Elimination of the Rayleigh Wave Effect on Low Strain

Integrity Test Results

(Part 2: Rayleigh Wave Elimination Technique)

J.P. Seidel1
S. K. Tan2

ABSTRACT
Following the experimental investigation, a technique was subsequently proposed to eliminate the effect
of Rayleigh Wave on low strain integrity test results. The technique involves simultaneous measurements
of the horizontal and vertical velocities during low strain integrity testing. By measuring the horizontal
velocity of Rayleigh waves, the vertical velocity of Rayleigh waves could be deduced by applying an
appropriate factor and subsequently be eliminated from the vertical velocity responses. It is shown that
this technique can eliminate the effect of Rayleigh Wave. However, a limitation of the technique is that
the ‘net’ velocity response applying the technique does not seem to indicate the reliable profile
consistently.
Keywords: surface waves, Rayleigh waves, sonic echo method, integrity testing

1. PROPOSED TECHNIQUE
Some understanding on the behaviour of Rayleigh waves reflecting back and forth along the pile top was
achieved based on the experimental work done. It was found that there was a correlation between the
horizontal velocity and vertical velocity of Rayleigh waves. It is postulated that the velocity responses
obtained from sonic echo tests contain the vertical velocity component of both compression waves and
Rayleigh waves. It is essential to obtain only the vertical velocity component of the compression waves
for a reliable interpretation of the test results. If the vertical velocity component of Rayleigh waves can be
eliminated from the velocity responses, then a pure response due to only compression waves can be
obtained.
A technique was subsequently proposed to enable the elimination of the vertical velocity of Rayleigh
waves from the vertical velocity responses. The key aspect of the technique is the simultaneous
measurements of the horizontal and vertical velocities during low strain integrity testing. By measuring
the horizontal velocity of Rayleigh waves, the vertical velocity of Rayleigh waves could be deduced by
applying an appropriate factor and subsequently be eliminated from the vertical velocity responses.

2. VALIDITY OF THE PROPOSED TECHNIQUE

In order to investigate the validity of the proposed technique, the technique was applied to velocity
responses of selected piles with intentionally built-in defects. The summary of the pile profiles is
presented in Table 1. A medium sized hammer of weight 450 g was used as a hammer impact. Sonic echo

1
Managing Director, FoundationQA Pty Ltd, Australia
2
Geotechnical Engineer, Gue and Partners Sdn Bhd, Malaysia
tests were performed on the two piles with three different locations of hammer impact and accelerometer.
The locations are:
• A@edge F@edge
• A@edge F@centre
• A@edge F@other edge
Figure 1 illustrates a schematic diagram of locations of the hammer impact and the accelerometer.

Table 1: Profiles of piles

Defect
Test pile Pile Length (m)
Thickness (mm) Depth from top Area (%)
(m)
4 12.10 300 10.10 26.0
5 11.89 300 9.89 18.0

edge other
edge
centre

Figure 1: Schematic diagram of locations of the hammer impact and the accelerometer
A magnification factor of 2 and a wavelet length of 1.0 m were applied to all velocity responses for
consistency and to enhance the presentation of the velocity responses. No low pass or high pass filter was
used as it was found to be unnecessary.

According to Kolsky(1963), the path of any particle is an ellipse and the its major axis is normal to the
surface. For particles at the surface (z = 0), the ratio between the major and minor axes of the ellipse is 1.468.
This suggests that the ratio of the maximum amplitude of vertical velocity to horizontal velocity component
of Rayleigh waves should be 1.468. Therefore, in the analysis, a constant ‘correction’ factor of 1.468 was
applied to the horizontal velocity to deduce the vertical velocity of the Rayleigh waves. This deduced vertical
velocity response of Rayleigh waves was then subtracted from the vertical velocity response of the
individual pile to eliminate the effect of Rayleigh waves.

Figure 3 shows the velocity responses of piles 4 and 5 before the correction factor was applied while Figure
4 shows the velocity responses of piles 4 and 5 after the correction factor was applied. As can be seen from
the graphs, the general shape of the ‘corrected’ velocity responses of pile 4 and 5 is similar and consistent.
However, from the analysis, it was found that the correction factor to eliminate the effect of Rayleigh waves
varied for different locations (see Table 2). The difference in the correction factor applied may indicate that
there is a superposition effect of the multiple paths travelled by Rayleigh waves upon reaching the
accelerometer as shown in Figure 5.
Demo V ersion - No Commercial Use 1 7/0 5/02
DISTA NCE
C:\pit\Monas h_pile\H_V \DISTA NCE\Dis ta nc e.PIT PIT- W ™ V er. 20 01- 1.DEMO
cm/s 3: # 21 PILE 4
0.20
A@edge F@edge
m edium
0.10 16/05/02 10:06:10 PM
2W 2.00 m 1011 Hz
A1: 0m
0.00 A2: 0m

x 2 L/D=20 (D=60 cm )
-0.10 V 0.178 cm /s (0.217)
12.10 m (4045 m /s)
V2 0.080 cm /s (0.094)
0 5 10 15 m
cm/s 3: # 23 PILE 4
0.20
A@edge F@centre
m edium
0.10 16/05/02 10:08:08 PM
2W 2.00 m 1054 Hz
A1: 0m
0.00 A2: 0m

x 2 L/D=20 (D=60 cm )
-0.10 V 0.134 cm /s (0.150)
12.10 m (4215 m /s)
V2 0.041 cm /s (0.066)
0 5 10 15 m
cm/s 3: # 25 PILE 4
0.20
A@edge F@other edge
m edium
0.07 16/05/02 10:09:01 PM
2W 2.00 m 1000 Hz
A1: 0m
-0.07 A2: 0m

x 2 L/D=20 (D=60 cm )
-0.20 V 0.044 cm /s (0.089)
12.10 m (4000 m /s)
V2 0.026 cm /s (0.090)
0 5 10 15 m
cm/s 3: # 27 PILE 5
0.20
A@edge F@edge
m edium
0.10 16/05/02 10:10:53 PM
2W 2.00 m 1007 Hz
A1: 0m
0.00 A2: 0m

x 2 L/D=20 (D=60 cm )
-0.10 V 0.188 cm /s (0.223)
11.89 m (4029 m /s)
V2 0.099 cm /s (0.115)
0 5 10 15 m
cm/s 3: # 29 PILE 5
0.20
A@edge F@centre
m edium
0.10 16/05/02 10:11:53 PM
2W 2.00 m 1045 Hz
A1: 0m
0.00 A2: 0m

x 2 L/D=20 (D=60 cm )
-0.10 V 0.123 cm /s (0.138)
11.89 m (4181 m/s)
V2 0.063 cm /s (0.071)
0 5 10 15 m
cm/s 3: # 31 PILE 5
0.20
A@edge F@other edge
m edium
0.10 16/05/02 10:12:39 PM
2W 2.00 m 1000 Hz
A1: 0m
0.00 A2: 0m

x 2 L/D=20 (D=60 cm )
-0.10 V 0.038 cm /s (0.057)
11.89 m (4000 m /s)
V2 0.019 cm /s (0.034)
0 5 10 15 m

Horizontal-Vertical velocity responses for piles 4 and 5

 cm/s 3: # 21 PILE 4
0.40
A@edge F@edge
m edium
0.20 16/05/02 10:06:10 PM
2W 2.00 m 1011 Hz

0.00

x 2 L/D=20 (D=60 cm )
-0.20 V+ 0.279 cm /s (0.217)
12.10 m (4045 m /s)
0 5 10 15 m
cm/s 3: # 23 PILE 4
0.40
A@edge F@centre
m edium
0.20 16/05/02 10:08:08 PM
2W 2.00 m 1054 Hz

0.00

x 2 L/D=20 (D=60 cm )
-0.20 V+ 0.186 cm /s (0.150)
12.10 m (4215 m /s)
0 5 10 15 m
cm/s 3: # 25 PILE 4
0.40
A@edge F@other edge
m edium
0.20 16/05/02 10:09:01 PM
2W 2.00 m 1068 Hz

0.00

x 2 L/D=20 (D=60 cm )
-0.20 V+ 0.201 cm /s (0.089)
12.10 m (4272 m/s)
0 5 10 15 m
cm/s 3: # 27 PILE 5
0.40
A@edge F@edge
m edium
0.20 16/05/02 10:10:53 PM
2W 2.00 m 1015 Hz

0.00

x 2 L/D=20 (D=60 cm )
-0.20 V+ 0.255 cm /s (0.222)
11.89 m (4061 m /s)
0 5 10 15 m
cm/s 3: # 29 PILE 5
0.40
A@edge F@centre
m edium
0.20 16/05/02 10:11:53 PM
2W 2.00 m 1045 Hz

0.00

x 2 L/D=20 (D=60 cm )
-0.20 V+ 0.179 cm /s (0.138)
11.89 m (4181 m /s)
0 5 10 15 m
cm/s 3: # 31 PILE 5
0.40
A@edge F@other edge
m edium
0.20 16/05/02 10:12:39 PM
2W 2.00 m 1050 Hz

0.00

x 2 L/D=20 (D=60 cm )
-0.20 V+ 0.140 cm /s (0.057)
11.89 m (4199 m /s)
0 Figure 3: Horizontal-Vertical velocity responses for piles 4 and 5
5 10 15 m

Figure 4: ‘Corrected’ velocity responses for piles 4 and 5

Although the technique is successful in obtaining a unique response which is location-independent, the
‘corrected’ velocity response does not seem to indicate the true profile of the pile reliably. As shown in Table
1, there is a defect located at the depth of 10.10 m in pile 4. However, there is no indication of this defect in
the corrected velocity responses of pile 4 for all the three locations. Similar phenomenon is also observed in
the corrected velocity responses of pile 5 for the location A@edge F@other edge. In pile 5, there is a defect
located at a depth of 9.89 m. There is a barely discernible peak identified at a depth of approximately 10 m in
the corrected velocity responses for locations A@edge F@edge and A@edge F@centre. However, the peak
is not discernible in the corrected velocity response for the location A@edge F@other edge. These results
suggest that the orientation of defects within the pile could possibly affect the velocity responses.

Table 2: Correction factor used on velocity responses of piles 4 and 5

Location Pile
4 5
A@edge F@edge 0.80 0.80
A@edge F@centre 1.10 1.00
A@edge F@other edge -1.20 -1.47

edge other
edge

Figure 5: Possible travel paths of Rayleigh waves

3. CONCLUSIONS
Through the implementation of the technique, it was demonstrated that the horizontal velocity could be
used to deduce the vertical velocity of Rayleigh waves despite the difference in the amplitude of the
‘corrected’ horizontal and the vertical velocity in the reflected Rayleigh waves.
A technique was subsequently proposed to enable the elimination of Rayleigh waves effect in a velocity
response and hence to obtain a location-independent response. The key aspect of the proposed technique is
the simultaneous measurements of the horizontal velocity and the vertical velocity and eliminating
Rayleigh waves effect by applying an appropriate correction factor to the horizontal velocity. The
technique was applied to the velocity responses of selected piles to assess the validity of the technique.
Through the implementation of the technique, it was demonstrated that the technique is capable of
eliminating Rayleigh waves effect in a velocity response, which lead to a location-independent response.
The successful application of the technique shows that Rayleigh waves are responsible for the location
dependence effect. Although the technique is successful in obtaining a location-independent response, it
has a limitation; the ‘corrected’ velocity response does not seem to indicate the true profile of the pile
consistently. Future research needs to be carried out to investigate the causes for this occurrence and to
improve the correlation of the ‘corrected’ velocity response with the pile profile.

References
Kolsky, H. (1963). Propagation in an extended elastic medium. Stress Waves in Solids. New York, New
York Dover, Inc.: pp. 16-24