Conference PaperPDF Available

The influence of soil conditions on railway induced ground-borne vibration and relevant mitigation measures

Conference Paper

The influence of soil conditions on railway induced ground-borne vibration and relevant mitigation measures

Abstract

Railway induced ground-borne vibration at low frequency, below about 50 Hz, has a great impact on the nearby environment. In many cases, it is recognised as a major source of annoyance for line-side residents. Within the EU research project RIVAS (Railway Induced Vibration Abatement Solutions), a number of mitigation methods were proposed and investigated to reduce the vibration level under typical European soil conditions. These included various measures applied in the transmission path such as open trenches, soft or stiff buried barriers and subgrade stiffening techniques. Although only selected cases could be studied, it has been shown that the soil conditions have an important effect both on the ground-borne vibration induced by passing trains and on the predicted performance of these mitigation measures. Based on various case studies, this paper aims to give a clearer view of the relationship between soil conditions (including homogeneous and layered ground with various soil properties) and the performance of these mitigation methods. Copyright © (2014) by the International Institute of Acoustics & Vibration All rights reserved.
ICSV21, Beijing, China, 13-17 July 2014 1
The 21
st
International Congress on Sound and Vibration
13-17 July, 2014, Beijing/China
THE INFLUENCE OF SOIL CONDITIONS ON RAILWAY
INDUCED GROUND-BORNE VIBRATION AND RELE-
VANT MITIGATION MEASURES
Jian Jiang
1
, Martin G.R. Toward
1
, Arne Dijckmans
2
, David J. Thompson
1
,
Geert Degrande
2
, Geert Lombaert
2
and Mohammed F.M. Hussein
1
1
Inst. of Sound and Vibration Research, University of Southampton, SO17 1BJ, UK
2
KU Leuven, Dept. of Civil Engineering, Kasteelpark Arenberg 40, 3001 Leuven, Belgium
e-mail: J.Jiang@soton.ac.uk
Railway induced ground-borne vibration at low frequency, below about 50 Hz, has a great
impact on the nearby environment. In many cases, it is recognised as a major source of an-
noyance for line-side residents. Within the EU research project RIVAS (Railway Induced
Vibration Abatement Solutions), a number of mitigation methods were proposed and investi-
gated to reduce the vibration level under typical European soil conditions. These included
various measures applied in the transmission path such as open trenches, soft or stiff buried
barriers and subgrade stiffening techniques. Although only selected cases could be studied, it
has been shown that the soil conditions have an important effect both on the ground-borne
vibration induced by passing trains and on the predicted performance of these mitigation
measures. Based on various case studies, this paper aims to give a clearer view of the rela-
tionship between soil conditions (including homogeneous and layered ground with various
soil properties) and the performance of these mitigation methods.
1. Introduction
Ground-borne vibration from railways is recognised as a major source of annoyance for line-
side residents. It can also cause malfunction of sensitive equipment. Most of the vibration is in-
duced at the wheel/rail interface due to various mechanisms, including track unevenness or rough-
ness, the moving quasi-static load, and transient effects from rail joints, switches and crossings
1
.
Feelable vibration can be perceived over a frequency range from 1 Hz to 80 Hz. However, higher
frequencies are generally attenuated rapidly with distance through the ground.
Although the ground-borne vibration can be attenuated by introducing isolation measures at
the source, i.e. the track, or at the receiver, i.e. a building, for low frequency vibration (< 50 Hz),
interrupting the vibration transmission path is more effective. Measures that have been considered
for use in the transmission path include open trenches
2
, buried stiff
3, 4
or soft
5
wall barriers, sub-
grade stiffening
6
and wave impeding blocks
7
.
The open trench is a classic measure which is designed to reflect the incident surface propa-
gating wave. Previous studies have shown that, for a homogeneous half-space, the effectiveness of
an open trench is mostly related to the Rayleigh wavelength. The generally accepted criterion is that
the depth of the trench should be at least 0.6 times the Rayleigh wavelength
4
. Buried stiff or soft
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 2
barriers are used as practical alternatives due to the difficulty of constructing an open trench. Alt-
hough they still act as wave reflector, they are normally less effective and the performance is influ-
enced by the impedance and stiffness contrast between the in-fill material and the soil.
For railway tracks on soft soils, subgrade stiffening is often applied to improve the soil be-
neath the track with the aim of reducing settlements or track displacements. This technique is also
known to lead to reduced levels of ground-borne vibration. Similar to subgrade stiffening, wave
impeding blocks have been proposed for application under or next to a railway track. However, they
are placed at some depth beneath the track. This results in a change of cut-on frequency of the upper
soil layer (the frequency above which waves that are limited to the upper layer can propagate).
Therefore, theoretically, an infinitely long wave impeding block can reduce the low frequency vi-
bration by increasing the existing cut-on frequency.
Although many research projects have been carried out to investigate the performance of
these mitigation measures, few of them have considered them in association with a range of soil
conditions. Soil condition (material properties and soil layering) has a great impact on the way
waves propagate. As a result, this can make a big difference to both the amplitude and frequency
range of perceived vibration 7-9. It is certainly important for those mitigation measures relying on
differences in properties between the soil and the materials used (e.g. buried barriers). Even for an
open trench, the depth of the upper soil layer, which introduces a cut-on frequency, would also in-
fluence the performance.
Within the EU FP7 project RIVAS (Railway Induced Vibration Abatement Solutions), a
number of mitigation methods have been proposed and carefully investigated with the aim of reduc-
ing the vibration level under typical European soil conditions. In this paper, results of four mitiga-
tion measures are presented: an open trench, a buried soft wall barrier, subgrade stiffening and wave
impeding blocks. These are applied under several soil conditions consisting of both layered and
homogeneous ground with different soil stiffnesses. Parametric studies have been carried out by
using two-and-a-half-dimensional (2.5D) finite element - boundary element (FE-BE) models. De-
tails of the models will be introduced in Section 2. Section 3 will give the results of these four miti-
gation measures under typical homogeneous half-space and layered ground. Performance of these
mitigation measures at three selected real reference sites (across Europe) will also be compared in
Section 3. Conclusions will be given in Section 4.
2. Methodology
Within the RIVAS project, parametric studies of various mitigation measures to reduce rail-
way induced vibration were carried out by project partners (KU Leuven and ISVR) using the 2.5D
FE-BE method 8-10. The 2.5D approach requires that the system investigated has invariant geometry
in one direction, which is satisfied for most railway structures. By using a Fourier transform with
respect to the longitudinal coordinate, the three-dimensional (3D) response of the structure and the
radiated wave field can be determined using a two-dimensional (2D) FE-BE mesh in the frequency
domain.
At KU Leuven, the classical 2.5D FE method is combined with the 2.5D BE method using
2.5D Green's functions of a horizontally layered half-space 11, 12. In this way, the free surface and
the layer interfaces of the half-space do not have to be discretized with boundary elements, avoiding
spurious reflections at mesh truncations. The BE mesh can be limited to the interface between the
structure and the soil, significantly reducing the size of the BE mesh.
The 2.5D coupled FE and BE model used at ISVR 13 uses fundamental solutions of a homo-
geneous full space for the boundary elements. Therefore the ground surface and any layer interfaces
need to be carefully meshed to a sufficient distance. A special edge element is used to avoid the
reflections at the end of the ground or layer mesh. Fig. 1 shows a typical example of the structure
considered in the study and part of the mesh used in the numerical model: an open trench placed
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 3
next to the track and penetrating through the first layer of soil. The ground was meshed using 3-
node boundary elements with appropriate element size ensuring that there were at least 6 nodes per
wavelength up to 100 Hz for the soil considered in this study. The track system, including rail, rail
pad, sleeper and ballast, was meshed with 4-node and 8-node rectangular finite elements.
Figure 1. Geometry of the 2.5D model for an open trench next to a railway track. Only part of the free sur-
face and layer interface are shown
3. Results and discussion
3.1 Ground without measures
Investigations were first carried out using a typical two-layer soil condition where a soft top
layer is located over a much stiffer half-space. The properties of these two layers of soil (including
layer thickness h, shear wave speed Cs, compressional wave speed Cp, density ρ, and material
damping ratio β) are shown in Table 1. To show the impact of the layer depth on wave propagation,
the depth of the top layer was intentionally varied, taking values 0 m, 3 m, 6 m and infinite.
Table 1. Soil properties for typical two-layer ground.
Layer h [m] Cs [m/s] Cp [m/s] β [-] ρ [kg/m³]
1 0 / 3 / 6/ 150 298 0.03 2000
2 600 1191 0.03 2000
Figure 2. Receptance at 24 m for top layer depth
varying between 0 m (blue solid), 3 m (red dashed),
6m (purple dot-dashed) and infinite (black dotted).
Figure 3. Vertical velocity level at 24 m when train
passing by for soil conditions with top layer depth is
varying between 0 m (blue solid), 3 m (red dashed)
and infinite (black dotted).
The ground transfer receptances for cases without any measure are shown in Fig. 2. The track
was not considered here. From Fig. 2, for a homogeneous half-space, i.e. when the depth of the top
layer h1 = 0 m or infinite, the receptance is relatively flat. It is higher for the soft ground (h1 = infi-
nite) than the stiff ground (h1 = 0 m). When a soft top layer is present, surface waves are dispersive
-4 -2 0 2
-1
-0.5
0
0.5
y (m)
z (m)
2 4 8 16 31.5 63 100
10
-10
10
-9
1/3 Octave band centre frequency [Hz]
Receptance [m/N]
h1=0 m
h1=3 m
h1=6 m
h1=inf
2 4 8 16 31.5 63 100
20
40
60
80
100
120
1/3 Octave band centre frequency [Hz]
Vertical velocity level [dB re 10
-9
m/s]
h1=0 m
h1=3 m
h1=6m
h1=inf
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 4
due to the variation of the soil properties with depth. Not all surface waves propagate over the
whole frequency range: there exist cut-on frequencies such that the waves remain evanescent below
this frequency. From Fig. 2, the waves in the upper layer have cut-on above about 10 Hz for h1 = 6
m and 20 Hz for h1 = 3 m. Thus, a shallower top soft layer will result in a higher cut-on frequency.
Above the cut-on frequency, the response of the layered ground tends to that of the softer homoge-
neous soil.
Vertical velocity levels at 24 m from the centre of the track during a train pass-by are obtained
by using the TGV model from ISVR 14. The track system considered here is a reference track used
within the RIVAS project and the train considered is a four-coach EMU train which is also used as
reference train in RIVAS project 15. Results are shown in Fig. 3. This shows that the train will in-
duce high vibration levels for the soft homogeneous soil (h1 = infinite) from a very low frequency.
For a layered ground (h1 = 3 m), the vibration level jumps, reaching a peak around 20 Hz, which is
about the cut-on frequency. For a stiff homogeneous ground (h1 = 0 m), the vibration level is lower
than for the others, but still reaches a high level at high frequencies.
3.2 Open trench and soft wall barrier
Fig. 4 compares the insertion loss of a soft barrier and an open trench. Both are 3 m deep and
the results are shown at 24 m from the excitation. To simulate a line source, a series of point loads
were applied at various locations to represent the axle locations of a train. The track is omitted here
as from previous study 16 the track mobility has been found not to be affected by the trench or barri-
er. The open trench is 0.5 m wide and located 8 m from the loads. The soft barrier is 50 mm wide
and filled with a material with Young’s modulus 1 MPa, Poisson’s ratio 0.4 and density 700 kg/m3
(based on Regupol PL). Although the soft barrier is used as a practical alternative to an open trench,
the attenuation is compromised by the fill material. For a layered ground, both trench and soft wall
barrier start to have an effect from the cut-on frequency, around 20 Hz. For a homogeneous half-
space, the attenuation of the open trench rises at a frequency at which the depth is around 0.6 times
the Rayleigh wavelength: 12.5 Hz for the soft ground and 50 Hz for the stiff ground. The insertion
loss of the soft wall barrier also rises at around the same frequency.
Figure 4. Insertion loss of soft wall barrier (blue solid line) and open trench (red dashed line) at 24 m away
from the excitation for top soil layer depth of (a) h1 = 0 m, (b) h1 = 3 m and (c) h1 = infinite.
As the depth of trench and barrier has a great influence on the frequency range where the
measure is effective, it was studied for these three soil conditions. Fig. 5 shows the results for the
open trench. The trench is again 8 m away from the excitation, has a width of 0.5 m and its depth is
1.5 m, 3 m (the same as in Fig. 4) or 6 m. For a homogeneous half-space, as mentioned above, an
open trench is effective when the depth is at least 0.6 times the Rayleigh wavelength. A deeper
trench is thus able to attenuate Rayleigh waves with longer wavelengths (i.e. lower frequencies).
This can be seen in the results in Fig. 5: by increasing the depth of trench the effective frequency
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(a) h
1
= 0 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(b) h
1
= 3 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(c) h
1
= inf
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
Soft wall barrier
Open trench
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 5
range extends to lower frequencies and the attenuation at higher frequencies is greater. The inser-
tion loss is also larger for a softer ground.
For a layered ground, the performance is also influenced by the depth of the soil layers.
Above the cut-on frequency the waves are constrained within the upper layer and the trench pro-
vides some attenuation. This attenuation is greatest when the depth of the trench is sufficient to
penetrate fully the soft upper layers. As seen in Fig. 5b, increasing the depth of the trench has little
effect on the cut-on frequency but the effectiveness of the trench improves. When it penetrates
through the top layer (3 m deep), a further increase in the depth of the trench from 3 m to 6m has
little additional benefit in terms of performance.
Figure 5. Insertion loss at 24 m away from the excitation with varying depth of 0.5 m wide open trench for
typical soil conditions (Table 1), where top layer depth is (a) h1 = 0 m, (b) h1 = 3 m and (c) h1 = infinite.
 
Figure 6. Effect of varying soft barrier depth at
typical layered ground site from 1.5 m, 3 m to
6 m. Receiver position at 24 m.
Figure 7. Effect of varying soft barrier material properties
at typical layered ground site. Receiver position at 24 m.
The same investigation was carried out for the soft barrier. Fig. 6 shows results for a 50 mm
wide, soft wall barrier of various depths. This clearly shows that, similar to the open trench, increas-
ing the depth can improve the performance of the soft barrier, but the improvement is much less
than for the open trench. In the same way as for the open trench, when the soft barrier penetrates
through the top layer, further increase of the depth of barrier does not provide any more benefit.
Since the properties of the fill materials play an important role for a soft wall barrier, their in-
fluence was also studied. Results are shown in Fig. 7. The Young’s modulus was reduced from
1 MPa to 0.5 MPa and 0.25 MPa. The shear wave speed was also changed from 22 m/s to 11 m/s
while keeping the Young’s modulus constant (by changing the density). As seen from the results,
the Young’s modulus has a bigger impact on the insertion loss above the cut-on frequency than the
shear wave speed.
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(a) h
1
= 0 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(b) h
1
= 3 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(c) h
1
= inf
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
d=1.5 m
d=3 m
d=6 m
4 8 16 31.5 63 100
0
10
20
h
1
= 3 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
d=1.5 m
d=3 m
d=6 m
4 8 16 31.5 63 100
0
5
10
15
20 h1 = 3 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
E=1 MPa, Cs=22 m/s
E=0.5 MPa, Cs=22 m/s
E=0.25 MPa, Cs=22 m/s
E=1 MPa, Cs=11 m/s
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 6
3.3 Subgrade stiffening and wave impeding blocks
The results of introducing a 6 m wide and 1 m thick concrete block under the track are shown
in Fig. 8. A line of point sources is again used here but the track system is also included in the mod-
el as the track mobility is affected by these two mitigation measures 15. The result for a 3 m deep
open trench from Fig. 4 is also shown for comparison. The distance between the top surface of the
block and surface of the soil, h, varies from 0 m (corresponding to subgrade stiffening) to 1 m and 2
m (which forms a wave impeding block). From the results, for the frequencies larger than ~16 Hz,
the open trench gives a higher insertion loss in all cases. For lower frequencies (<16 Hz), the wave
impeding block and subgrade stiffening can achieve a higher insertion loss for the layered ground.
Although wave impeding blocks are intended to shift the cut-on frequency, this effect does not seem
very apparent compared to the corresponding subgrade stiffening. However, the performance in
each case drops at higher frequencies, and this seems to be related to the location of the block. For a
deeper block the performance drops at a lower frequency. This is seen most clearly for the soft ho-
mogeneous ground (Fig. 8c) but is also apparent for the layered ground (Fig. 8b). From the results
in Fig. 8 it appears that the wave impeding block does not change the cut-on frequency but appears
to act in the same way as the subgrade stiffening.
Figure 8. Effect of varying depth of concrete block using typical soil condition. The depth varies from 0 m
(blue solid line), 1 m (red dashed line), to 2 m (purple dot-dashed line). Receiver position at 24 m.
Fig. 9 shows the results varying the thickness of subgrade stiffening block, placed directly
under the track. A thicker block gives a better insertion loss, though this difference is less obvious
for a homogeneous half-space. For layered ground, a big increase of insertion loss can be found
when the block penetrates through the top layer.
Figure 9. Effect of varying thickness of subgrade stiffening concrete block using typical soil condition. Re-
ceiver position at 24 m.
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(a) h
1
= 0 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(b) h
1
= 3 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(c) h
1
= inf
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
h = 0 m
h = 1 m
h = 2 m
Open trench
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(a) h
1
= 0 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(b) h
1
= 3 m
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
4 8 16 31.5 63 100
-5
0
5
10
15
20
25
(c) h
1
= inf
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
d = 1 m
d = 2 m
d = 3 m
d = 4 m
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 7
3.4 Comparisons at reference sites
Studies were carried out by applying the same measures at three different reference sites. The
mitigation measures investigated include open trench (3 m deep, 1 m wide), soft barrier (6 m deep,
0.05 m wide, filled with Regupol PL), subgrade stiffening block (concrete, 6 m wide, 0.5 m thick)
and wave impeding block (concrete, 6 m wide, 0.5 m thick, 0.9 m under the ground surface). The
three soil reference sites are Horstwalde (a homogeneous half-space), Lincent (a layered ground
which has soft top layers with a total top layer depth of 4.1 m) and Furet (a layered ground which
has soft top layers with a total top layer depth of 12 m). Soil properties of these three reference sites
are listed in Table 2. Insertion losses at 24 m for all three reference sites are shown in Fig. 10.
Table 2. Soil properties for the reference sites.
Layer h [m] Cs [m/s] Cp [m/s] β [-] ρ [kg/m³]
Horstwalde 1 250 1470 0.025 1945
Lincent 1 1.4 128 286 0.044 1800
2 2.7 176 286 0.038 1800
3 355 1667 0.037 1800
Furet 1 2 154 375 0.025 1800
2 10 119 290 0.025 1850
3 200 490 0.025 1710
From Fig. 10, at Horstwalde the effectiveness of the open trench is mostly related to the Ray-
leigh wavelength. It has a larger insertion loss at Lincent than at Furet at most frequencies, because
the depth of trench, 3 m, is closer to its top layer depth (4.1 m) than at Furet (12 m).
The soft barrier works similarly at Lincent and Furet, but has a better performance at Horst-
walde where the soil is stiffer.
Subgrade stiffening and wave impeding blocks work similarly to each other at these three
sites. For the layered grounds they work better at low frequencies, although a negative effect ap-
pears at high frequencies.
Figure 10. Insertion loss of same measure at different reference sites.
4. Conclusions
Four different mitigation measures were investigated, which can be applied in the transmis-
sion path to reduce railway induced vibration. Studies were carried out under different soil condi-
tions. An open trench gives the best attenuation for most cases. For a layered ground its perfor-
mance is maximised when the trench penetrates through the soft upper layer. A soft wall barrier has
816 31.5 63 100
-5
0
5
10
15
20
25
(a)Horstwalde
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
816 31.5 63 100
-5
0
5
10
15
20
25
(b) Lincent
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
816 31.5 63 100
-5
0
5
10
15
20
25
(c) Furet
1/3 Octave band centre frequency [Hz]
Insertion loss [dB]
Open trench
Soft barrier
Subgrade stiff ening
Wave impeding block
21st International Congress on Sound and Vibration (ICSV21), Beijing, China, 13-17 July 2014
ICSV21, Beijing, China, 13-17 July 2014 8
a much lower effect than the corresponding open trench but it can still achieve 5-10 dB above a
certain frequency in most cases. For a suitable choice of fill material, it can be a more practical al-
ternative to an open trench. Subgrade stiffening and a wave impeding block can give good insertion
loss, about 5-10 dB in the effective frequency bands. Comparing the same mitigation method at
three example sites, it is shown that a change in soil condition (soil structure and properties) can
affect the performance of the designed mitigation measures by 5-10 dB. Therefore, a good under-
standing of the soil is essential in the design procedure.
Acknowledgements
The results presented in this paper have been obtained within the frame of the EU FP7 project
RIVAS (Railway Induced Vibration Abatement Solutions) under grant agreement No. 26575.
REFERENCES
1 Thompson, D. J., Railway Noise and Vibration. Mechanisms, modelling and means of control, Elsevier, (2008).
2 Alzawi, A. and ElNaggar, M., Full scale experimental study on vibration scattering using open and in-filled
(GeoFoam) wave barriers, Soil Dynamics and Earthquake Engineering, 31, 306–317, (2011).
3 Ahmad, S. and Al-Hussaini, T. M., Simplified design for vibration screening by open and infilled trenches,
Journal of Geotechnical Engineering, Proceedings of the ASCE, 117(1), 67– 88, (1991).
4 Beskos, D. E., Dasgupta, B., and Vardoulakis, I. G., Vibration isolation using open or filled trenches. Part I: 2-D
homogeneous soil, Computational Mechanics, 1, 43–63, (1986).
5 Kim, M., Lee, P., Kim, D. and Kwon, H., Vibration isolation using flexible rubber chip barriers. In Schmid, G.
and Chouw, N. (editors), Proceedings of the International Workshop Wave 2000, Wave propagation, Moving
load, Vibration reduction, 289–298. A.A Balkema, Amsterdam, (2000).
6 Andersen, L. and Nielsen, S. R. K., Reduction of ground vibration by means of barriers or soil improvement
along a railway track, Soil Dynamics and Earthquake Engineering, 25, 701– 716, (2005).
7 Peplow, A. T., Jones, C. J. C., and Petyt, M., Surface vibration propagation over a layered halfspace with an in-
clusion, Applied Acoustics, 56, 283–296, (1999).
8 Sheng, X., Jones, C. J. C. and Thompson, D. J., A theoretical study on the influence of the track on train-induced
ground vibration, Journal of Sound and Vibration, 272, 909-936, (2004).
9 Sheng, X., Jones, C. J. C. and Thompson, D. J., A theoretical model for ground vibration from trains generated
by vertical track irregularities, Journal of Sound and Vibration, 272, 937-965, (2004).
10 Galvín, P., François, S., Schevenels, M., Bongini, E., Degrande, G. and Lombaert, G., A 2.5D coupled FE-BE
model for the prediction of railway induced vibrations, Soil Dynamics and Earthquake Engineering, 30(12),
1500-1512, (2010).
11 François, S., Schevenels, M., Lombaert, G., Galvín, P. and Degrande, G., A 2.5D coupled FE-BE methodology
for the dynamic interaction between longitudinally invariant structures and a layered halfspace, Computer Meth-
ods in Applied Mechanics and Engineering, 199(23-24), 1536-1548, (2010).
12 Schevenels, M., François, S., and Degrande, G., EDT: An ElastoDynamics Toolbox for MATLAB, Computers
& Geosciences, 35(8),1752-1754, (2009).
13 Nilsson, C. M., and Jones, C. J. C., Theory manual for WANDS 2.1, ISVR Technical Memorandum 975, Uni-
versity of Southampton, (2007).
14 Sheng, X., Jones, C.J.C. and Petyt, M., Ground vibration generated by a load moving along a railway track,
Journal of Sound and Vibration, 228 (1), 129–156, (1999).
15 Toward, M.G.R., Jiang, J., Dijckmans, A., Coulier, P., Thompson, D.J., Degrande, G., Lombaert, G. and Hussein,
M.F.M, Mitigation of railway induced vibrations by using subgrade stiffening and wave impeding blocks, 9th In-
ternational Conference on Structural Dynamics (EURODYN 2014), Porto, Portugal , 30 June – 2 July, (2014).
16 Jiang J., Toward M., Dijckmans A., Thompson D., Degrande G., Lombaert G. and Ryue J., Reducing railway in-
duced ground-borne vibration by using trenches and buried soft barriers. In Nielsen J., et al. (eds.) 11th Interna-
tional Workshop on Railway Noise, Uddevalla, Sweden, 9-13 September, 615-622, (2013).
... This modified wave propagation regime results in the reduction of low-frequency vibration in the transmission path. Besides, Jiang et al. [46] found that WIB could achieve better performance in layered ground but the increased depth of stiffened block has little effect on the improvement of the performance, which should be considered in the design period. It was found that WIB can suppress the ground vibrations between 6 and 8 dB in a low-frequency range (<30 Hz) [47,48]. ...
Article
Full-text available
Noises and vibrations caused by operating transport systems can seriously affect people’s health and environmental ecosystems. Railway-induced vibrations in urban settings can cause disturbances and damages to surrounding buildings, infrastructures and residents. Over many decades, a number of mitigation methods have been proposed to attenuate vibrations at the source, in the transmission path, or at the receiver. In fact, low-frequency or ground-borne vibration is turned out to be more difficult to be mitigated at source, whilst some attenuation measures in propagation path can be applicable. To broaden the mitigating range at the low-frequency band, the applications of meta-materials/structures have been established. In railway systems, periodic structures or resonators can be installed near the protected buildings to isolate the vibrations. Despite a large number of proposed attenuation methods, the sustainability of those methods has not been determined. Based on rational engineering assumptions, the discounted cash flows in construction and maintenance processes are analysed in this study to evaluate lifecycle costs and the quantity of materials and fuels, as well as the amount of carbon emissions. This study is the world’s first to identify the efficacy and sustainability of some transmission path attenuation methods in both normal and adverse weather conditions. It reveals that geofoam trenches and wave impeding blocks are the most suitable methods. Although metamaterial applications can significantly mitigate a wider range of lower frequency vibrations, the total cost and carbon emissions are relatively high. It is necessary to significantly modify design parameters in order to enable low-cost and low-carbon meta-materials/structures in reality.
... Berbagai penelitian telah diupayakan untuk mengurangi efek vibrasi ini termasuk mengenal lebih baik karakteristik vibrasi yang dihasilkan. Penelitian tersebut antara lain dalam bidang mitigasi vibrasi (Kim and Lee, 2000), (Bahrekazami et al, 2004), (Bei Su, 2005), (Erkal, 2010), (Dijckmans et al, 2012), (Motazedian et al, 2012), dan (Jiang, 2014). Dalam bidang simulasi numerik vibrasi antara lain (Holm, 2002), (Kogut, 2003), (XueCheng et al, 2008) dan (Zhang et al, 2016). ...
Conference Paper
Full-text available
Abstrak Keberadaan kereta api trans kalimantan telah lama diidam-idamkan oleh penduduk Pulau Kalimantan. Salah satu aspek teknis yang seringkali menjadi kendala terealisasinya infrastruktur ini adalah kondisi tanah kalimantan yang memiliki daya dukung yang buruk bagi fondasi jalan kereta api, sehingga biaya konstruksi fondasi menjadi sangat mahal. Terlepas dari hal tersebut, aspek teknis lainnya yang perlu memperoleh perhatian adalah vibrasi (getaran) yang dihasilkan baik pada masa konstruksi maupun masa operasional kereta api. Pengaruh vibrasi ini semakin penting untuk ditinjau mengingat salah satu trase kereta api tersebut melewati wilayah Kota Pontianak yang padat permukiman. Penelitian dilakukan dengan menggunakan formulasi teoritis dengan memakai persamaan empiris kecepatan partikel puncak. Hasil kalukulasi menunjukkan bahwa besaran PPV berkisar antara 3700 mm/s hingga 9900 mm/s dengan kisaran frekuensi antara 6,5 Hz hingga 22 Hz. Hasil ini selanjutnya dibandingkan dengan standar getaran yang dikeluarkan oleh Departemen Lingkungan Hidup No. KEP-48/MENLH//11/96. Hasil perbandingan menunjukkan bahwa vibrasi yang ditimbulkan akibat keberadaan kereta api menimbulkan beberapa dampak negatif bagi permukiman antara lain kebisingan dan penurunan fondasi bangunan yang sebagian besar dibangun diatas tanah lunak Kota Pontianak. Katakunci: vibrasi, kereta api trans kalimantan, tanah lunak Kota Pontianak Abstract The existence of Trans Kalimantan Railways has been long time expected by the people of Kalimantan. One of the technical aspect that hamper this infrastructure to be realized is kalimantan soil's condition that has a low bearing capacity for the foundation of a railways, so that the construction of the foundation become very expensive. Beyond this, there is also another technical aspect that should be taken into acccount namely vibration. Vibration should be considered either at the construcion stage or at the operational stage. The effect of vibration is signifacantly have to be observed considering that the railways will cross through the City of Pontianak that is dense with dwellings. The research is conducted using theoretical formulation with the help of emperical equation of Peak Particle Velocity (PPV). The results show that the PPV is ranging between 3700 mm/s until 9900 mm/s with frequency between 6,5 Hz until 22 Hz. Furthermore, the results are compared to the vibration standards issued by Department of Living Environment No. KEP-48/MENLH//11/96. The results show that the vibrations cause negative effect on the dwelings among other noise and foundation settlement due to most of the dwellings are built on Pontianak soft soil.
Article
Ground-borne vibrations due to high-speed trains passage strongly depend, apart from the speed of the train, on the geometry of the railways as well as the properties of the underlying soil layer(s). The main aim of this study is to investigate the effectiveness of expanded polystyrene (EPS) blocks in mitigating soil vibrations induced on railway embankments for different subsoil and railway embankment material conditions. The EPS blocks are placed in suitable locations, either as embankment's side fill material, or trench filling material, or combination of the above. An efficient three-dimensional numerical model has been developed -in conjunction with a user-developed subroutine for applying the moving loads-to accurately calculate the dynamic response of the coupled embankment-soil model. Four typical soil types - categorized as rock, dense sand with gravels, stiff and soft clay - are investigated. In addition, the mechanical properties of the embankment material have been altered to assess to what extend they can affect the HST vibrations.
Article
Full-text available
[For parts I,II see: the last two authors and B. Dasgupta, ibid. 1, 43-63 (1986; Zbl 0614.73093), and ibid. 6, 129-142 (1990; Zbl 0725.73076).] The problem of isolating structures from surface waves by open or filled trenches under conditions of plane strain is numerically studied. The soil is assumed to be an isotropic, linear elastic or viscoelastic nonhomogeneous (layered) half-space medium. Waves generated by the harmonic motion of a rigid surface machine foundation are considered. The formulation and solution of the problem are accomplished by the frequency domain boundary element method. The Green function of Kausel-Peek-Hull for a thin layered half-space is employed and this essentially requires only a discretization of the trench perimeter and the soil-foundation interface. The proposed methodology is used for the solution of a number of vibration isolation problems and the effect of soil inhomogeneity on the wave screening effectiveness of trenches is discussed.
Article
Full-text available
This paper studies the efficiency of subgrade stiffening next to the track as a mitigation measure for railway induced vibrations by means of a two-and-a-half-dimensional coupled finite element–boundary element methodology. An analysis in the frequency–wavenumber domain for a homogeneous halfspace reveals that the block of stiffened soil next to the track can act as a wave impeding barrier. It is demonstrated that the wave impeding effect depends on the relation between the Rayleigh wavelength in the soil and the free bending wavelength in the block of stiffened soil, as the transmission of plane waves in the soil with a longitudinal wavelength smaller than the bending wavelength is hindered. This leads to a critical frequency from which this mitigation measure starts to be effective, depending on the stiffness contrast between the soil and the block of stiffened soil. The existence of a critical angle delimiting an area where vibration levels are reduced in case of harmonic excitation on the rail is also demonstrated. Two applications involving a layered halfspace are finally discussed to demonstrate that the performance of this mitigation measure critically depends on the soil characteristics.
Article
Full-text available
Rectangular open or in-filled trenches (wave barriers) are often used in engineering practice to reduce the ground vibrations caused by propagating surface (Rayleigh) waves of relatively small wavelengths. This paper presents simple design expressions for estimating the vibration screening effectiveness of rectangular wave barriers in homogeneous soil deposits. The design formulas are developed by conducting an extensive numerical investigation on the influence of various geometrical and material parameters on the vibration screening effectiveness of the barriers. An advanced direct Boundary Element Method (BEM) incorporating higher-order isoparametric elements and a sophisticated self-adaptive numerical integration scheme is used for this study. Dimensionless parameters that govern a barrier's performance are then identified, and models are developed taking these parameters into account in a simplified way. Through some comparisons with results from the rigorous BEM code and available experimental data, the applicability of the model is demonstrated. Some layered soil cases are also studied, and the important effects of layering on the performance of a trench as a wave barrier are presented.
Conference Paper
Full-text available
To reduce railway induced low frequency vibration, two mitigation measures - open trenches and buried soft wall barriers have been studied in this paper by using coupled finite element-boundary element models. These models were developed at KU Leuven and ISVR, and have been cross-validated within the EU FP7 project RIVAS (Railway Induced Vibration Abatement Solutions). Variations in the width, depth, location of trench and properties of soft barrier material are considered under various soil conditions. Results show that in all ground conditions, the notional rectangular open trench performs better than the other constructions. The width of an open trench has little influence on its performance, whereas increasing the width of a filled trench reduces the stiffness of the barrier, improving the performance of the trench. Likewise, fill materials with lower Young’s modulus give higher insertion losses.
Article
Full-text available
The isolation of structures from ground transmitted waves by open and infilled trenches in a three-dimensional context is numerically studied. The soil medium is assumed to be elastic or viscoelastic, homogeneous and isotropic. Waves generated by the harmonic motion of a surface rigid machine foundation are considered in this work. The formulation and solution of the problem is accomplished by the boundary element method in the frequency domain. The infinite space fundamental solution is used requiring discretization of the trench surface, the soil-foundation interface and some portion of the free soil surface. The proposed methodology is first tested for accuracy by solving three characteristic wave propagation problems with known solutions and then applied to several vibration isolation problems involving open and concrete infilled trenches. Three-dimensional graphic displays of the surface displacement pattern around the trenches are also presented.
Article
Full-text available
The problem of structural isolation from ground transmitted vibrations by open or infilled trenches under conditions of plane strain is numerically studied. The soil medium is assumed to be linear elastic or viscoelastic, homogeneous and isotropic. Horizontally propagating Rayleigh waves or waves generated by the motion of a rigid foundation or by surface blasting are considered in this work. The formulation and solution of the problem is accomplished by the boundary element method in the frequency domain for harmonic disturbances or in conjunction with Laplace transform for transient disturbances. The proposed method, which requires a discretisation of only the trench perimeter, the soil-foundation interface and some portion of the free soil surface on either side of the trench appears to be better than either finite element or finite difference techniques. Some parametric studies are also conducted to assess the importance of the various geometrical, material and dynamic input parameters and provide useful guidelines to the design engineer.
Book
Railways are an environmentally friendly means of transport well suited to modern society. However, noise and vibration are key obstacles to further development of the railway networks for high-speed intercity traffic, for freight and for suburban metros and light-rail. All too often noise problems are dealt with inefficiently due to lack of understanding of the problem. This book brings together coverage of the theory of railway noise and vibration with practical applications of noise control technology at source to solve noise and vibration problems from railways.
Article
The ElastoDynamics Toolbox (EDT) version 2.1 offers an extensive set of MATLAB functions to model elastodynamic wave propagation in horizontally layered media. The toolbox is based on the direct stiffness method and the thin layer method. These methods provide stiffness matrices for a homogeneous layer and a homogeneous halfspace, which are formulated in the frequency-wavenumber domain. EDT 2.1 can be used to solve a variety of problems governed by wave propagation in the soil, such as (1) site amplification, (2) the computation of dispersive wave modes in layered soils, and (3) the calculation of the forced response of the soil due to harmonic and transient loading. The toolbox serves as an electronic learning environment for the simulation and processing of seismic wave propagation in layered media. It has also been used by various authors to model wave propagation in layered soils