that are each 900 mm wide and 2,170 mm high that rotate about a
fixed axle with the aid of a hydraulic accumulator. The concept of energy storage upon
lifting and release on impact theoretically increases the potential energy imparted to the
ground; however, there is little, if any, published information that quantifies the
magnitude of the energy that can be stored and released by either the double-spring-
linkage system or the hydraulic accumulator.
31 |
ADE | The Impact of Rolling Dynamic Compaction
Figure 2.10: 3-sided RDC module (Source: Landpac.com)
In an attempt to quantify the effects of the spring-linkage system, Clifford and Bowes
(1995) analysed the change in angular velocity of the module before and after impact.
They did not, however, quantify the contribution of spring energy in terms of the
potential energy imparted to the ground. Whilst differences in impact roller
configuration may account for some of the disparity in the estimates provided by Heyns
(1998) and Clifford and Bowes (1995), there is clear disagreement as to whether the use
of potential energy or kinetic energy provides more accurate estimates. It is also
apparent that research is required to determine the effects of the double-spring-linkage
system and the hydraulic accumulator to be able to accurately quantify the total
potential energy delivered by the 4- and 3-sided impact rollers, respectively.
From both field trials undertaken, it is evident that the towing speed of the module
influences the pressure imparted to the ground, suggesting that gravitational potential
energy alone does not accurately capture the ground response of RDC. Whilst Heyns
(1998) found that towing speed influenced the energy imparted to the ground at towing
speeds higher than the typical range, these findings present compelling evidence that the
magnitude of the energy imparted to the ground is a function of towing speed, even
within the typical operating range of 9-12 km/h. Clifford and Bowes (1995) argued that
module speed was a critical parameter, and that the continuous rolling action must be
more beneficial than the equivalent falling weight that relied solely on gravitational
potential energy. However, the magnitude of peak pressures measured in the ground
with changes in towing speeds strongly suggests that the use of total kinetic energy does
32 |
ADE | The Impact of Rolling Dynamic Compaction
not accurately describe it either. If it did, greater changes in pressure would have been
evident with varying speed. The observations indicate that total kinetic energy over-
estimates the contribution of towing speed, and therefore does not provide a reliable
estimate of the energy imparted to the ground. Combining the findings of past research
and the trials presented in this paper, the energy imparted to the ground appears to be a
function of both potential and kinetic energies. To determine the magnitude of energy
imparted to the ground by a single blow, it is necessary to analyse the potential and
kinetic energy before and after impact in more detail, which is addressed below.
2.3.1 Energy Imparted by RDC
In order to estimate the energy imparted to the ground as a consequence of RDC, the
conclusions from the high-speed photography undertaken by Clifford and Bowes (1995)
are adopted. They indicated that, when compared to the average, the module velocity
decreased by 10-20% during the lifting phase of the module, and increased by 10-20%
during the falling phase. The module frame is towed at a relatively constant speed,
therefore the speed of the module after impact with the ground is slower than that prior
to impact, but is not zero as implied by Clifford and Bowes (1995) for their use of total
kinetic energy to be correct. For calculation purposes, a module mass, m, has a velocity
increase of +10% prior to impact, v, and a velocity decrease of -10% after impact, v,
i f
when compared to the average. These correspond to lower bound values stated by
Clifford and Bowes (1995), to determine the work done due to the change in kinetic
energy, W , which is equal to DKE, as defined using Eq. (2.3). The results are presented
ke
in Table 2.2.
𝑊 = ∆KE = 1 m𝑣2 −1 m𝑣2 [2.3]
𝑘𝑒 𝑖 𝑓
2 2
33 |
ADE | The Impact of Rolling Dynamic Compaction
Table 2.2: Predicted change in kinetic energy based on high-speed photography by Clifford and
Bowes (1995)
v v vi vf DKE
(km/h) (m/s) (m/s) (m/s) (kJ)
8 2.22 2.44 2.00 7.8
9 2.50 2.75 2.25 10.0
10 2.78 3.06 2.50 12.5
11 3.06 3.36 2.75 14.9
12 3.33 3.67 3.00 17.8
13 3.61 3.97 3.25 20.8
Note:v = speed of towing unit; v = module velocity prior to impacting the ground; v = module velocity
i f
after impacting the ground.
The change in potential energy, DPE , is equal to the work done due to gravity, W ,
g g
therefore, the module falling to the ground surface can be described by Eq. (2.4), in
which the module drop height after impact, h , is equal to zero; hence for an 8-tonne
2
mass, m, and a lift height (h ) of 0.15 m, DPE ≈12 kJ.
1 g
𝑊 = ∆𝑃𝐸 = 𝑚𝑔ℎ − 𝑚𝑔ℎ [2.4]
𝑔 𝑔 1 2
It should be emphasised that Eq. (2.4) gives the maximum potential energy that can be
delivered to the ground. This energy will not be delivered with every impact as the full
gravitational potential energy will only be reached when the module is compacting soil
that is hard enough to allow the full lift height to be achieved. It is noted that using
high-speed photography will also capture changes in module velocity due to the spring-
linkage system, or due to energy losses in the system (such as frictional forces that act
between the module and the ground surface). The net work done, W, as described by
Eq. (2.5), is a combination of both the change in potential and kinetic energies, as work
is being done against gravity, as well as inertia and frictional resistive forces, and is
considered a more appropriate means to describe the energy delivered by RDC, rather
than relying solely on gravitational potential or total kinetic energy.
34 |
ADE | The Impact of Rolling Dynamic Compaction
W = ∆PE + ∆KE [2.5]
The high-speed photography approach used by Clifford and Bowes (1995) quantified
the spring energy in terms of a change in module rotational velocity as the springs are
compressed and subsequently released. However, spring energy, as defined by Halliday
et al. (1993), is a form of potential energy, therefore the contribution of the dual springs
in the linkage system should, more appropriately, be quantified in terms of potential
energy.
2.3.1.1 Contribution of the spring-linkage system
The double-spring-linkage system consists of two springs: a large outer spring and a
smaller inner spring that fits within the internal diameter of the larger spring. To
determine the contribution of each of the springs to the energy imparted by the module,
the stiffness of both springs was determined. Each spring was placed separately in a
large compression machine whereby the load versus displacement response was
quantified. The maximum compression of the dual springs was governed by the limiting
compression distance of the outer spring, as both springs compress together in the
towing frame. The force in the spring is determined using Hooke’s law in Eq. (2.6),
where the spring force, F , is a function of the spring stiffness, k, and the compression
s
distance of the spring, x:
𝐹 = −𝑘𝑥 [2.6]
𝑠
Based on Halliday et al. (1993), the work done by a spring, W can be determined using
s
Eq. (2.7):
𝑊 = ∫0 (𝐹)𝑑𝑥 = 1 𝑘𝑥2 [2.7]
𝑠 −𝑥𝑚𝑎𝑥 𝑠 2 𝑚𝑎𝑥
35 |
ADE | The Impact of Rolling Dynamic Compaction
Where x is the maximum spring compression. Using Eq. (2.7), it is possible to
max
determine the work done, W, by both the inner and outer springs with varying spring
s
compression distances up to the maximum (limiting) compression, x . Whilst having
max
different spring stiffnesses, k, both the inner and outer springs compress by the same
magnitude in the double-linkage mechanism, the work done by the springs is equal to
the change in spring potential energy, DPE , as described by Eq. (2.8).
s
𝑊 = ∆𝑃𝐸 = (1 𝑘𝑥2 ) +(1 𝑘𝑥2 ) [2.8]
𝑠 𝑠 𝑚𝑎𝑥 𝑚𝑎𝑥
2 𝑖𝑛𝑛𝑒𝑟 2 𝑜𝑢𝑡𝑒𝑟
The outer spring was found to contribute 84% of the work done by the dual springs
combined, due to the larger spring stiffness (k = 370 N/mm), compared to the inner
spring (k = 70 N/mm). As observed in Fig. 2.11, the work done by the springs is
approximately 5 kJ at the maximum spring compression. This is the maximum energy
that the springs are able to deliver, but the full potential energy of the springs will not be
delivered with every blow, as both the geotechnical properties of the ground and the
undulating surface profile significantly affect the behaviour of the module.
36 |
ADE | The Impact of Rolling Dynamic Compaction
6
Dual Spring
Combination
5 Outer Spring
4
)J
k
( 3
e
n
o
D
k
r
o 2
W
1
0
0 30 60 90 120 150
Spring Compression (mm)
Figure 2.11: Energy contribution of the dual springs in the linkage system of the 4-sided impact
roller.
A summary of the work done with varying speed is presented in Fig. 2.12. It is observed
that the change in gravitational and spring potential energies is constant for all speeds.
The maximum spring energy is more likely to be realised at faster towing speeds;
however, further research involving more direct measurement techniques is needed to
confirm this. As stated previously, the change in kinetic energy, as quantified by
Clifford and Bowes (1995), accounts for spring effects and this is supported by
Fig. 2.12, where DPE < DKE. Without taking into account the spring energy
s
contribution twice, the total work done is equal to the sum of the change in gravitational
potential, and kinetic energies (Eq. (2.5)). This yields values of total work done between
22 kJ and 30 kJ for typical towing speeds of 9 km/h and 12 km/h, respectively. For the
same speeds, Clifford and Bowes (1995) predicted 30 kJ - 54 kJ, respectively, using
Eq. (2.2) and assuming that the spring-linkage system increases the landing velocity of
the module by 10%. The predicted energy that is imparted to the ground by Bradley et
al. (2019) does support the assumptions made by Clifford and Bowes (1995) regarding
the relationship between towing speed and module velocity that were used in this study
to estimate the change in kinetic energy. Bradley et al. (2019) quantified the change in
37 |
ADE | The Impact of Rolling Dynamic Compaction
energy due to a single module impact from high-speed photography, and estimated that
the energy imparted to the ground due to a single module impact was 23 kJ (±4 kJ) for a
towing speed of 10 km/h, consistent with the findings of this study.
Work Done (Equation 2.5) Change in KE (Equation 2.3)
Change in PE(g) (Equation 2.4) Change in PE(s) (Equation 2.8)
30
25
20
)J
k
(
e
n
o
D 15
k
r
o
W
10
5
0
9 10 11 12
Towing Speed (km/h)
Figure 2.12: Increasing energy for typical towing speeds of the 4-sided impact roller.
2.4 Conclusions
This paper examined the effect of towing speed on the energy imparted to the ground
from the 4-sided impact roller. This involved combining theory from Halliday et al.
(1993), observations from two full-scale field trials, high-speed photography by Clifford
and Bowes (1995), and estimates of energy imparted to the ground for the 3-sided roller
by Heyns (1998). The maximum imparted energy delivered to the ground by the 4-sided
impact roller was found to lie in the range between 22 kJ and 30 kJ, for typical towing
speeds of 9-12 km/h.
It is proposed that the energy imparted by RDC to the ground needs to be considered in
terms of work done, which is due to the change in both potential and kinetic energies.
Current practice of describing the energy imparted to the ground using total kinetic
energy should be avoided as it overestimates the energy imparted to the ground.
38 |
ADE | The Impact of Rolling Dynamic Compaction
Describing the energy via the use of gravitational potential energy should also be
avoided, but for a different reason; it is counter-productive for the impact rolling
industry to develop specifications stipulating target towing speeds when the rollers are
described solely in terms of their gravitational potential energy.
The change in potential energy is derived from a combination of both gravitational and
spring energies for the 4-sided impact roller. The values presented in this paper for the
potential energy delivered by the springs (5 kJ) and gravitational potential energy
(12 kJ) are the maximum values that are theoretically possible. However, they are not
values that will be achieved with every impact, as favourable ground conditions are
needed for the full potential energy to be delivered. The change in kinetic energy is a
function of the friction between the module and the ground surface. Quantifying the
friction at the module-soil interface is extremely difficult to evaluate theoretically, as it
depends on several variables associated with the module, such as the roughness of the
module face in contact with the ground, the presence of wear plates or anti-skid bars,
the contact area between the module and soil, and the towing speed. Properties relating
to the ground are also significant, with soil type, grading, moisture content, density,
elastic modulus and surface geometry all providing different frictional resistance, which
makes it complex and extremely difficult to estimate the energy needed to overcome
friction as it is material-dependent.
If the energy imparted to the ground was only due to potential energy, then it would be
theoretically independent of towing speed and would be limited to a maximum value of
17 kJ. The findings of this research confirm that towing speed does influence the energy
imparted to the ground. There is, therefore, a need for specifications to detail a target
towing speed range for RDC. Based on the authors’ experiences, the optimum speed
will vary depending on site conditions. To optimise the use of the 4-sided impact roller,
a towing speed range of 10-12 km/h is recommended, which is consistent with the
findings of the field trials reported in this paper.
39 |
ADE | The Impact of Rolling Dynamic Compaction
2.5 Declaration of Competing Interest
The authors wish to confirm that there are no known conflicts of interest associated with
this publication and there has been no significant financial support for this work that
could have influenced its outcome.
2.6 Acknowledgements
The authors wish to acknowledge the following civil engineering students from the
University of Adelaide who contributed to the field work referred to in this paper:
Andrew Bradley, Gianfranco Canala, Chris Gauro, Dapeng Liu, Jackson March, Chris
Smith and Richard Strapps. The authors are grateful to the instrumentation and
laboratory staff at the University of Adelaide who have provided invaluable assistance
over the years and to Cathy Cates for her assistance with word processing. The authors
are also grateful to Broons and Monarto Quarries personnel for access to equipment and
the test site, without their help and support, the work undertaken in the field trials
featured in this paper would not have been possible.
2.7 References
ASTM D6913-04. Standard test methods for particle-size distribution (gradation) of
soils using sieve analysis. West Conshohocken, PA, USA: ASTM International; 2009.
ASTM D2216-10. Standard test methods for laboratory determination of water
(moisture) content of soil and rock by mass. West Conshohocken, PA, USA: ASTM
International; 2010.
ASTM D4318-10. Standard test methods for liquid limit, plastic limit, and plasticity
index of soils. West Conshohocken, PA, USA: ASTM International; 2010.
ASTM D698-12. Standard test methods for laboratory compaction characteristics of soil
using standard effort. West Conshohocken, PA, USA: ASTM International; 2012.
ASTM D1557-12. Standard test methods for laboratory compaction characteristics of
soil using modified effort. West Conshohocken, PA, USA: ASTM International; 2012.
40 |
ADE | The Impact of Rolling Dynamic Compaction
Avalle DL, Scott BT, Jaksa MB. Ground energy and impact of rolling dynamic
compaction – results from research test site. In: Proceedings of the 17th International
Conference on Soil Mechanics and Geotechnical Engineering, Vol. 3, Alexandria,
Egypt, 2009: 2228–2231.
Avsar S, Bakker M, Bartholomeeusen G, Vanmechelen J. Six sigma quality
improvement of compaction at the New Doha International Airport Project. Terra et
aqua, 2006, 103: 14-22.
Berry AD. Development of a volumetric strain influence ground improvement
prediction model with special reference to impact compaction. Dissertation submitted in
partial fulfilment of the requirements of the degree of Master of Engineering, University
of Pretoria, Pretoria, South Africa, 2001. [Dissertation]
Bradley AC, Jaksa MB, Kuo YL. Examining the kinematics and energy of the four-
sided impact roller. Proceedings of the Institution of Civil Engineers – Ground
Improvement, 2019, 172 (4): 297-304.
Chen Z, Xu C, Ye G, Shen C. Impact Roller Compaction of Dry Sand in Laboratory
Tests. In: Proceedings Geo-Shanghai, Ground Improvement and Geosynthetics,
Shanghai, China, 2014: 258-269.
Clifford JM. An introduction to impact rollers. National Institute for Road Research,
Internal report RP/2/75, South Africa, 1975.
Clifford JM. Evaluation of Compaction Plant and Methods for the Construction of
Earthworks in Southern Africa. Dissertation in partial fulfilment of the requirements for
the degree of Master of Science in Engineering, University of Natal, Durban, South
Africa, 1978 [Dissertation]
Clifford JM. The development and use of impact rollers in the construction of
earthworks in southern Africa. CSIR Report 373, National Institute for Transport and
Road Research Bulletin 16, Pretoria, South Africa, 1980.
Clifford JM, Bowes G. Calculating the Energy delivered by an Impact Roller. A trilogy
of Papers for the Sept. 1995 Lecture Tour and International Seminars to commemorate
the 10th Anniversary of the BH 1300 Standard Impact Roller, Paper Two, 1995.
41 |
ADE | The Impact of Rolling Dynamic Compaction
Dave TN, Dasaka SM. A review on pressure measurement using earth pressure cell.
International Journal of Earth Sciences and Engineering, 2011, 4 (6): 1031-1034.
Ellis CI. Pavement engineering in developing countries. Transport and Road Research
Laboratory, TRRL Supplementary Report 537, Crowthorne, Berkshire, 1979.
Halliday D, Resnick R, Walker J. Fundamentals of Physics, 4th Edition. John Wiley &
Sons, Incorporated, Canada, 1993: 182-250.
Heyns S. Response analysis of an impact compactor. Report LGI98/013, Project No
020-DP, Laboratory for Advanced Engineering Pty Ltd., University of Pretoria, South
Africa, 1998.
Mayne PW, Jones JS. Impact stresses during dynamic compaction. Journal of
Geotechnical Engineering, 1983, 109 (10): 1342-1346.
McCann K. The use of impact compaction for the near surface compaction on dredged
sand land reclamation projects. In: Proceedings 12th Australia New Zealand Conference
on Geomechanics, New Zealand Geotechnical Society, Wellington, New Zealand, 2015,
1193-1200.
Ranasinghe RATM, Jaksa MB, Kuo YL, Pooya Nejad F. Application of artificial neural
networks for predicting the impact of rolling dynamic compaction using cone
penetrometer test results. Journal of Rock Mechanics and Geotechnical Engineering,
2017, 9 (2): 340-349.
Rinehart RV, Mooney MA. Measurement of Roller Compactor Induced Triaxial Soil
Stresses and Strains. Geotechnical Testing Journal, 2009, 32 (4): 347-357.
Scott BT, Jaksa MB. The effectiveness of rolling dynamic compaction – a field based
study. In: Indraratna B, Chu J, Rujikiatkamjorn C, eds. Ground Improvement Case
Histories: Compaction, Grouting and Geosynthetics. Kidlington, Oxford: Elsevier,
2015: 429-452.
Scott B, Jaksa M, Mitchell P. Ground response to rolling dynamic compaction.
Geotechnique Letters, 2019, 9 (2): 99-105.
42 |
ADE | The Impact of Rolling Dynamic Compaction
Depth of influence of rolling dynamic compaction (Paper 2)
Abstract
The depth of influence of rolling dynamic compaction (RDC) was investigated in a field
trial using the 4-sided impact roller. Earth pressure cells (EPCs) were placed at varying
depths at a site consisting of homogeneous soil conditions. EPCs measured pressures
imparted by RDC at 3.85 m depth; however, the largest magnitudes of pressure were
confined to the top 2 m beneath the ground surface. These results were complemented
by field density data, penetrometer and geophysical testing. A number of published case
studies using the 8-tonne, 4-sided impact roller, for either improving ground in situ or
compacting soil in thick layers, are summarised. Finally, equations are presented that
predict first, the effective depth of improvement, appropriate for determining the depth
to which ground can be significantly improved in situ, and, second, the depth of major
improvement for RDC, appropriate for thick layer compaction.
List of notation
D depth of soil compacted due to gravitational potential energy (m)
d particle size at per cent finer of 50%
50
g free-fall acceleration (9.81 m/s2)
h maximum module drop height (m)
k ratio of energy imparted to the ground divided by the gravitational potential
energy
m module mass (t)
n empirical factor in depth of improvement equation
r reduction factor for determining the depth of major improvement
v towing speed (m/s)
v module velocity after impacting the ground (m/s)
f
v module velocity prior to impacting the ground (m/s)
i
DKE change in kinetic energy (kJ)
49 |
ADE | The Impact of Rolling Dynamic Compaction
3.1 Introduction
There is an increasing need for civil engineers to provide cost-effective solutions for
construction on marginal or difficult sites, in particular an understanding of the
advantages and limitations of ground improvement options is essential to ensure that
technically feasible and constructible solutions are adopted. Compaction is a prevalent
ground improvement technique that aims to increase the density of soil by applying
mechanical energy to increase soil strength and decrease differential and total
settlements within a desired depth range beneath the ground surface. This paper is
concerned with a specific type of dynamic compaction known as rolling dynamic
compaction (RDC) which involves traversing the ground with a non-circular roller.
Typical module designs have three, four or five sides. As the module rotates, it imparts
energy to the soil as it falls to impact the ground. High energy impact compaction
(HEIC) and high impact energy dynamic compaction (HIEDYC) are alternative names
found in different parts of the world, or used by different contractors, for RDC.
When compared to circular drum rollers, RDC can compact thicker layers due to a
greater depth of influence beneath the ground’s surface. This is derived from a
combination of a heavy module mass, the shape of the module and the speed at which it
is towed; typically in the range of 9-12 km/h. Depths of improvement for RDC have
been found to vary significantly and the factors that affect it are not fully understood.
The depth of influence of RDC is often quantified by comparing in situ test results,
before and after compaction. At sites containing significant soil variability, the use of
pre- and post-compaction testing can be problematic. To overcome this limitation, this
paper describes a compaction trial where earth pressure cells (EPCs) were placed at
different locations beneath the ground surface in homogeneous soil conditions to
quantify the depths to which RDC improves the ground.
3.2 Background
Published case studies involving the standard 4-sided impact roller that have improved
the ground in situ, and have compacted soil in thick layers, are summarised in Tables
3.1 and 3.2, respectively. In addition to the referenced published articles, the authors
have reviewed dozens of unpublished reports that have utilised the 4-sided, 8-tonne
roller in a variety of soil conditions. Their findings are in general agreement with the
improvement depths and layer thicknesses summarised in Tables 3.1 and 3.2,
50 |
ADE | The Impact of Rolling Dynamic Compaction
respectively. It is clear from Tables 3.1 and 3.2 that the depth of improvement of RDC
varies significantly depending upon the soil material type. It is reasonable to conclude
that RDC has a greater depth of influence in granular soils compared to clays. It is also
evident that the thickness of compacted layers is less than the depth of improvement in
the same soil type, as the compacted layer thickness is typically tailored to meet a target
specification.
Table 3.1: Improvement depths for compacting in situ
No. Reference Soil type Improvement Depth
(m)
1 Clifford (1978) Sand >2.5
2 Clifford (1978) Sand >2.0
3 Avalle and Young (2004) Fill (clay) 1.0
4 Avalle (2004) Fill (sand) >2.0
5 Avalle and Grounds (2004) Fill (mixed) 1.5
6 Avalle and Mackenzie (2005) Fill (clay) 2.0
7 Avalle and Carter (2005) Fill (sand) over natural sand 3.0
8 Avalle (2007) Fill (sand) 2.5
9 Scott and Suto (2007) Fill (gravelly clay) 1.5
10 Whiteley and Caffi (2014) Fill (mixed) 1.5
11 Scott and Jaksa (2014) Fill (clayey sand) over natural clay 1.75
51 |
ADE | The Impact of Rolling Dynamic Compaction
Table 3.2: Thickness of compacted layers
ID Reference Soil type Layer Thickness (m)
A Wolmarans and Clifford (1975) Sand 1.5
B Wolmarans and Clifford (1975) Clay 0.6
C Clifford (1980) Clay 0.5
D Clifford and Coetzee (1987) Fill (coal discard material) 0.5
E Avalle and Grounds (2004) Fill (gravel) 1.0
F Avalle (2007) Sandy clay / clayey sand 0.7
G Scott and Jaksa (2012) Fill (mixed) 1.0
H Scott and Jaksa (2014) Fill (clayey sand) 1.0
Whilst not summarised in these tables, other variables such as moisture content, ground
water conditions and the number of passes applied also affect the depth to which ground
can be improved using RDC. When reviewing Tables 3.1 and 3.2 it is important to note
that the target specification, testing methods used to quantify improvement, and the
interpretation of how the depth of improvement is both defined and quantified, varies
between the listed references, making it difficult to draw definitive conclusions as to the
maximum improvement depth or layer thickness possible. In current practice, it is often
the responsibility of the project engineer to predict whether the use of RDC will
improve the ground sufficiently for the desired project application. The variable and
unknown depth of influence of RDC is a key reason why this ground improvement
technique is not used more commonly, and highlights why further research is needed.
Kim (2010) performed finite element simulations on impact rollers of different shapes
with the aim of determining the stress distribution and influence depth, which was
defined as the depth at which the vertical stress decreased to one-tenth of the applied
stress at the surface. This study held module mass, diameter and width of each roller
consistent; only the shape and number of sides varied. This study identified that
influence depth is a function of both contact area and applied stress, with greater contact
area and surface contact pressures resulting in increased depths of influence. A key
limitation of this study, given the definition of influence depth adopted, was that the
surface contact stresses modelled for impact rolling were not verified using field test
results. Significantly, Kim’s analysis illustrated stress wave propagation to depths much
greater than those typically influenced by static loading. Nazhat (2013) analysed the
behaviour of sand subjected to dynamic loading, and identified compaction shock bands
via the use of high-speed photography and image correlation techniques from
52 |
ADE | The Impact of Rolling Dynamic Compaction
laboratory-based testing. As explained by Nazhat (2013), it is evident that
improvements in the ability to measure and quantify dynamic effects are helping to
increase knowledge of unseen processes beneath the ground surface; however, it is clear
that more research is needed to fully understand the kinematic behaviour of soils
subjected to dynamic loading.
3.3 Dynamic Compaction
Dynamic compaction is a ground improvement technique that usually employs a large
crane to lift a heavy tamper, which is then dropped onto the ground in a regular grid
pattern. Menard and Broise (1975) improved the mechanical characteristics of fine
saturated sands using this method, and were the first to propose a relationship between
the thickness to be compacted, D, the pounder mass, m, and the drop height, h, as given
by Eq. (3.1).
𝐷 = √𝑚ℎ [3.1]
Menard and Broise (1975) observed that greater depths of improvement could be
achieved for partially immersed soils than for soils completely out of water. The initial
density and grading were factors that influenced the time taken to reach a liquefied
state, after which the low frequency, high amplitude vibrations from dynamic
compaction caused the sand particles to be reorganised into a more dense state. In
subsequent years, this theory was applied to a wider range of soils conditions, including
unsaturated soils, where it was found that in many cases the maximum depth of
influence was found to be less than that predicted by Eq. (3.1). A number of different
authors, including Leonards et al. (1980), Lukas (1980), Charles et al. (1981) and Lukas
(1995) investigated the variation of an empirical factor (n) with different soil conditions
and for varying drop heights, h, and pounder masses, m. The general consensus is that n
varies with different soil conditions, with lower values for fine-grained soils and larger
values for coarse-grained soils, resulting in varying estimations for the depth of
improvement, as per Eq. (3.2).
53 |
ADE | The Impact of Rolling Dynamic Compaction
𝐷 = 𝑛 √𝑚ℎ [3.2]
Alternatively, Eq. (3.2) can be re-written as shown in Eq. (3.3). In this form, the right-
hand side of the equation is a function of gravitational potential energy, mgh, and the
material characteristics, described by the parameter n.
𝑛2
𝐷 = √ (𝑚𝑔ℎ) [3.3]
𝑔
The value of n was investigated in detail by Mayne et al. (1984) who collated data from
over 120 sites and found that n typically varied between 0.3-0.8, but could be as high as
1.0 in some instances. As explained by Mayne et al. (1984) and Lukas (1995) the
variation in predicted depth of improvement is not simply a function of the tamper
weight and drop height, but is also influenced by other variables such as tamper surface
area, total energy applied, contact pressure of the tamper, efficiency of the dropping
mechanism, initial soil conditions and ground water levels.
Applying Eq. (3.2) to the range of plotted values for n (0.3-0.8) in Mayne et al. (1984)
to an 8-tonne, 4-sided impact roller, using the maximum physical drop height of the
module that is available on a flat surface (h=0.15 m), the depth of improvement
predicted would be in the range of 0.33-0.88 m. Hamidi et al. (2009) applied Eq. (3.2)
to RDC and indicated that the use of this equation was subject to controversy as larger
depths of improvement have been reported. Table 3.1 confirms the use of dynamic
compaction formulae as under-estimating the improvement depths that are achievable
using RDC. Whilst the application of deep dynamic compaction theory to RDC without
modification is not suitable, the use of a more appropriate n value does warrant further
investigation, as both dynamic compaction theory and Table 3.1 indicate that soil type is
a key variable that influences the depth of improvement.
For dynamic compaction applications, Slocombe (2004) defines the ‘effective depth of
influence’ as being the maximum depth at which significant improvement is
measurable. The ‘zone of major improvement’ is typically half to two-thirds of the
54 |
ADE | The Impact of Rolling Dynamic Compaction
effective depth of influence. As explained by Slocombe (2004), the aforementioned
terms have been adopted in the United Kingdom but may have alternative meanings in
different parts of the world.
Impact rolling is routinely undertaken in unsaturated soils, whereby the application of
mechanical energy expels air from the voids to reduce the void ratio. Within the
influence depth of RDC, repeated loading induced stresses imparted into a granular soil
are sufficient to cause a permanent rearrangement of soil particles, resulting in increased
density and soil settlement. Below the influence depth, the soil remains elastic and does
not undergo volume change. Berry (2001) developed an elastoplastic model to
determine the depth to which there was permanent deformation using surface settlement
as the main input parameter. Whilst Berry’s model did not quantify the energy to
achieve a particular surface settlement, it was observed that a depth of 3 times the
module width was considered appropriate for a 3-sided impact roller. At sites with a
shallow water table, it is possible for the high amplitude and low frequency vibrations
associated with RDC to induce pore pressures to rise to the surface. In order to prevent
liquefaction from occurring the number of passes is typically limited to allow pore
water pressures to dissipate. Rather than competing with, impact rollers are often used
to complement deeper ground improvement techniques that leave soils within the top
2 m of the surface in a disturbed and weakened condition. Avsar et al. (2006) describe
an example of a large land reclamation project whereby impact rolling successfully
complemented deeper ground improvement techniques.
In this paper, the depth to which RDC improves the ground is measured in full-scale
field trials in homogeneous soil conditions. The measured data are compared to
predictions based on dynamic compaction theory to determine the relevance of this
approach to RDC applications.
3.4 Field trial to determine depth of improvement
A field trial was conducted using a Broons BH-1300 (8-tonne) 4-sided impact roller as
shown in Fig. 3.1 at the Iron Duke Mine located on the Eyre Peninsula in South
Australia during June 2011. The test pad was constructed in three separate lifts as
illustrated in Fig. 3.2, which also shows the locations of embedded EPCs in plan and
elevation. The test pad was constructed using haul trucks, end tipping loose tailings
material in stockpiles where a loader and excavator subsequently spread the material
55 |
ADE | The Impact of Rolling Dynamic Compaction
and liquid limit 22%). According to the Unified Soil Classification System (USCS), the
fill material used for this compaction trial could be described as a well-graded sand
(SW).
Table 3.3: Particle size distribution, compaction and field moisture test results
Material d50 Gravel Sand Fines Standard Standard FMC Modified Modified
(mm) (%) (%) (%) optimum maximum (%) OMC (%) maximum
moisture dry unit dry unit
content weight weight
(OMC) (%) (kN/m3) (kN/m3)
Magnetite
0.7 14 80 6 6.6 23.9 5.1 5.7 25.8
tailings
d particle size at per cent finer of 50%.
50,
3.4.2 Earth pressure cells
Four Geokon Model 3500 (230 mm diameter, 6 mm thick) earth pressure cells (EPCs)
were used to measure the dynamic pressures imparted by RDC. As shown in Fig. 3.2,
the initial lift (1,200 mm thick containing buried EPC1 and EPC2) was first compacted,
this was repeated for the second lift of 1,530 mm (containing EPC3) and the third and
final lift (1,460 mm containing EPC4). In plan, the EPCs were placed one-half of one
rotation of the roller apart (2.9 m) from each other in the forward direction of travel.
The EPCs were connected to a bespoke data acquisition system and Labview software
program (National Instruments, 2019). A sampling frequency of 2 kHz (i.e. one sample
every 0.0005 seconds) was adopted to capture sudden increases in pressure caused by
the module impacting the ground. Prior to compaction the EPCs were used to measure
the self-weight of the impact rolling module for the roller in an ‘at rest’ condition,
centered above each EPC. The measured pressures were compared to predictions using
Fadum’s chart (Fadum, 1948) using elastic theory, the results of which are shown in
Fig. 3.3. The measured pressures follow the same general trend, but are less than the
predicted pressures; the difference between the predicted and measured values is an
average of 38% over the depths measured. The most likely explanation for this is that
the non-uniform shape of the module face impacting the ground does not produce a
uniform pressure distribution, this is exacerbated for shallow EPC depths. A towing
speed of 10.5 km/h was selected for all 16 passes that were conducted on each layer.
58 |
ADE | The Impact of Rolling Dynamic Compaction
The staged construction process resulted in the dynamic pressure imparted by RDC to
be measured at 9 different depths.
Figure 3.3: Measured and predicted pressures versus depth for impact roller at rest
3.4.3 In situ testing
Various in situ testing methods were performed after 0, 8 and 16 passes to quantify soil
improvement with increasing compactive effort. The in situ tests were undertaken in the
centre of Lane A in layer 3 as shown in Fig. 3.2. The tests conducted included field
density measurements (ASTM 2008), the spectral analysis of surface waves (SASW)
geophysical technique and dynamic cone penetration tests (DCPs) to measure and infer
changes in density as a function of the number of module passes. SASW testing was
conducted using a GDS Surface Wave System using six 4.5 Hz geophones spaced at
1 m intervals with a sledge hammer source impacting a metal strike plate 1 m from the
first geophone. DCP testing was undertaken in accordance with the procedure described
59 |
ADE | The Impact of Rolling Dynamic Compaction
in AS 1289.6.3.3 (Australian Standards, (SA, 1997)). Verification of RDC was also
undertaken using settlement monitoring to quantify the change in ground surface level
with the number of passes. This was achieved using a level and staff to measure
settlement at 9 points across the test pad in adjacent low points in the undulating
surface, as is the normal practice. Due to space constraints, a discussion of testing
methods generally employed to verify RDC is not presented here. This is, however,
discussed in detail by Avalle and Grounds (2004) and Scott and Jaksa (2008).
3.5 Results of field trial
This section provides details of the results obtained from the field trial; specifically
those obtained from the EPCs, in situ and geophysical testing, and settlement
monitoring.
3.5.1 Earth pressure cell data
Fig. 3.4 illustrates the results obtained for a typical pass of the impact roller traversing
over the first lift of the test pad, where EPCs 1 and 2 were buried at depths of 0.67 m
and 0.87 m, respectively. As expected, the shallower EPC recorded the greatest
pressure. Fig. 3.5 presents the variation of measured peak pressure with depth, where it
is observed that peak pressures greater than 100 kPa were recorded at depths above
2.0 m. The EPC results generally supported other test data that indicated that most of
the quantifiable ground improvement occurred within 2 m of the surface. Even the
deepest EPC (buried at a depth of 3.85 m below the ground surface) registered positive
pressure readings due to the impact roller, suggesting that the depth to which RDC had
an influence extended beyond this depth. Whilst the fitted trend line illustrates a good fit
to the measured data, extrapolating for shallower than the measured depths is not
recommended. A limitation of using EPCs, is that they should not be placed at, or close
to the ground surface due to the high probability of damaging the sensors, with the
manufacturer’s guidelines recommending that no heavy equipment be used over the
cells unless at least 500 mm of material is placed above them (Geokon, 2007). Fig. 3.6
illustrates the measured peak pressures, plotted on a log scale, that were recorded by
each EPC as the impact roller traversed directly above (lane A) and in the lanes adjacent
to the buried EPCs, representing lateral offset distances of 2.5 m and 5.0 m. For a lateral
offset of 2.5 m, a maximum peak pressure was measured at a depth of 2.0 m. For a
60 |
ADE | The Impact of Rolling Dynamic Compaction
in density, or reduction in void ratio, is evident, regardless of magnitude. To determine
this, predictive models such as that proposed by Berry (2001) could be adopted;
applying this theory to the 4-sided roller yields an influence depth of 3.9 m.
Alternatively, sensitive measuring equipment, such as EPCs, or intrusive site
investigation techniques, such as the cone penetration test and dilatometer test could be
used.
Here, no attempt is made to quantify the depth to which RDC has a small positive
influence. Instead, an energy-based approach is proposed to provide estimations for
depths capable of being significantly improved in situ, and layer thicknesses capable of
being compacted by RDC. Gravitational potential energy forms part of the total energy
imparted to the ground. Other factors include the potential energy due to the double-
spring-linkage system, and the kinetic energy due to friction between the soil and
module interface. The effects of the double-spring-linkage system can be quantified via
a change in module velocity, hence considered part of the kinetic energy component
delivered by the impact roller. For the towing speed adopted in the field trial reported in
this paper, the change in potential and kinetic energies are listed in Table 3.4.
Table 3.4: Predicted change in potential and kinetic energies for a towing speed of 10.5 km/h
v v vi vf mgh DKE
(km/h) (m/s) (m/s) (m/s) (kJ) (kJ)
10.5 2.92 3.21 2.63 11.8 13.6
Note:v = speed of towing unit; v = module velocity after impacting the ground; v = module velocity
f i
prior to impacting the ground.
The second definition is applicable when improving ground in situ; in such cases,
depths shallower than the maximum capable by RDC are typically targeted for
improvement. Working within the limitations of RDC ensures that quantifiable
improvement occurs and the properties of the ground are improved such that a specified
target criterion is met. The concept of an effective depth of improvement (EDI) is most
relevant for applications involving improving ground in situ (as per the case studies
referenced in Table 3.1). The EDI can be considered as the equivalent of the term
described by Slocombe (2004) for dynamic compaction, being the maximum depth to
67 |
ADE | The Impact of Rolling Dynamic Compaction
which significant improvement occurs. As shown in Eq. (3.4), new parameter EDI is
calculated as the product of Eq. (3.2) (based on module mass, m, lift height, h, and
empirical factor n from dynamic compaction theory), and a new term k, defined as the
ratio of the energy imparted to the ground divided by the gravitational potential energy,
as calculated in Table 3.5.
Table 3.5: Values of k for different towing speeds based on change in potential and kinetic energies
v mgh DKE mgh + DKE k
(km/h) (kJ) (kJ) (kJ)
9 11.8 10.0 21.8 1.8
10.5 11.8 13.6 25.4 2.2
12 11.8 17.8 29.6 2.5
Note:v = speed of towing unit; k = ratio of the energy imparted to the ground divided by gravitational
potential energy.
𝐸𝐷𝐼 = 𝑘(𝑛 √𝑚ℎ) [3.4]
Alternatively, Eq. (3.4) can be re-written as shown in Eq. (3.5). In this form, EDI is
written in terms of the material characteristics, n, gravitational potential energy, mgh
and a variable k, which depends upon the towing speed, as per Table 3.5.
𝑘2𝑛2
𝐸𝐷𝐼 = √ (𝑚𝑔ℎ) [3.5]
𝑔
Third, for determining the maximum layer thickness that can be compacted in thick
lifts, the concept of a depth of major improvement (DMI) is appropriate. This applies to
situations where a target criterion that is comparable to what can be achieved by
conventional compaction equipment in thin lifts is required. Consistent with the
description adopted by Slocombe (2004) to determine the zone of major improvement
68 |
ADE | The Impact of Rolling Dynamic Compaction
For the field trial described in this paper, RDC was measured to have an influence at a
depth of 3.85 m; however, the majority of improvement occurred within the top 2.0 m
from the surface, consistent with the definition of EDI. Whilst RDC improves soil
beneath this so-called effective depth, for a uniform soil profile, the magnitude of
improvement beyond this depth is less significant. A maximum dry density ratio of 95%
with respect to modified compaction was obtained for a layer thickness of 1.2 m (DMI).
The values for EDI and DMI obtained are consistent with Table 3.6 for an n value of
0.8, reasonable for granular soils, and a k value of 2.2, consistent for the 10.5 km/h
towing speed adopted in the trial. Table 3.6 suggests that the depths to which RDC can
improve and compact granular soils is influenced more by towing speed than for clay
soils. However, not all ground conditions can sustain a towing speed of 12 km/h for the
8-tonne 4-sided impact roller; therefore in the absence of site specific information, a
median towing speed of 10.5 km/h is recommended for use in Table 3.6.
3.7 Conclusions
This paper has examined improving ground in situ and compaction of soil in thick
layers as they are two distinctly different applications for RDC that, in the authors’
opinion, need to be treated independently. For a towing speed of 10.5 km/h for the
8-tonne 4-sided impact roller, the effective depth of improvement, EDI, is estimated to
be 0.73 m for clay soils (n = 0.3), and 1.94 m for granular soils (n = 0.8). This highlights
that soil type is the single most important variable in quantifying the depth to which
RDC can improve soil. A relationship to evaluate EDI is presented as a function of the
energy imparted to the ground by RDC, and is appropriate for determining the depths to
which ground can be improved in situ. For the field trial presented in this paper, an EDI
of 2.0 m was measured using buried EPCs and complementary in situ testing.
A second relationship to determine the depth of major improvement, DMI, is also
introduced, and is appropriate for determining the thickness of layers that can be
compacted using RDC, typically half to two thirds of EDI. For the field trial presented
in this paper, a DMI of 1.2 m was measured using in situ testing. The equations
presented in this paper augment the relationship for dynamic compaction first proposed
by Menard and Broise (1975). In addition to soil type, module mass and drop height, the
equations presented also incorporate the effect of towing speed. Whilst the equations
presented in this paper are relatively simple in nature, the proposed energy-based
approach yields estimations for depths capable of being significantly improved in situ,
70 |
ADE | The Impact of Rolling Dynamic Compaction
and layer thicknesses capable of being compacted by RDC, that are in broad agreement
with the findings of the field trial presented, and the results of published case studies
involving the 8-tonne 4-sided impact roller over the past four decades.
3.8 Acknowledgements
The authors wish to acknowledge the following final year civil engineering students
from the University of Adelaide who contributed to the field work referred to in this
paper: Nicole Mentha, Simon Pointon, Erfan Syamsuddin, Aidan Symons and
Penelope Wrightson. The authors are grateful to the instrumentation and laboratory staff
at the University of Adelaide who have provided invaluable support over the years and
to Cathy Cates for her assistance with word processing. The authors are also grateful to
Broons and HWE Mining personnel for access to equipment and the test site, without
their help and support, the work undertaken in the field trial featured in this paper would
not have been possible.
3.9 References
ASTM (2007) D 698: Standard test methods for laboratory compaction characteristics
of soil using standard effort. ASTM International, West Conshohocken, PA, USA.
ASTM (2008) D 5195: Standard test methods for density of soil and rock in-place at
depths below surface by nuclear methods. ASTM International, West Conshohocken,
PA, USA.
ASTM (2009a) D 6913: Standard test methods for particle-size distribution (gradation)
of soils using sieve analysis. ASTM International, West Conshohocken, PA, USA.
ASTM (2009b) D 1557: Standard test methods for laboratory compaction
characteristics of soil using modified effort. ASTM International, West Conshohocken,
PA, USA.
ASTM (2010a) D 2216: Standard test methods for laboratory determination of water
(moisture) content of soil and rock by mass. ASTM International, West Conshohocken,
PA, USA.
ASTM (2010b) D 4318: Standard test methods for liquid limit, plastic limit, and
plasticity index of soils. ASTM International, West Conshohocken, PA, USA.
71 |
ADE | The Impact of Rolling Dynamic Compaction
Avalle DL (2004) Ground improvement using the “square” impact roller – case studies.
In Proceedings of the 5th International Conference on Ground Improvement Techniques,
Kuala Lumpur, Malaysia, pp. 101-108.
Avalle DL (2007) Trials and validation of deep compaction using the “square” impact
roller. Symposium Advances in Earthworks, Australian Geomechanics Society, Sydney,
Australia, pp. 63-69.
Avalle DL and Carter JP (2005) Evaluating the improvement from impact rolling on
sand. In Proceedings of the 6th International Conference on Ground Improvement
Techniques, Coimbra, Portugal, pp. 153-160.
Avalle D and Grounds R (2004) Improving pavement subgrade with the "square"
impact roller. 23rd Annual Southern African Transport Conference, Pretoria, South
Africa, pp. 44-54.
Avalle DL and Mackenzie RW (2005) Ground improvement of landfill site using the
“square” impact roller. Australian Geomechanics, 40(4): 15-21.
Avalle D and Young G (2004) Trial programme and recent use of the impact roller in
Sydney. Earthworks Seminar, Australian Geomechanics Society, Adelaide, Australia,
5pp.
Avsar S, Bakker M, Bartholomeeusen G and Vanmechelen, J. (2006) Six sigma quality
improvement of compaction at the New Doha International Airport Project. Terra et
aqua, 103: 14-22.
Berry AD (2001) Development of a volumetric strain influence ground improvement
prediction model with special reference to impact compaction. Dissertation submitted in
partial fulfilment of the requirements of the degree of Master of Engineering, University
of Pretoria, Pretoria, South Africa.
Charles JA, Burford D and Watts KS (1981) Field studies of the effectiveness of
dynamic consolidation. In Proceedings of the 10th International Conference on Soil
Mechanics and Foundation Engineering, Stockholm, Sweden, 3: 617-622.
Clifford JM (1975) An introduction to impact rollers. National Institute for Road
Research, Internal report RP/2/75, South Africa, pp. 1-30.
72 |
ADE | The Impact of Rolling Dynamic Compaction
Clifford JM (1978) Evaluation of Compaction Plant and Methods for the Construction
of Earthworks in Southern Africa. Dissertation in partial fulfilment of the requirements
for the degree of Master of Science in Engineering, University of Natal, Durban, South
Africa.
Clifford JM (1980) The development and use of impact rollers in the construction of
earthworks in southern Africa. CSIR Report 373, National Institute for Transport and
Road Research Bulletin 16, Pretoria, South Africa, pp. 1-53.
Clifford JM and Coetzee SD (1987) Coal and discard stockpiling with an impact roller.
In Proceedings of the International Conference on Mining and Industrial Waste
Management, Johannesburg, South Africa, pp. 75-79.
Ellis CI (1979) Pavement engineering in developing countries. Transport and Road
Research Laboratory, TRRL Supplementary Report 537, Crowthorne, Berkshire, 35pp.
Fadum RE (1948) Influence values for estimating stresses in elastic foundations. In
Proceedings of the 2nd International Conference ISSMFE, Rotterdam, Netherlands, 3:
77-84.
Geokon (2007) Instruction Manual Model 3500, 3510, 3515, 3600 Earth Pressure Cells.
Geokon Incorporated, Lebanon, USA, 24pp.
Hamidi B, Nikraz H and Varaksin S (2009) A review of impact oriented ground
improvement techniques. Australian Geomechanics 44(2): 17-24.
Kim K (2010) Numerical simulation of impact rollers for estimating the influence depth
of soil compaction. Dissertation in partial fulfilment of the requirements for the degree
of Master of Science, Texas A&M University, Texas, USA.
Leonards GA, Holtz RD and Cutter WA (1980) Dynamic compaction of granular soils.
Journal of Geotechnical Engineering Division ASCE 106(1): 35-44.
Lukas RG (1980) Densification of loose deposits by pounding. Journal of Geotechnical
Engineering Division ASCE 106(4): 435-446.
73 |
ADE | The Impact of Rolling Dynamic Compaction
Ground response to rolling dynamic compaction (Paper 3)
Abstract
Rolling dynamic compaction (RDC) is typically used for improving ground in situ or
compacting fill in thick lifts. In many project applications, the effects of RDC are
verified via testing that is undertaken pre- and/or post-compaction. This study presents
results from a full-scale field trial that involved placing an earth pressure cell (EPC) and
accelerometers at a depth of 0.7 m within a 1.5 m thick layer of homogeneous sandy
gravel to measure the response to RDC in real-time. Double integration of acceleration-
time data enabled settlement to be inferred, whilst the EPC measured the change in
stress due to impact. The maximum change in vertical stress recorded over the 80 passes
undertaken was approximately 1,100 kPa. During a typical module impact, the loading
and unloading response occurred over a duration of approximately 0.05 seconds. The
acceleration response of RDC was measured in three orthogonal directions, with the
vertical accelerations dominant.
List of notation
d particle size at per cent finer of 50%
50
g free-fall acceleration (9.81 m/s2)
W elastic work done (energy imparted to ground that is recovered elastically)
elastic
W plastic work done (energy imparted in ground that causes permanent settlement)
plastic
W total work done (area under the load-displacement curve) = W +W
total elastic plastic
Dt duration of applied load
elastic (rebound) settlement
elastic
plastic (permanent) settlement
plastic
4.1 Introduction
Rolling dynamic compaction (RDC) imparts energy to the ground through the use of a
heavy, non-circular module that falls to impact the ground. A limitation of many past
field investigations to verify the effects of RDC is that testing is typically performed
pre- and/or post-compaction. Such investigations often serve their intended purpose for
determining if a project specification has been met (or otherwise) but they do not
81 |
ADE | The Impact of Rolling Dynamic Compaction
capture the dynamic effects of a heavy module impacting the ground in real-time. This
study has used a buried earth pressure cell (EPC) and accelerometers to better
understand the ground response beneath the surface during the passage of an impact
roller.
4.2 Background
Impact rollers with different module masses, shapes and drop heights, have been
compared to predict the energy imparted into the ground (McCann 2015). A limitation
with such a prediction is that a RDC module that is towed across the surface impacts the
ground in a different manner to a dynamic compaction pounder that is a function of
mass, drop height and vertical acceleration due to gravity. A RDC module, shown in
Fig. 4.1, impacts the ground in a similar way to a falling hinged trap-door; the geometry
and surface area of the module that is in contact with the ground is non-uniform; as is
the impact velocity of the module when it contacts the ground.
In RDC applications, accelerometers have been placed on an impact roller to measure
the ground surface response. Heyns (1998), and McCann and Schofield (2007) both
noted that an increase in the magnitude of decelerations is commonly measured with
increasing passes, as the surface soil stiffness increases. This finding is consistent with
the work of Clifford (1978), who observed that the module drop height increases as the
ground surface becomes harder; the cross-sectional area of the module that is in contact
with the ground changes with drop height due to the geometry of the rounded corners
and how far they embed in the ground. The energy imparted by the roller is spread over
a smaller area as the stiffness of the surficial soil increases; this results in greater contact
pressures being imparted to the soil with increasing passes. The use of module mounted
accelerometers has proven useful in identifying less stiff near-surface soils that typically
exhibit lower decelerations (McCann and Schofield 2007); however, there is no
guarantee that measuring the response of an impact roller as it passes over the ground
surface gives a true indication of the soil response below the surface. Inferring
improvement due to RDC from surface measurements can be challenging given RDC
typically disturbs the near-surface soils, which can be further complicated by sites
containing inherent soil variability. Mooney and Rinehart (2007) carried out a field
investigation using a smooth drum vibrating roller. They performed multiple passes
across test areas comprising both heterogeneous and homogeneous soils. They found
that soil heterogeneity presented significant challenges for interpreting instrumented
82 |
ADE | The Impact of Rolling Dynamic Compaction
roller data. This study overcomes previous limitations by attaching accelerometers to an
EPC and burying them in homogeneous fill material to quantify the loading induced
stress and ground deceleration beneath the ground surface, yet within the expected zone
of influence of RDC.
4.2.1 Comparisons with dynamic compaction
Measuring the ground response of deep dynamic compaction has been studied by
Mayne and Jones (1983), who attached an accelerometer to a 20.9 tonne pounder to
monitor the deceleration upon impact with the ground surface after falling a distance of
18.3 m; the deceleration-time response of the impact blow occurred over a duration of
only 0.05 seconds. Also of significance, the magnitude of decelerations recorded were
in the order of 70-85g, and a trend of increasing magnitude with number of drops was
observed. Clegg (1980) attached an accelerometer to a falling weight and found that the
peak deceleration of the weight upon impact with the soil was directly related to the soil
resistance, described as a combination of both soil stiffness and shearing resistance.
Chow et al. (1990) developed a theoretical framework that was based on matching
deceleration measurements of a dynamic compaction pounder impacting the ground
using an accelerometer that was attached to the pounder near the centre of gravity. The
one-dimensional model that was developed was similar to pile driving analyses where
the impact velocity was obtained by integrating measured decelerations. Yu (2004)
double integrated the acceleration-time response of a vertically falling plate to generate
the load-displacement relationship, which was integrated to quantify the work done.
Analysis of a load-displacement response due to impact was also undertaken by Jha et
al. (2012) who investigated energy dissipation to quantify the elastic energy that was
recovered during unloading of multi-phase cementitious materials. They plotted the
load-displacement response for cementitious materials subjected to nano-indentation
and determined the area under the loading and unloading curves and quantified the work
done. Key aims of this study are to measure the loading induced stresses and
displacements that soil particles beneath the ground surface experience, and to quantify
the work done from measured force-displacement data.
4.3 Research test site
Fig. 4.1 shows a 4-sided 8-tonne impact roller (1,450 mm square and 1,300 mm wide
module) that was used at a dedicated research site located at Monarto Quarries,
83 |
ADE | The Impact of Rolling Dynamic Compaction
approximately 60 km south-east of Adelaide, Australia. Whilst conducting a full-scale
trial that is not associated with a client funded project is expensive, a research focussed
trial provided an ability to control a number of variables that can often conceal the true
effects of RDC. Significantly, natural soil was excavated to a depth of 1.5 m and
replaced with homogeneous fill; a crushed rock with a maximum particle size of 20 mm
that was readily available and locally produced at the quarry. Six equal lifts of 250 mm
thickness were adopted; the material was placed using a Volvo L150E Loader, and was
lightly compacted using a 60 kg vibrating plate and wheel rolling from the loader. The
fill material was classified as a well-graded Sandy Gravel (GW) in accordance with the
Unified Soil Classification System. The fill was tested for homogeneity through the use
of particle size distribution and Proctor compaction testing; the results are given in
Table 4.1.
Figure 4.1: 8-tonne 4-sided impact roller.
84 |
ADE | The Impact of Rolling Dynamic Compaction
Table 4.1: Particle size distribution, compaction and field moisture test results of fill material
Gravel Sand Standard Standard Modified Modified
Material d50 size size Fines OMC MDD FMC OMC MDD
(mm) (%) (%) (%) (%) (kN/m3) (%) (%) (kN/m3)
20 mm
crushed 4.0 57 40 3 7.9 17.9 8.6 7.2 18.9
rock
Note:d = particle size at percent finer of 50%; OMC = optimum moisture content; MDD = maximum
50
dry density; FMC = field moisture content.
4.3.1 Earth pressure cells and accelerometers
Field trials undertaken by Avalle et al. (2009) and Scott et al. (2016) using the 4-sided
impact roller have shown that a module impacting the ground directly above embedded
instrumentation results in significantly higher ground decelerations being recorded,
compared to when the module strikes the ground off-set from embedded
instrumentation. A limitation of burying equipment at discrete locations is that it is not
possible to capture the maximum ground response from every impact. However, a key
advantage of this technique is that it does provide real-time data on dynamic pressures
and accelerations that are imparted into the ground that other testing methods are unable
to do.
A custom-built accelerometer cluster consisting of ±5g and ±16g accelerometers in the
Z-plane to measure vertical acceleration, and ±5g accelerometers in the X- and Y-
planes, to measure tilt perpendicular to, and in the direction of travel, respectively. A
total of 80 passes were undertaken. The accelerometer cluster was attached to an EPC
(230 mm diameter and 6 mm thick) that was buried at a depth of 0.7 m below the
ground surface, and connected to a bespoke data acquisition system and Labview
software program (refer Labview (2018)). The ability to capture an accurate ground
response using EPCs and accelerometers relies heavily on adopting a sufficiently high
sampling frequency. Given that displacement is to be quantified from the double
integration of acceleration-time data, a sampling frequency of 4 kHz (twice that adopted
by Avalle et al., 2009) was selected for this trial to ensure that the true peak pressure
and ground deceleration could be accurately captured. As discussed by Thong et al.
(2002), faster sampling rates can improve the accuracy of integration, but errors can
increase with the duration of the time interval over which integration is undertaken.
85 |
ADE | The Impact of Rolling Dynamic Compaction
4.4 Results and discussion
A single pass (No. 54 summarised in Table 4.2) was selected out of the 80 passes
undertaken for analysis as it featured a high peak pressure and the largest vertical
deceleration recorded. In Fig. 4.2 the module impact resulted in a measured peak
pressure of 1,077 kPa at a depth of 0.7 m. It can be observed that the impulse pressure
imparted to the ground was loaded and unloaded over a duration of approximately
0.05 seconds. Fig. 4.3 illustrates the vertical (Z-) acceleration-time response for the
same pass shown in Fig. 4.2, whereby a downward (negative) acceleration first occurs
as the soil is loaded. In response to loading, the soil resistance is mobilised, which
results in an upward acceleration before the acceleration trace dampens and returned to
zero less than 0.1 seconds after loading. Significantly, a peak deceleration (negative
acceleration) of 21g was measured before the soil resistance was mobilised. Table 4.3
includes a summary of passes 1-10, as well as every fifth pass thereafter. As observed in
Table 4.3, the magnitude of the peak downward acceleration was typically greater than
the peak upward acceleration, this trend was more defined for impacts that generated
large accelerations. Consequently, a shift in the baseline (zero) reading was undertaken
that enabled readings of –21g and +6.3g to be measured using a ±16g accelerometer
(range of 32g). Consistent with the findings of Mayne and Jones (1983), an increased
number of passes generally resulted in larger accelerations (and peak pressures) being
recorded. However, the variable location of the module landing on the ground surface
relative to buried instrumentation, analysed and discussed by Scott et al. (2016), was
also a contributing factor that would explain why some passes (e.g. pass 54) yielded
much larger peak pressures and vertical accelerations than others.
86 |
ADE | The Impact of Rolling Dynamic Compaction
Table 4.2: Summary of Pass No. 54 for test depth of 0.7 m
Peak Peak
Pass elastic plastic Wtotal Welastic Wplastic Peak Pressure Dt Dec. Acc.
No. (mm) (mm) (J) (J) (J) (kPa) (s)
(g) (g)
54 4 5 254 36 218 1077 0.05 -21.0 6.3
Note: = rebound settlement; = permanent settlement; W = total area under load-
elastic plastic total
displacement curve; W = elastic work done; W = plastic work done; Dt = duration of applied
elastic plastic
load; Peak Dec. = peak deceleration; Peak Acc. = peak acceleration.
Table 4.3: Summary of passes for test depth of 0.7 m
Impulse
P Na os .s ( mela msti )c (mpla msti )c Wtotal (J) W (e Jla )s tic W (p Jla )s tic Pr (P e ke s Psa auk )r e Dt DP ece .a (k
g )
AP cce .a (k
g )
(s)
1 2.0 0.5 13 9 4 230 0.07 -3.5 3.0
2 3 1 44 13 31 419 0.07 -5.5 3.8
3 3.5 0.5 35 25 10 371 0.08 -5.3 4.4
4 3 2 76 20 56 594 0.08 -4.6 2.5
5 6.5 0 108 53 55 656 0.07 -5.6 7.7
6 3 2 71 13 58 503 0.06 -11.6 5.2
7 3 2 64 20 44 550 0.08 -2.1 3.4
8 1 1 73 45 28 177 0.08 -1.3 0.6
9 2 1 22 6 16 258 0.05 -4.9 2.8
10 3 2 71 14 57 539 0.06 -8.5 3.9
15 3 2 56 15 41 490 0.08 -4.0 1.7
20 3 2 62 18 44 492 0.05 -9.6 4.8
25 2.5 1.5 35 14 21 324 0.06 -8.0 4.7
30 6 0.5 58 29 29 380 0.06 -10.5 9.6
35 2.5 1 22 7 15 272 0.05 -4.0 2.9
40 2 3 41 5 36 309 0.04 -6.6 4.4
45 2.5 0.5 12 4 8 166 0.05 -1.6 2.6
50 2 1 11 7 4 202 0.06 -1.8 1.7
55 3.5 2.5 98 24 74 680 0.05 -7.2 5.6
60 2.5 0.5 11 7 4 169 0.07 -2.4 2.5
65 3.5 3.5 177 14 163 873 0.05 -13.2 5.4
70 4 1.5 60 34 26 557 0.07 -4.9 3.8
75 1.5 6 136 18 118 731 0.07 -9.2 4.5
80 7.5 0.5 249 59 190 1115 0.05 -11.2 8.0
Note: = rebound settlement; = permanent settlement; W = total area under load-
elastic plastic total
displacement curve; W = elastic work done; W = plastic work done; Dt = duration of applied
elastic plastic
load; Peak Dec. = peak deceleration; Peak Acc. = peak acceleration, peak values in bold.
88 |
ADE | The Impact of Rolling Dynamic Compaction
Figure 4.5: X-acceleration response at time of module impact
Fig. 4.6 shows the variation of Z-acceleration and Z-displacement of the soil with time
in response to a single module impact, whereby displacement was calculated from
double integration of the acceleration-time response. From Fig. 4.6, it is evident that
approximately 9 mm total displacement occurred due to loading; however, upon
unloading, the permanent displacement due to the single impact was 5 mm. The same
impact blow is illustrated in Fig. 4.7, which shows the loading and unloading response
of the soil due to a single pass of the impact roller at a measured depth of 0.7 metres
beneath the ground surface. Force is determined by adopting the peak pressure at the
time of impact and multiplying it by the plan area of the EPC. Displacement is
evaluated from double integration of the acceleration-time response. In Fig. 4.7 the
portion of the curve between points A and B represents the loading of the soil. The
unloading portion of the curve is shown between points B and C. The distance between
90 |
ADE | The Impact of Rolling Dynamic Compaction
Figure 4.8: Force-displacement curves for consecutive passes (Pass Nos. 1-10)
4.5 Conclusions
To minimise soil variability, this study has captured the change in vertical stress due to
RDC at a depth of 0.7 m beneath the surface using an earth pressure cell buried in a
1.5 m thick layer of homogeneous sandy gravel. The maximum change in vertical stress
recorded over the 80 passes undertaken was approximately 1,100 kPa. During a typical
module impact, the loading and unloading of the soil occurred over a duration of
roughly 0.05 seconds. The acceleration response of a single module impact was also
measured in three orthogonal directions at 0.7 m depth, with the vertical accelerations
dominant. In project applications, there is typically a trade-off between layer thickness
and the number of passes required to significantly improve ground to meet a certain
specified criterion. Whilst the number of passes (80) undertaken in this study was
greater than what would economically be undertaken in practice, the results from buried
instrumentation indicate that 0.7 m is well within the depth range that can be
significantly improved by RDC. Quantifying the dynamic behaviour of the soil beneath
93 |
ADE | The Impact of Rolling Dynamic Compaction
the ground surface in real-time emphasises that the uneven module geometry results in
some passes imparting much greater pressure to the ground that others, this being a key
reason why many passes are needed to ensure adequate coverage of a site.
4.6 Acknowledgements
The authors wish to acknowledge the following final year civil engineering students at
the University of Adelaide who assisted with the field work referred to in this paper:
Stefan Chenoweth, Jordan Colbert, Julianne Saw and Ross Vince. The authors are also
indebted to instrumentation staff at the University of Adelaide who wrote the Labview
software program and built the accelerometer clusters used in the field trial. The authors
would also like to thank Cathy Cates for her help with word processing. The authors are
grateful to Mr Stuart Bowes from Broons who provided financial and in-kind support;
and Monarto Quarries personnel, who supplied plant and quarry material. Without their
help and support, the work featured in this paper would not have been possible.
4.7 References
Avalle DL, Scott BT and Jaksa MB (2009). Ground energy and impact of rolling
dynamic compaction – results from research test site. In Proceedings 17th International
Conference on Soil Mechanics and Geotechnical Engineering, Vol. 3, Alexandria,
Egypt, (eds M. Hamza, M. Shahien and Y. El-Mossallamy), vol. 3, pp. 2228–2231.
Amsterdam, the Netherlands: IOS Press BV.
Chow YK, Yong DM, Yong KY and Lee SL (1990). Monitoring of dynamic
compaction by deceleration measurements. Computers and Geotechnics, 10(3): 189-
209.
Clegg B (1980). An impact soil test as alternative to California bearing ratio. In
Proceedings 3rd Australia-New Zealand Conference on Geomechanics, Wellington,
New Zealand: New Zealand Institution of Engineers, pp. 225-230. Wellington, New
Zealand: New Zealand Institution of Engineers.
Clifford JM (1978). Evaluation of Compaction Plant and Methods for the Construction
of Earthworks in Southern Africa. Dissertation in partial fulfilment of the requirements
for the degree of Master of Science in Engineering, University of Natal, Durban, South
Africa.
94 |
ADE | The Impact of Rolling Dynamic Compaction
Chapter 5: Conclusions
Quantifying the ground response due to RDC, as well as the effects of towing speed,
and the depth of improvement, will enable RDC to be applied and validated more
appropriately for a range of applications and soil characteristics. A greater
understanding of RDC will reduce design conservatism and construction costs; and
importantly, reduce instances where the anticipated improvement did not occur.
Furthermore, it will enable RDC to be used more effectively and with greater
confidence in a range of applications. This will inevitably lead to more accurate
assessments of the geotechnical properties of ground compacted using RDC, and hence,
the optimised design of geotechnical systems such as pavements, foundations, dams,
embankments and retaining structures.
5.1 Research Contributions
This research has added to the current knowledge of RDC and has addressed the five
objectives of this thesis as follows:
1. This research has determined that towing speed influences the stress that is
imparted into the ground. At towing speeds less than 9 km/h, the dynamic
effects of the module are not maximised, compared with towing speeds of 10-12
km/h that were found to be optimal. At towing speeds above 12 km/h, higher
stresses could occasionally be imparted into the ground, but the kinematics of
the module impacting the ground changed if it was towed too quickly, causing
the module to skip and jump from corner to corner, rather than the sides of the
module falling to impact the ground in a predictable and reproducible manner.
2. This research proposes that the energy imparted to the ground due to RDC is a
function of the work done, which is equal to the sum of the change in
gravitational potential and kinetic energies. The energy that is delivered by a
single impact is dependent upon towing speed. This research has refined the
maximum estimated energy that is imparted to ground to between 22 kJ to 30 kJ
for typical towing speeds of 9 and 12 km/h, respectively. This is in contrast to
previous estimates for the 8-tonne 4-sided impact roller that predicted values of
between 30 kJ to 54 kJ for the same towing speed range based on kinetic energy.
97 |
ADE | The Impact of Rolling Dynamic Compaction
3. Understanding the depths beneath the surface that RDC is capable of improving
in different soil conditions is a key criterion, particularly in applications for
compacting in situ material (improving poor quality ground). This research has
proposed a new term for RDC, defined as the effective depth of improvement
(EDI), giving practitioners some guidance regarding what depths can be
improved in different soil conditions. When compacting deep layers of placed
fill material; the uncertainty regarding the depth of improvement should be less
of a concern to practitioners, as the placed layer thickness can be designed to be
limited to a depth that is well within the capability of the impact roller. This
research also offers guidance as to the thickness of fill layers that can be
adopted, via the use of another new term, defined as the maximum depth of
improvement (MDI).
4. This research found that the formula for deep dynamic compaction first
proposed by Menard and Broise (1975) could not be used directly for RDC
without modification. However, modifications to this formula can be used to
estimate the improvement depths for two distinctly different applications of
RDC: improving ground in situ (using EDI), and compacting soil in thick layers
(using MDI), respectively. Whilst these relationships are simplistic and have
limitations, they provide reasonable estimates based on the field trials
undertaken in this thesis and are in broad agreement with reported case studies
that have used the 8-tonne, 4-sided impact roller over the past four decades.
5. Arguably, the greatest contribution to current knowledge from this body of
work, is from capturing and analysing the in-ground response of RDC. It is
imperative to understand how the ground responds to a single impact if there is
any chance of trying to predict and understand what will happen after multiple
impacts (as undertaken in project applications). The duration of the pressure
impulse is significant; the time over which loading and unloading of the soil
occurred was less than 0.05 seconds. The soil response due to RDC was also
captured using accelerometers. Double integration of the acceleration-time
response allowed displacements to be evaluated, whereby the loading and
unloading response of the soil due to a single pass of the impact roller enabled
recoverable (elastic) and permanent (plastic) components to be identified
separately. Force was determined by adopting the peak pressure at the time of
98 |
ADE | The Impact of Rolling Dynamic Compaction
impact and multiplying it by the plan area of the EPC. Plotting force against
displacement enabled the in-ground response of RDC to be quantified in terms
of work done, thereby giving a measure of the actual energy imparted to the
ground at the measured depth.
5.2 Limitations of current research methods and existing RDC
practices
An underlying aim of this research is to measure the in-ground response of RDC. To do
so, instrumentation must obviously be buried in the ground. The single biggest
limitation of using buried instrumentation is that it is not possible for an RDC module to
land repeatedly on the same surface location relative to buried instrumentation that is
placed at a fixed location. To overcome this issue, it was necessary to conduct many
passes to determine trends and to measure the distance between the centre of the module
and the centre of buried instrumentation (offset distance).
A large number of field test sites were used in this research, many of which were
commercial project sites, giving the author opportunities to observe RDC in a variety of
soil conditions, whilst also providing the author with large quantities of test data. What
became apparent was that project data were often sufficient for proving conformance
with a project specification; however, they were insufficient to answer specific research
questions regarding how individual factors contributed to the performance of RDC.
Inherent soil variability was the greatest issue that was masking results. At filled (or
‘made-up’) sites, this was, unsurprisingly, an even bigger issue. Despite attempts to
quantify ground improvement using closely spaced boreholes and a suite of different in
situ tests within close proximity, there was still uncertainty when comparing pre- and
post-compaction data; the difference could be attributed to either soil variability,
improvement using the impact roller, or a combination of the two factors. To isolate and
quantify the effects of RDC, it was necessary to conduct tests in uniform soil
conditions. This required significant financial support as it meant placing uniform fill
material in significant quantities at sites where the costs of fill and earthmoving plant
and equipment had to be covered. Adopting comprehensive field trials at dedicated
research testing sites enabled targeted field trials of longer duration to be undertaken
that were not possible at commercial sites where field testing programs had to be more
efficient and time effective so as to not delay other site activities.
99 |
ADE | The Impact of Rolling Dynamic Compaction
avoided. It is also counter-productive for the industry to rate impact rollers in terms of
gravitational potential energy when specifications correctly dictate towing speed ranges
that must be adopted in order for RDC to be most effective.
Some projects still adopt method specifications (this typically involves adopting a
specified number of passes and not testing the soil post compaction). In applications
where the impact roller is being used to improve in situ soil or compact thick layers, the
focus should be on meeting the requirements of a performance specification. In future,
the author predicts there will be a greater focus on quantifying the change in soil
stiffness due to RDC, leading to increased testing and understanding of changes in soil
modulus.
Impact roller module design was primarily developed during the period between the
1950s and 1970s in South Africa. As described by Clifford (1975) the impact roller was
originally developed as there was a need to rapidly compact potentially collapsing sands
in Southern Africa up to nominal depths of up to 3 metres using towing units with
approximately 160-170 horsepower. Its success in improving the density of potentially
collapsing sands resulted in the impact roller being used in different soils and
applications as time progressed. Whilst improvements to module design have occurred
since then, in recent years the development of the towing units have continued to
increase, with modern equipment having significantly greater torque and horsepower
than their 1970s counterparts. As such, the author believes there is scope to optimise
and refine module design in the future.
5.4 Future research directions
The field testing for this research project was limited to the use of the standard (8-tonne)
4-sided impact roller. Preference was given to working with the same roller for the work
conducted in this thesis; however, there is a need to test both the 8- and 12-tonne
4-sided modules at the same site in uniform soil conditions containing buried
instrumentation. It is proposed both rollers would operate at the same speed to truly
isolate the effect of module mass. Similarly, there is a need for full-scale testing of
impact rollers with different numbers of sides in the same soil conditions using buried
pressure cells and accelerometers to measure the ground response; this will provide a
greater understanding of their similarities and differences.
101 |
ADE | The Impact of Rolling Dynamic Compaction
In Chapter 3 of this thesis, there is scope to examine if there are advantages in refining
and improving Equation 3.4 to include the change in kinetic energy term, DKE.
Equation 3.4 retains the form of the original Menard and Broise (1975) equation, along
with an empirical factor n to take into account soil type (as per dynamic compaction
theory). This equation was augmented for RDC by multiplying by an energy ratio
parameter, k, which varies with towing speed and is based on an estimation of DKE.
Augmenting the original Menard and Broise (1975) equation was deliberate to allow
practitioners to infer a depth of influence based on physical parameters associated with
the module (mass m and drop height h) that can be quantified easily. It could be argued
that incorporating the DKE term into Equation 3.4 may be logical and useful; however,
the difficulty in accurately quantifying the change in kinetic energy is the reason why
DKE was not included. Further research could aim to better quantify DKE, whilst it is
recognised that there are energy losses due to friction and noise, their magnitude
remains unknown.
It is acknowledged that equations developed for RDC in this thesis are simplistic,
relying only on module mass, drop height, n value (soil type) and k value (taking into
account that the energy imparted into the ground is not solely gravitational potential
energy). The equations presented in this thesis do not include variables such as moisture
content, number of passes and contact area of module with the ground. It is hoped that
this work will inspire more refined assessments to be made in the future so that RDC
can be compared more accurately with more traditional forms of compaction that have
better predictive models. More accurate assessments of RDC will lead to greater
knowledge and better informed decisions regarding circumstances that are appropriate
for adopting this ground improvement technique, or otherwise. The effect of module
shape (number of sides) and mass requires further investigation as the k value
introduced in this paper, would not be applicable for the heavy duty (12-tonne) 4-sided
roller, nor would it be applicable for impact rollers with 3- and 5-sided modules.
Whilst it is recognised that this thesis has included case studies of RDC working in
uniform soil conditions that seldom occur in practice, it is hoped that the benefit of
controlling soil variability in full-scale field trials can provide valuable information
regarding the depths (and magnitude) that RDC can improve ground. Controlling soil
variability is easy to do in a computer model, and comparatively easy in a laboratory
environment, but neither captures the real-world environment in which ground
102 |
ADE | The Impact of Rolling Dynamic Compaction
improvement methods such as RDC are used. The author has witnessed RDC used at
remote and challenging sites containing highly variable fill, where it was difficult to
characterise and sub-sample representative soil conditions. In such instances where
mixed or highly variable fill is present at marginal or difficult sites, field trials are
recommended to quantify the limitations and capabilities of RDC.
Rigorous analysis of high speed photography to capture the changes in module velocity
would significantly add to current knowledge and is arguably a reasonable way to
quantify the frictional effects between module and soil that contribute to the work done.
A fellow PhD student at The University of Adelaide is undertaking a quantitative
assessment of high speed photography undertaken at one of the research intensive field
full-scale trials, and is combining it with finite element modelling to predict the energy
imparted into the ground; this work is progressing and is expected to support the
findings summarised in this thesis that have quantified the energy imparted into the
ground in terms of work done.
Rating an impact roller in terms of the energy that can be imparted to the ground surface
under ideal conditions is highly theoretical; of greater importance is the actual in-
ground response to RDC and how much work done is imparted into the soil at depth,
and the resultant load-displacement response. Ultimately, RDC is a method used to
achieve compaction, with the primary aim of reducing voids (inducing settlement)
within a soil as a result of imparting mechanical energy into the ground. In future, there
is a need to to quantify changes in stiffness with increasing RDC passes. The use of
buried earth pressure cells captures compressive waves induced from the module;
however, there is a need to also quantify the change in shear waves with increasing
passes. To be able to do this effectively, a carefully planned and executed trial would
include taking stiffness measurements not just pre- and post-compaction but after each
pass to be able to quantify changes in shear wave speed with changes in void ratio.
In Chapter 2 of this thesis, the maximum imparted energy delivered to the ground
surface by the 4-sided impact roller was found to lie in the range between 22 kJ and
30 kJ, for typical towing speeds of 9-12 km/h. In Chapter 4, the maximum work done
that was measured by a 230 mm diameter EPC, was 250 J. Clearly, there is a large
discrepancy between the external work applied to the soil surface, and, the internal work
measured by a single EPC at a depth of 0.7 m below the ground surface. It is evident,
103 |
ADE | The Impact of Rolling Dynamic Compaction
that there is a need for further research to try and quantify the dissipation of internal
work done on the soil over the influence zone of the impact roller, particularly given
that the magnitudes of internal work and external work are substantially different. To be
able to accurately quantify the dissipation of internal work done on the soil would
require measurement of force and displacement over the influence zone of a single
impact of the impact roller. Using current EPC technology, there are limitations with
how close embedded EPCs can be spaced (both vertically and horizontally) for results
to be considered a reliable representation of quantifying in situ soil stress. Whilst it may
be possible that future advances in technology will make it easier to quantify the
dissipation of work done that is imparted into the ground at full scale, initial endeavours
to quantify this may be better achieved via numerical models, with calibration against
full scale measurements such as those undertaken in this study.
There is a need to conduct rigorous vibration monitoring of RDC to determine
appropriate safe distances for which RDC can safely operate in varying soil conditions.
Studying the effects of vibrations induced from RDC is beyond the scope of this thesis,
but is nonetheless important, as vibration effects can potentially limit the use of RDC in
some instances. A series of charts that compare measurements of peak particle
velocities, the industry standard for quantifying ground vibrations, with the distance
from the impact roller will provide increased confidence regarding operational safe
distances from structures of various types to prevent damage. The development of a
ground vibration model (or similar) would enable ground vibrations to be predicted in a
variety of applications and soil types. Together, these will enable RDC to be specified,
for a particular application and soil conditions, with much more certainty than is
currently the case.
There are three main focus areas for future research into RDC: full-scale field testing,
numerical modelling and physical scale model testing. Research using numerical
modelling is being undertaken by a fellow PhD student at The University of Adelaide,
and has the obvious advantage of being able to conduct sensitivity analyses. In a
computer model it is much easier to vary parameters such as soil type, moisture content,
layer thickness, module shape or number of passes. However, computer modelling
alone without calibration is of little value, therefore it is hoped that the output from this
research (and future targeted full-scale field testing) will be of use to numerical
modelling studies involving RDC.
104 |
ADE | The Impact of Rolling Dynamic Compaction
Conducting research in a laboratory using a physical scale model is another future
direction for research into RDC. A 1:13 scale model 4-sided impact roller has been
constructed and is currently being tested and calibrated against full-scale results from
this thesis at The University of Adelaide. Geometric and kinematic similarity is needed
in order to obtain meaningful results from a scale model test; however, this method is
showing promise in its developmental phase.
Research into RDC is still immature when compared to other compaction techniques
such as conventional circular static and vibratory rolling, and, deep dynamic
compaction. Given the complex nature of RDC, much research is still needed; it is
hoped that this thesis will inspire further research into RDC so that our current
knowledge and understanding of limitations into RDC can catch up to the
aforementioned compaction techniques. The development of a compaction model that
quantifies the performance of RDC (in terms of improvement in density, strength,
stiffness or permeability) as a function of the characteristics of the compactor module
(size, shape and number of passes, towing speed) and the geotechnical properties of
underlying soil profile is the ultimate aim. This research has provided steps towards
achieving this overarching aim.
105 |
ADE | The Impact of Rolling Dynamic Compaction
Chapter 6: References
ASTM (2007). D 698: Standard test methods for laboratory compaction characteristics
of soil using standard effort. ASTM International, West Conshohocken, PA, USA.
ASTM (2008). D 5195: Standard test methods for density of soil and rock in-place at
depths below surface by nuclear methods. ASTM International, West Conshohocken,
PA, USA.
ASTM (2009a). D 6913: Standard test methods for particle-size distribution (gradation)
of soils using sieve analysis. ASTM International, West Conshohocken, PA, USA.
ASTM (2009b). D 1557: Standard test methods for laboratory compaction
characteristics of soil using modified effort. ASTM International, West
Conshohocken, PA, USA.
ASTM (2010a). D 2216: Standard test methods for laboratory determination of water
(moisture) content of soil and rock by mass. ASTM International, West
Conshohocken, PA, USA.
ASTM (2010b). D 4318: Standard test methods for liquid limit, plastic limit, and
plasticity index of soils. ASTM International, West Conshohocken, PA, USA.
Avalle DL (2004). Ground improvement using the “square” impact roller – case studies.
In Proceedings of the 5th International Conference on Ground Improvement
Techniques, Kuala Lumpur, Malaysia, pp. 101-108.
Avalle DL (2007) Trials and validation of deep compaction using the “square” impact
roller. Symposium Advances in Earthworks, Australian Geomechanics Society,
Sydney, Australia, pp. 63-69.
Avalle DL and Carter JP (2005). Evaluating the improvement from impact rolling on
sand. In Proceedings of the 6th International Conference on Ground Improvement
Techniques, Coimbra, Portugal, pp. 153-160.
107 |
ADE | The Impact of Rolling Dynamic Compaction
Avalle D and Grounds R (2004). Improving pavement subgrade with the "square"
impact roller. 23rd Annual Southern African Transport Conference, Pretoria, South
Africa, pp. 44-54.
Avalle DL and Mackenzie RW (2005). Ground improvement of landfill site using the
“square” impact roller. Australian Geomechanics, 40(4): 15-21.
Avalle DL, Scott BT and Jaksa MB (2009). Ground energy and impact of rolling
dynamic compaction – results from research test site. In Proceedings 17th
International Conference on Soil Mechanics and Geotechnical Engineering, Vol. 3,
Alexandria, Egypt, pp. 2228-2231.
Avalle D and Young G (2004). Trial programme and recent use of the impact roller in
Sydney. Earthworks Seminar, Australian Geomechanics Society, Adelaide, Australia,
5pp.
Avsar S, Bakker M, Bartholomeeusen G and Vanmechelen, J (2006). Six sigma quality
improvement of compaction at the New Doha International Airport Project. Terra et
aqua, 103: 14-22.
Berry AD (2001). Development of a volumetric strain influence ground improvement
prediction model with special reference to impact compaction. Dissertation submitted
in partial fulfilment of the requirements of the degree of Master of Engineering,
University of Pretoria, Pretoria, South Africa.
Bradley AC, Jaksa MB, Kuo YL. Examining the kinematics and energy of the four-
sided impact roller. Proceedings of the Institution of Civil Engineers – Ground
Improvement, 2019, 172 (4): 297-304.
Charles JA, Burford D and Watts KS (1981). Field studies of the effectiveness of
dynamic consolidation. In Proceedings of the 10th International Conference on Soil
Mechanics and Foundation Engineering, Stockholm, Sweden, 3: 617-622.
108 |
ADE | The Impact of Rolling Dynamic Compaction
Chen Z, Xu C, Ye G and Shen C (2014). Impact Roller Compaction of Dry Sand in
Laboratory Tests. In Proceedings Geo-Shanghai, Ground Improvement and
Geosynthetics, Shanghai, China, pp. 258-269.
Chow YK, Yong DM, Yong KY and Lee SL (1990). Monitoring of dynamic
compaction by deceleration measurements. Computers and Geotechnics, 10(3): 189-
209.
Clegg B (1980). An impact soil test as alternative to California bearing ratio. In
Proceedings 3rd Australia-New Zealand Conference on Geomechanics, New Zealand
Institution of Engineers, Wellington, New Zealand, pp. 225-230.
Clifford JM (1975). An introduction to impact rollers. National Institute for Road
Research, Internal report RP/2/75, South Africa, pp. 1-30.
Clifford JM (1978). Evaluation of Compaction Plant and Methods for the Construction
of Earthworks in Southern Africa. Dissertation in partial fulfilment of the
requirements for the degree of Master of Science in Engineering, University of Natal,
Durban, South Africa.
Clifford JM (1980). The development and use of impact rollers in the construction of
earthworks in southern Africa. CSIR Report 373, National Institute for Transport and
Road Research Bulletin 16, Pretoria, South Africa, pp. 1-53.
Clifford JM and Coetzee SD (1987). Coal and discard stockpiling with an impact roller.
In Proceedings of the International Conference on Mining and Industrial Waste
Management, Johannesburg, South Africa, pp. 75-79.
Clifford JM and Bowes G (1995). Calculating the Energy delivered by an Impact
Roller. A trilogy of Papers for the Sept. 1995 Lecture Tour and International
Seminars to commemorate the 10th Anniversary of the BH 1300 Standard Impact
Roller, Paper Two.
109 |
ADE | The Impact of Rolling Dynamic Compaction
Leonards GA, Holtz RD and Cutter WA (1980). Dynamic compaction of granular soils.
Journal of Geotechnical Engineering Division ASCE 106(1): 35-44.
Lukas RG (1980). Densification of loose deposits by pounding. Journal of Geotechnical
Engineering Division ASCE 106(4): 435-446.
Lukas RG (1995). Dynamic compaction. Geotechnical Engineering Circular No. 1, U.S.
DOT, Publication No. FHWA-SA-95-037, Federal Highway Administration,
Washington, D.C.
Mayne PW and Jones JS (1983). Impact stresses during dynamic compaction. Journal of
Geotechnical Engineering, ASCE, 109(10): 1342-1346.
Mayne PW, Jones JS and Dumas JC (1984). Ground response to dynamic compaction.
Journal of Geotechnical Engineering ASCE 110(6): 757-774.
McCann K (2015). The use of impact compaction for the near surface compaction on
dredged sand land reclamation projects. In Proceedings 12th Australia - New Zealand
Conference on Geomechanics, New Zealand Geotechnical Society, Wellington, New
Zealand, pp. 1193-1200.
McCann K and Schofield N (2007). Innovative methods in the in-situ determination of
design parameters on heterogeneous sites subject to ground treatment using deep
impact compaction. In Proceedings 10th Australia New Zealand Conference on
Geomechanics, Australian Geomechanics Society, Vol. 2, Brisbane, Australia, pp.
126-131.
Menard L and Broise Y (1975). Theoretical and practical aspects of dynamic
consolidation. Geotechnique, 25(1): 3-18.
Mooney MA and Rinehart RV (2007). Field monitoring of roller vibration during
compaction of subgrade soil. Journal of Geotechnical and Geoenvironmental
Engineering, ASCE, 133(3): 257-265.
111 |
ADE | The Impact of Rolling Dynamic Compaction
National Instruments (2019) http://www.ni.com/en-us/shop/labview.html (accessed
03/05/2019).
Nazhat Y (2013). Behaviour of sandy soil subjected to dynamic loading. Thesis
submitted in fulfilment of the requirement for the degree of Doctor of Philosophy,
University of Sydney.
Ranasinghe RATM, Jaksa MB, Kuo YL and Pooya Nejad F (2017). Application of
artificial neural networks for predicting the impact of rolling dynamic compaction
using cone penetrometer test results. Journal of Rock Mechanics and Geotechnical
Engineering, 9(2): 340-349.
Rinehart RV and Mooney MA (2009). Measurement of Roller Compactor Induced
Triaxial Soil Stresses and Strains. Geotechnical Testing Journal, 32(4): 347-357.
SA (Standards Australia) (1997) AS 1289.6.3.3: Methods of testing soils for
engineering purposes – Soil strength and consolidation tests – Determination of the
penetration resistance of a soil – Perth sand penetrometer test. Standards Australia,
Sydney, Australia.
Salgado R and Yoon S (2003) Dynamic cone penetration test (DCPT) for subgrade
assessment. Joint Transportation Research Program, West Lafayette, IN, USA,
Report FHWA/IN/JTRP-2002/30.
Scott BT and Jaksa MB (2008). Quantifying the influence of rolling dynamic compaction.
In Proceedings of the 8th Australia New Zealand Young Geotechnical Professionals
Conference, New Zealand Geotechnical Society, Wellington, New Zealand, pp.
199-204.
Scott BT and Jaksa MB (2012). Mining applications and case studies of Rolling Dynamic
Compaction. In Proceedings of the 11th Australia New Zealand Conference on
Geomechanics, Australian Geomechanics Society, Melbourne, Australia, pp. 961-966.
112 |
ADE | B.T.Scottetal./JournalofRockMechanicsandGeotechnicalEngineering12(2020)126e134 127
withanormalstressmeasurementrangeof6000kPatocapturethe
expectedhigherloadsfromtheimpactroller.
It has been well documented by researchers (e.g. Weiler and
Kulhawy, 1982; Rinehart and Mooney, 2009) that a buried cell
can influence localised stress fields and therefore any measure-
ments may not be representative of the true loading-induced
stresses. They discussed that errors can be minimised via the
choice of pressure cell design, by undertaking calibration and by
theuseofcorrectfieldplacementtechniques.Giventhechallenges
associatedwithmeasuringinsitustressaccurately,itwasimpor-
tanttocharacterisetheuncertaintyinthemeasurementtechniques
adopted.Awholesystemcalibrationwasperformedbothpre-and
post-testing,wherebytheworst-casescenariowasadifferenceof
Fig.1. 4-sidedRDCmodule(Broons). 8.5%.Thismagnitudeoferrorisgenerallyconsistentwiththatre-
ported by Dave and Dasaka (2011) who compared different cali-
bration techniques for EPCs and stated that pressure cell output
material containing a granular/cohesive mixture could reduce
couldbeconsideredreliablewithinanerrorofapproximately10%.
lateralshearingeffectsandaidedtractionofthemodulefortypical
The dynamic frequency response (peak capture) was affected by
towing speeds. Clifford (1978) described a case study where an thedataacquisitionrateandanyinternalfilteringusedinthesignal
insufficiently thick capping layer was adopted which resulted in
path. The data acquisition rate selected was 2000 samples per
individual impact blows punching through to the underlying second,andthefilterusedwassetat800Hz.FastFouriertransform
dredgedfill;thesitewasalsodividedintoaseriesofsmallworking
analysis of the data indicated that the fundamental frequency of
areas inwhich the roller was unable tomaintain a towing speed impulsesduetoRDCwaslessthan800Hz,confirmingthatthepeak
withinthetypicalrange.AccordingtoClifford(1978),bothfactors valueswerenotattenuatedbytheadoptedfilter.
causeareductioninspeedandarethekeyreasonsthatbetterre-
sultscouldnotbeobtained.
2.1. TrialA
Clifford (1980) discussed that there is an upper speed limit
beyondwhichanimpactblowisnotdeliveredbythefaceof the A field trial was undertaken at Monarto Quarries, located
module.Attowingspeedsgreater thanthetypicalrange,Clifford
approximately 60 km southeast of Adelaide, South Australia. The
(1980)statedthat the rollercan spin as a circular mass and only
testsitewasprimarilychosenbecausetherewasaccesstoearth-
contact the ground with its corners, a condition that should be
moving equipment, and importantly, homogeneous quarry mate-
avoided.Avsaretal.(2006)describedthecompactionofa22-km2
rialwasusedforthefieldtrial.Anareawithinthequarrywherethe
reclamation area for the new Doha International Airport Project. groundwasflat,closetomaterialstockpiles,yetawayfromquarry
They identified towing speed as one of the most important in-
operationswaschosenforthetrial.Naturalsoilwasremovedtoa
dicatorsthatdirectlyinfluencedtheinsitudrydensitythatcould
depthof1.75m,overaplanareathatwas10mlongand5.5mwide.
beachieved;anoptimumtowingspeedofthe4-sidedrollerforthat
ThreeGeokonModel3500EPCswereburiedatnominaldepthsof
project was found to be 11 km/h. Chen et al. (2014) conducted a 0.5m,1mand1.5mwithinthequarryfillmaterialthatwasplaced
laboratory investigation on a scale model impact roller device in
in seven lifts of 250 mm thickness. Bedding sand was placed
loosedrysand,byexaminingtheeffectofmoduleweight,sizeand
immediately below and above each pressure cell to ensure hori-
towingspeed.TheyusedaChineseconepenetrationtesttoconfirm
zontal placement and to prevent gravel sized particles of the fill
thattowingspeedwasoneofthemostimportantfactorscontrib-
materialfromdamagingthecells.Eachliftwaswheel-rolledusinga
utingtotheeffectivenessoftheimpactroller.Theaforementioned
Volvo L150E loader; a vibrating plate compactor was used to
casesgenerallysupporttheconceptthattowingspeedinfluenced
compact soil within 250 mm from each EPC to prevent possible
theeffectiveness,asdidthefindingsofScottandSuto(2007),who
damage.
statedthatgroundneartheperimeterofafencedsitecouldnotbe
improved as successfully as the rest of site due to access-related 2.1.1. Materialclassification
issues that reduced the towing speed of the module. This paper Thefillmaterialplacedforthetrialwasacrushedrockwitha
presents the findings of two full-scale field trials that were un-
maximum particle size of 20 mm that was readily available and
dertaken to quantify the effect of towing speed for the 4-sided
locally produced. A summary of the particle size distribution and
impactroller.
ProctorcompactiontestresultsforTrialAisgiveninTable1.ForTrial
A, particle size distribution (ASTM D6913-04(2009), 2009) results
2. Testingmethodology aretheaverageofninetests,andthestandard(ASTMD698-12,2012)
andmodified(ASTMD1557-12,2012)Proctorcompactionresultsare
Eachtimethemoduleofanimpactrollerstrikestheground,a the average of three curves. The field moisture content (ASTM
pressure wave is created that travels through the soil from the D2216-10,2010)reportedistheaverage ofnine testsundertaken.
surface. A keyaim of the trial is to measure the loading-induced Atterberglimittesting(ASTMD4318-10,2010)confirmedthatthe
stressesbelowthegroundduetoRDC.EPCsallowreal-timemea- finesconsistedofclayoflowplasticity.AccordingtotheUnifiedSoil
surements of stresses imparted to the ground. Rinehart and ClassificationSystem(USCS),thefillmaterialusedforthiscompac-
Mooney (2009) successfully used Geokon Model 3500 semi- tiontrialcouldbedescribedaswell-gradedgravel(GW).
conductor type EPCs in a field trial to measure dynamic loading TheaimofTrialAundertakeninAugust2012wastomeasure
inducedfromvibratorycirculardrumrollers.Theyused100mm- theloading-inducedstressatthreedifferentdepthsfor40passesin
diameter cells that were 10 mm thick with normal stress mea- total;10passesoftherollerwereconductedateachofthetowing
surementrangesof250kPa,400kPaand1000kPa.Thesametype speedsof9,10,11and12km/h.Towingspeedwascontrolledvia
of cells were selected to measure the pressure imparted into the the control panel in the towing unit (i.e. tractor) but was subse-
soilduetoRDC,albeit230mm-diametercellsof6mmthickness quently validated by dividing the distance between EPCs by the |
ADE | GroundImprovement Depthofinfluenceofrollingdynamic
compaction
Scott,JaksaandMitchell
sites containing significant soil variability, the use of pre- and important to note that the target specification, the testing
post-compaction testing can be problematic. To overcome this methods used to quantify improvement and the interpretation
limitation, this paper describes a compaction trial where earth of how the depth of improvement is both defined and quanti-
pressurecells(EPCs)wereplacedatdifferentlocationsbeneath fied vary between the listed references, making it difficult to
thegroundsurfaceinhomogeneoussoilconditionstoquantify draw definitive conclusions as to the maximum improvement
thedepthstowhichRDCimprovestheground. depthorlayerthicknesspossible.Incurrentpractice,itisoften
theresponsibilityoftheprojectengineertopredictwhetherthe
2. Background use of RDC will improve the ground sufficiently for the
Published case studies involving standard four-sided impact desired project application. The variable and unknown depth
rollers that have improved the ground in situ and have com- of influence of RDC is a key reason why this ground-improve-
pacted soil in thick layers are summarised in Tables 1 and 2, ment technique is not used more commonly, and highlights
respectively. In addition to the referenced published articles, whyfurtherresearchisneeded.
the authors reviewed dozens of unpublished reports on the
use of a four-sided 8 t roller in a variety of soil conditions. Kim (2010) performed finite-element simulations on impact
Their findings are in general agreement with the improvement rollers of different shapes with the aim of determining the
depths and layer thicknesses summarised in Tables 1 and 2, stress distribution and influence depth, which was defined as
respectively. It is clear from Tables 1 and 2 that the depth of thedepth atwhich thevertical stress decreased to one-tenth of
improvement of RDC varies significantly depending on the the applied stress at the surface. In that study, the module
soilmaterialtype.ItisreasonabletoconcludethatRDChasa mass, diameter and width of each roller were held consistent;
greater depth of influence in granular soils than in clays. It is only the shape and number of sides varied. This study ident-
also evident that the thickness of compacted layersis lessthan ified that the influence depth is afunction of both the contact
the depth of improvement in the same soil type, as the com- area and applied stress, with greater contact area and surface
pacted layer thickness is typically tailored to meet a target contact pressures resulting in increased depths of influence. A
specification. key limitation of this study, given the definition of influence
depth adopted, was that the surface contact stresses modelled
While not summarised in these tables, other variables such as for impact rolling were not verified using field test results.
moisture content, groundwater conditions and the number of Significantly, Kim’s analysis illustrated stresswave propagation
passes applied also affect the depth to which ground can be to depths much greater than those typically influenced by
improved using RDC. When reviewing Tables 1 and 2, it is static loading. Nazhat (2013) analysed the behaviour of sand
Table1. Improvementdepthsforcompactinginsitu
Reference Soiltype Improvementdepth:m
Clifford(1978) Sand >2·5
Clifford(1978) Sand >2·0
AvalleandYoung(2004) Fill(clay) 1·0
Avalle(2004) Fill(sand) >2·0
AvalleandGrounds(2004) Fill(mixed) 1·5
AvalleandMackenzie(2005) Fill(clay) 2·0
AvalleandCarter(2005) Fill(sand)overnaturalsand 3·0
Avalle(2007) Fill(sand) 2·5
ScottandSuto(2007) Fill(gravellyclay) 1·5
WhiteleyandCaffi(2014) Fill(mixed) 1·5
ScottandJaksa(2014) Fill(clayeysand)overnaturalclay 1·75
Table2. Thicknessofcompactedlayers
Reference Soiltype Layerthickness:m
WolmaransandClifford(1975) Sand 1·5
WolmaransandClifford(1975) Clay 0·6
Clifford(1980) Clay 0·5
CliffordandCoetzee(1987) Fill(coaldiscardmaterial) 0·5
AvalleandGrounds(2004) Fill(gravel) 1·0
Avalle(2007) Sandyclay/clayeysand 0·7
ScottandJaksa(2012) Fill(mixed) 1·0
ScottandJaksa(2014) Fill(clayeysand) 1·0
2
Downloaded by [] on [17/10/19]. Published with permission by the ICE under the CC-BY license |
ADE | GroundImprovement Depthofinfluenceofrollingdynamic
compaction
Scott,JaksaandMitchell
subjected to dynamic loading, and identified compaction The value of n was investigated in detail by Mayne et al.
shock bands bywayof the use of high-speed photographyand (1984),who collated datafromover 120sites andfoundthatn
imagecorrelationtechniquesfromlaboratory-basedtesting.As typically varied between 0·3 and 0·8, but could be as high as
explainedbyNazhat (2013),it isevident that improvements in 1·0insomeinstances.AsexplainedbyMayneetal.(1984)and
the ability to measure and quantify dynamic effects are Lukas(1995), thevariation inpredicted depthof improvement
helping to increase knowledge of unseen processes beneath the isnotsimplyafunctionofthetamperweightanddropheight,
ground surface; however, it is clear that more research is but is also influenced by other variables such as the tamper
needed to fully understand the kinematic behaviour of soils surface area, total energy applied, contact pressure of the
subjectedtodynamicloading. tamper, efficiencyof the dropping mechanism, initial soil con-
ditionsandgroundwaterlevels.
3. Dynamiccompaction Applying Equation 2 to the range of plotted values for
Dynamic compaction is a ground-improvement technique that n (0·3–0·8) in Mayne et al. (1984) to an 8t four-sided impact
usually employs a large crane to lift a heavy tamper, which is roller, using the maximum physical drop height of the module
then dropped onto the ground in a regular grid pattern. that is available on a flat surface (h=0·15m), the depth of
MenardandBroise (1975)improvedthemechanicalcharacter- improvement predicted would be in the range of 0·33–0·88 m.
istics of fine saturated sands using this method, and were the Hamidietal.(2009)appliedEquation2toRDCandindicated
firsttoproposearelationshipbetweenthethicknesstobecom- that the use of this equation was subject to controversy as
pacted, D, the pounder mass, m, and the drop height, h, largerdepthsofimprovementhavebeenreported.Table1con-
asgivenby firms the use of dynamic compaction formulae as under-
pffiffiffiffiffiffiffi estimating the improvement depths that are achievable using
1: D¼ mh RDC. While the application of deep dynamic compaction
theorytoRDCwithoutmodificationisnotsuitable,theuseof
a more appropriate nvalue doeswarrant further investigation,
Menard and Broise (1975) observed that greater depths of as both dynamic compaction theory and Table 1 indicate
improvement could be achieved for partially immersed soils that soil type is a key variable that influences the depth of
than for soils completelyout of water. The initial density and improvement.
grading were factors that influenced the time taken to reach a
liquefied state, after which the low-frequency, high-amplitude For dynamic compaction applications, Slocombe (2004)
vibrations from dynamic compactioncaused thesand particles
definesthe‘effectivedepthofinfluence’asbeingthemaximum
tobereorganisedintoamoredensestate.Insubsequentyears, depth at which significant improvement is measureable. The
this theory was applied to a wider range of soil conditions, ‘zone of major improvement’ is typically half to two-thirds
including unsaturated soils, and it was found that in many of the effective depth of influence. As explained by
cases the maximum depth of influence was less than that pre- Slocombe (2004), these terms have been adopted in the UK
dictedbyEquation1.Anumberofdifferentauthors,including but may have alternative meanings in different parts of the
Leonards et al. (1980), Lukas (1980, 1995) and Charles et al. world.
(1981), investigated the variation of an empirical factor (n)
with different soil conditions and for varying drop heights, h, Impact rolling is routinely undertaken in unsaturated soils,
and pounder masses, m. The general consensus isthatnvaries whereby the application of mechanical energy expels air
with different soil conditions, with lower values for fine- from the voids to reduce the void ratio. Within the influence
grained soils and larger values for coarse-grained soils, result- depth of RDC, repeated loading-induced stresses imparted
ing in varying estimations for the depth of improvement, as into a granular soil are sufficient to cause a permanent
perEquation2. rearrangement of soil particles, resulting in increased density
and soil settlement. Below the influence depth, the soil
pffiffiffiffiffiffiffi
2: D¼n mh remains elastic and does not undergo volume change. Berry
(2001) developed an elastoplastic model to determine the
depth to which there was permanent deformation using
Alternatively, Equation 2 can be re-written as shown in
surface settlement as the main input parameter. While Berry’s
Equation3.Inthisform,theright-handsideoftheequationis
model did not quantify the energy to achieve a particular
a function of gravitational potential energy, mgh, and the
surface settlement, it was observed that a depth of three
materialcharacteristics,describedbytheparametern.
times the module width was considered appropriate for a
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi three-sided impact roller. At sites with a shallow water table,
n2 it is possible for the high-amplitude and low-frequency
3: D¼ ðmghÞ
g vibrations associated with RDC to induce pore pressures to
rise to the surface. In order to prevent liquefaction from
3
Downloaded by [] on [17/10/19]. Published with permission by the ICE under the CC-BY license |
ADE | GroundImprovement Depthofinfluenceofrollingdynamic
compaction
Scott,JaksaandMitchell
occurring, the number of passes is typically limited to allow
2900 2900
pore-water pressures to dissipate. Rather than competing with,
impact rollers are often used to complement deeper ground- Lane A 2500
improvement techniques that leave soils within the top 2m of EPC1 and EPC4 EPC2 EPC3
thesurfaceinadisturbedandweakenedcondition.Avsaretal.
(2006)describe anexampleof alargeland reclamation project
whereby impact rolling successfully complemented deeper Lane B 2500
ground-improvementtechniques.
In the work described in this paper, the depth to which RDC
Lane C 2500
improves the ground measured in full-scale field trials
in homogeneous soil conditions. The measured data were
compared with predictions based on dynamic compaction (a)
theory to determine the relevance of this approach to RDC
846
applications. Layer 3 1460
EPC4
980
Layer 2 1530
4. Field trial to determinedepth of EPC3
improvement
A field trial was conducted using a Broons BH-1300 8 t EPC1 670 EPC2 870 Layer 1 1200
four-sided impact roller (Figure 1) at the Iron Duke mine (b)
located on the Eyre Peninsula in South Australia during June Figure2. (a)Planand(b)elevationviewsoftestpadincluding
2011. The test pad was constructed in three separate lifts, as EPClocations(alldimensionsinmm)
illustrated in Figure 2, which also shows the locations of
embedded EPCs in plan and elevation. The test padwas con-
structed using haul trucks, end tipping loose tailings material
instockpileswherealoaderandexcavatorsubsequentlyspread 4.1 Materialclassification
the material over the test pad. The placement process caused The test pad was constructed using iron magnetite tailings,
the soil to be partially compacted by the self-weight of the which are a by-product of a consistent rock-crushing process.
plant; however, this method was deemed representative of the In order to classify and determine the compaction character-
proposed construction method for the mine site and therefore istics of the tailings, particle-size distribution tests were per-
was consistent with the generic aim of the field compaction formed,aswellasstandardandmodifiedcompactiontests,the
trial to be as representative as possible given the site con- results of which are summarised in Table 3. The particle-size
straints. Aswell as undertaking the trial for research purposes, distribution (ASTM, 2009a) results are the average of nine
to determine thedepthof influence,therewasaneedto ascer- tests and the standard (ASTM, 2007) and modified (ASTM,
tainthelayerthicknessthatcouldbeplacedtoachieveatarget 2009b) Proctor compaction results are the average of three
density of 95% of maximum modified dry density for future curves. The large dry unit weights are a consequence of the
projectsatthemine. sand-sized particles consisting of crushed magnetite. The field
moisture content (FMC) (ASTM, 2010a) reported is the
averageof15testsundertaken.Atterberglimittesting(ASTM,
2010b) confirmed that the fines consisted of clay of low plas-
ticity (plastic limit 11% and liquid limit 22%). According to
the Unified Soil Classification System, the fill material used
for this compaction trial could be described as a well-graded
sand(SW).
4.2 EPCs
Four Geokon model 3500 (230mm diameter, 6 mm thick)
EPCswereusedtomeasurethedynamicpressuresimpartedby
RDC. As shown in Figure 2, the initial lift (1200mm thick
containing buried EPC1 and EPC2) was first compacted; this
was repeated for the second lift of 1530mm (containing
EPC3) and the third and final lift (1460mm containing
EPC4).Inplan,theEPCswereplacedone-halfofonerotation
Figure1. 8tfour-sidedimpactroller
of the roller apart (2·9 m) from each other in the forward
4
Downloaded by [] on [17/10/19]. Published with permission by the ICE under the CC-BY license |
ADE | GroundImprovement Depthofinfluenceofrollingdynamic
compaction
Scott,JaksaandMitchell
Table3. Particle-sizedistribution,compactionandfieldmoisturetestresults
Standard Standard Modified
optimum maximum maximum
moisture dryunit dryunit
content weight: Modified weight:
Material d :mm Gravel:% Sand:% Fines:% (OMC):% kN/m3 FMC:% OMC:% kN/m3
50
Magnetitetailings 0·7 14 80 6 6·6 23·9 5·1 5·7 25·8
d ,particlesizeatpercentfinerof50%
50
direction of travel. The EPCs were connected to a bespoke 4.3 Insitutesting
data-acquisition system and the Labview software program Various in situ testing methods were performed after 0, 8 and
(National Instruments, 2019). A sampling frequencyof 2kHz 16 passes to quantify soil improvement with increasing com-
(i.e.onesampleevery0·0005 s)wasadoptedtocapturesudden pactiveeffort.Theinsitutestswereundertakeninthecentreof
increases in pressure caused by the module impacting the lane A in layer 3, as shown in Figure 2. The tests conducted
ground. Prior to compaction, the EPCs were used to measure included field density measurements (ASTM, 2008), the spec-
theself-weightoftheimpactrollingmodulefortherollerinan tral analysis of surface waves (SASW) geophysical technique
‘at rest’ condition, centred above each EPC. The measured and dynamic cone-penetration (DCP) tests to measure and
pressures were compared to predictions using Fadum’s chart inferchangesindensityasafunctionofthenumberofmodule
(Fadum, 1948) using elastic theory, the results of which are passes.SASWtestingwasconductedusingaGDSInstruments
shown in Figure 3. The measured pressures followed the same surfacewave systemusing six 4·5 Hz geophones spaced at 1 m
general trend, but were less than the predicted pressures; the intervalswithasledgehammersourceimpactingametalstrike
difference between the predicted and measured values was an plate 1m from the first geophone. DCP testing was under-
average of 38% over the depths measured. The most likely taken in accordance with the procedure described in AS
explanation for this is that the non-uniform shape of the 1289.6.3.3 (SA, 1997). Verification of RDC was also under-
modulefaceimpactingthegrounddoesnotproduceauniform taken using settlement monitoring to quantify the change in
pressure distribution and this is exacerbated for shallow EPC ground surface level with the number of passes. This was
depths. A towing speed of 10·5 km/h was selected for all 16 achieved using a level and staff to measure settlement at nine
passesthatwereconductedoneachlayer.Thestagedconstruc- points across the test pad in adjacent low points in the undu-
tion process resulted in the dynamic pressure imparted by lating surface, as is the normal practice. Due to space con-
RDCtobemeasuredatninedifferentdepths. straints, a discussion of testing methods generallyemployed to
verify RDC is not presented here. Theyare however, discussed
in detail by Avalle and Grounds (2004) and Scott and Jaksa
Pressure: kPa
(2008).
0 5 10 15 20 25 30 35
0 0
5. Results of thefield trial
0·5 This section provides details of the results obtained from the
0·5 field trial; specifically those obtained from the EPCs, in situ
andgeophysicaltestingandsettlementmonitoring.
1·0
1·0
1·5 5.1 EPCdata
Figure4illustratestheresultsobtainedforatypicalpassofthe
2·0 1·5 impact roller traversing over the first lift of the test pad, where
EPC1 and EPC2 were buried at depths of 0·67 and 0·87 m,
respectively. As expected, the shallower EPC recorded the
2·5
2·0
greatest pressure. Figure 5 presents the variation of measured
peakpressurewith depth, whereit isobservedthatpeakpress-
3·0
ures greater than 100kPa were recorded at depths above
2·5
2·0m. The EPC results generally supported other test data
3·5 Measured Predicted
that indicated that most of the quantifiable ground improve-
3·0 ment occurred within 2m of the surface. Even the deepest
4·0
EPC (buried at a depth of 3·85m below the ground surface)
Figure3. Measuredandpredictedpressuresagainstdepthfor
registered positive pressure readings due to the impact roller,
impactrolleratrest
suggesting that the depth to which RDC had an influence
5
Downloaded by [] on [17/10/19]. Published with permission by the ICE under the CC-BY license
m
:htpeD
htdiw
eludom/htpeD |
ADE | GroundImprovement Depthofinfluenceofrollingdynamic
compaction
Scott,JaksaandMitchell
1000 lateral offset distances of 2·5 and 5·0m. Fora lateral offset of
900 EPC1 2·5m, a maximum peak pressure was measured at a depth of
800 Direction of travel 2·0m. For a lateral offset of 5·0 m, all measured peak press-
700 ures were considered negligible. Further information on the
600 EPC2 lateral influence of RDC is discussed by Scott and Jaksa
500 (2014).
400
300 5.2 Insitutestresults
200 Figure 7 compares the average modified dry density ratio in
100 accordance with ASTM (2009b) against depth after eight
0 passes. From the trend line fitted to the data, it is estimated
0 1 2 3 4 5 6 7 8 9 10
that eight passes will achieve a dry density ratio of 95%, pro-
Time: s
vided that the layer thickness does not exceed 1·2 m. Due to
Figure4. Exampleresultsofpressureagainsttimeforasingle
time constraints on site, density testing was not undertaken
passoftheimpactroller:lift1containingEPC1andEPC2
after16passes.
The SASW technique was used in conjunction with DCP
Peak pressure: kPa tests to assess the improvement with depth at intervals of
eight passes. Results for layer 2 are shown in Figure 8, where
0 200 400 600 800 1000
0 0 it can be observed that an increased numberof passes resulted
in an increase in shear modulus between depths of 0·5 and
2·1m; this is an indication of increased soil density. Below
0·5
0·5 a depth of 2·1m the results were inconclusive due to insuffi-
cientdata.
1·0
1·0 Figure 9 summarises the number of DCP blows per 50 mm
1·5 penetration with respect to depth below the ground surface.
The tests were terminated at penetration depths of 850mm
2·0 1·5 due to the limited length of the penetrometer. Salgado and
Yoon (2003) found that increasing blow counts are indirectly
related to an increase in soil dry density. An increase in blow
2·5
2·0 count is evident with a greater number of passes to depths of
between 0·3 m and beyond the 0·85 m limit of the penetrom-
3·0 eter. Loosening of near-surface soils (<0·3m) as a conse-
2·5 quence of RDC is consistent with the findings of Clifford
(1975) and Ellis (1979), who both suggested that RDC is
3·5 y = 35·7x–0·6
R2 = 0·95 unsuitableasafinishingroller.
3·0
4·0
5.3 Surfacesettlementmonitoring
Figure5. Measuredpeakpressureagainstdepthwithtrendline
The average surface settlement across the test pad against
fittedtodata
number of passes was also measured. It was found that the
majority of settlement occurred within the first eight passes;
theaverage surface settlement measuredwas 106 and128mm,
extended beyond this depth. While the fitted trend line illus- aftereightand16passes,respectively.
trates a good fit to the measured data, extrapolating for
shallower than the measured depths is not recommended. 6. Discussion
A limitation of using EPCs is that they should not be placed In current practice, the influence depth of RDC can be inter-
atorclosetothegroundsurfaceduetothehighprobabilityof preted differentlyasthere are many in situ techniquesthat can
damaging the sensors, with the manufacturer’s guidelines be, and are, used to measure it. In essence, these estimates are
recommending that no heavy equipment be used over the only as good as the quality of the pre- and post-compaction
cells unless at least 500mm of material is placed above testing undertaken. It is suggested that three basic definitions
them (Geokon, 2007). Figure 6 illustrates the measured peak are relevant in this context. First, the depth of influence, in
pressures, plotted on a log scale, that were recorded by each simple terms, is the depth to which some improvement in
EPC as the impact roller traversed directly above (lane A) density or reduction in void ratio is evident, regardless of
and in the lanes adjacent to the buried EPCs, representing magnitude. To determine this, predictive models such as that
6
Downloaded by [] on [17/10/19]. Published with permission by the ICE under the CC-BY license
aPk
:erusserP
m
:htpeD
htdiw
eludom/htpeD |
ADE | GroundImprovement Depthofinfluenceofrollingdynamic
compaction
Scott,JaksaandMitchell
Peak pressure: kPa Peak pressure: kPa Peak pressure: kPa
1 10 100 1000 1 10 100 1000 1 10 100 1000
0 0 0 0 0 0
0·5 0·5 0·5
0·5 0·5 0·5
1·0 1·0 1·0
1·0 1·0 1·0
1·5 1·5 1·5
2·0 1·5 2·0 1·5 2·0 1·5
2·5 2·5 2·5
2·0 2·0 2·0
3·0 3·0 3·0
2·5 2·5 2·5
3·5 3·5 3·5
3·0 3·0 3·0
4·0 4·0 4·0
(a) (b) (c)
Figure6. MeasuredpeakpressureagainstdepthforvaryinglateraldistancesfromthecentreoflaneA:(a)0m;(b)2·5m;(c)5·0m
MDD ratio: % Shear modulus: MPa
90 95 100 0 50 100 150 200 250
0 0
0·5
0·2
1·0
0·4
1·5
0·6 2·0
0·8 2·5
Trend line
3·0
1·0
3·5
1·2
4·0
Zero passes Eight passes 16 passes
1·4
Figure8. Geophysical(SASW)testresultsforzero,eightand16
Figure7. Modifiedmaximumdrydensityratioagainstdepthafter passes
eightpasses
ground. Other factors include the potential energy due to the
proposed by Berry (2001) could be adopted; applying this double-spring−linkage system and the kinetic energy due to
theory to the four-sided roller yielded an influence depth of friction between the soil and module interface. The effects of
3·9 m. Alternatively, sensitive measuring equipment, such as the double-spring−linkage system can be quantified by wayof
EPCs, or intrusive site-investigation techniques, such as the achange in module velocity, and hence considered part of the
cone-penetrationtestanddilatometertest,couldbeused. kinetic energy component delivered by the impact roller. For
the towing speed adopted in the field trial reported in this
Here, no attempt is made to quantify the depth to which paper, the changes in potential and kinetic energies are listed
RDC has a small positive influence. Instead, an energy-based inTable4.
approachisproposedtoprovideestimationsofthedepthscap-
ableofbeingsignificantlyimprovedinsituandthelayerthick- The second definition is applicablewhen improving ground in
nesses capable of being compacted by RDC. Gravitational situ; in such cases, depths shallower than the maximum
potentialenergyformspartofthetotalenergyimpartedtothe capable by RDC are typically targeted for improvement.
7
Downloaded by [] on [17/10/19]. Published with permission by the ICE under the CC-BY license
m
:ecafrus
dnuorg
woleb
htpeD
m
:htpeD
htdiw
eludom/htpeD m
:htpeD
m
:ecafrus
dnuorg
woleb
htpeD
htdiw
eludom/htpeD m
:htpeD
htdiw
eludom/htpeD |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.