Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering

journal homepage: www.elsevier.com/locate/soildyn

Field testing and analysis of high speed rail vibrations

D.P. Connolly a,n, G. Kouroussis b, P.K. Woodward a, P. Alves Costa c, O. Verlinden b,
M.C. Forde d
a
Heriot-Watt University, Institute for Infrastructure & Environment, Edinburgh, UK
b
Department of Theoretical Mechanics, Dynamics and Vibrations, University of Mons, 31 Boulevard Dolez, B-7000 Mons, Belgium
c
University of Porto, Faculty of Engineering, Porto, Portugal.
d
University of Edinburgh, Institute for Infrastructure and Environment, School of Engineering, AGB Building, The Kings Buildings, Edinburgh, UK

art ic l e i nf o a b s t r a c t

Article history: This paper outlines an experimental analysis of ground-borne vibration levels generated by high speed
Received 20 May 2014 rail lines on various earthwork profiles (at-grade, embankment, cutting and overpass). It also serves to
Received in revised form provide access to a dataset of experimental measurements, freely available for download by other
29 August 2014
researchers working in the area of railway vibration (e.g. for further investigation and/or the validation
Accepted 31 August 2014
Available online 26 September 2014
of vibration prediction models).
First, the work outlines experimental investigations undertaken on the Belgian high speed rail
Keywords: network to investigate the vibration propagation characteristics of three different embankment
High speed rail conditions. The sites consist of a 5.5 m high embankment, an at-grade section and a 7.2 m deep cutting.
Railway
The soil material properties of each site are determined using a ‘Multichannel Analysis of Surface Waves’
Vibration prediction
technique and verified using refraction analysis. It is shown that all sites have relatively similar material
Ground borne vibration
Download data properties thus enabling a generalised comparison.
Embankment Vibration levels are measured in three directions, up to 100 m from the track due to three different
Cutting train types (Eurostar, TGV and Thalys) and then analysed statistically. It is found that contrary to
Environmental impact assessment commonly accepted theory, vertical vibrations are not always the most dominant, and that horizontal
EIA vibrations should also be considered, particularly at larger offsets. It is also found that the embankment
Ground-borne noise earthworks profile produced the lowest vibration levels and the cutting produced the highest.
Experimental testing
Furthermore, a low (positive) correlation between train speed and vibration levels was found. A
MASW
selection of the results can be downloaded from www.davidpconnolly.com.
In-situ testing
& 2014 Elsevier Ltd. All rights reserved.

1. Introduction the train weight (quasi-static excitation), from changes in support

stiffness (e.g. regularly spaced sleepers) and irregularities in the
The rapid uptake of high speed rail has been in-part due to its wheel/rail geometry (dynamic excitation) [50,12]. Additionally,
superior economic, social and environmental benefits [6] in vibration amplitude levels may be elevated if the train speed
comparison to other modes of transport. On-going research into becomes comparable with the natural Rayleigh wave speed in the
aerodynamics, construction materials and motor technology has supporting soil [17,21,24,37,43], or if the excitation frequency is
allowed for the development of lightweight trains capable of close to a track natural frequency [23].
reaching increasingly higher speeds. Japan holds the world record These vibrations can cause significant negative effects such as
for the fastest high speed rail velocity of 581 km/h which is close personal distress in communities residing close to the lines.
to the speed experienced by a typical commercial jet. Therefore it is important to predict vibration levels before the line
One negative environmental side effect of high speed rail is the is constructed [11,13]. A vast body of prediction models has been
elevated levels of ground-borne vibration generated [7]. These proposed for investigating vibration levels on at-grade track
vibrations are generated at the wheel/rail interface and arise from sections [3,15,26,36,37,40,47,51] and underground lines [1,27,29,
30,42,45,52]. Despite this, research related to railway vibrations
under different earthwork profile conditions is scarce.
n
Corresponding author. An advantage of an experimental study over a numerical one is
E-mail addresses: [email protected] (D.P. Connolly),
that a reduced number of modelling assumptions are required. For
[email protected] (G. Kouroussis),
[email protected] (P.K. Woodward), [email protected] (P. Alves Costa), example, [20] presented an analytical model for the investigation
[email protected] (O. Verlinden), [email protected] (M.C. Forde). of vibrations due to an embankment and it was shown that the

http://dx.doi.org/10.1016/j.soildyn.2014.08.013
0267-7261/& 2014 Elsevier Ltd. All rights reserved.
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 103

embankment was a source of high frequency vibration. Despite investigation and for the validation of numerical prediction
this, the embankment was assumed to have vertical sides and the models.
train excitation was uncoupled from a simplified track model.
Another approach was presented by [8] who used a 3D finite
element (FE) modelling approach to analyse vibrations within 2. Test site details
embankments with varying stiffness. It was shown that stiff
embankments provided superior vibration performance in com- 2.1. General
parison to soft ones. A drawback of the FE approach is that
assumptions must be made concerning the distribution of soil 2.1.1. Site 1—At-grade
properties, and high frequency content can be difficult to simulate. Site 1 consisted of an at-grade railway section (Figs. 1 and 2)
To overcome some of the limitations associated with numerical 4 km south of the town of Leuze-en-Hainaut. The track was a
analysis [18,25,31,32], performed experimental analysis on at- classically ballast track composed of ballast, subballast and sub-
grade railway tracks to analyse the characteristics of railway grade layers, with thicknesses 0.3, 0.2 and 0.5 m, respectively. The
vibration. Despite this, few investigations have been undertaken rails were continuously welded UIC 60 rails with a mass of 60 kg/
into embankment vibration. One of the few studies used accel- m3 and fixed to the prestressed concrete sleepers (300 kg mono-
erometers to record ground movement on the rail, sleeper and an block) via Pandrol clips (Fig. 3). The rails were also supported by
embankment made from compacted gravel [39]. It was found that railpads with thickness 0.01 m. The irregularity of the rails (for all
the dominant frequencies within the embankment were between test sites) was assumed to be very low because grinding had been
40 and 70 Hz, with the spectrum reducing in frequency with performed eight days before testing. It was also assumed that the
distance from the embankment shoulder. Unfortunately the standard of track geometry was high and identical across all test
results were not compared to non-embankment data. sites.
To the authors’ knowledge, there is no published literature Two distinct test setups were deployed, the first to record three
related to the experimental analysis of vibration from railway component vibration levels at distances of 9 –35 m from the
cuttings. Therefore this paper attempts to compare the vibration closest track (Table 1), and the second to record vertical vibration
levels generated by cuttings, embankments and at-grade track between 9 and 100 m from the track (Table 2). The first setup was
sections, via field experiments [10,33]. First, experimental inves- composed of 8 low frequency, 3 component, SM-6 geophones,
tigations are performed at three Belgian test sites. Vibration levels with sensitivity 28.8 V/m/s (Fig. 4). For the second setup, 24 low
are recorded in all three component directions and vertical frequency, 1 component (vertical), SM-6 geophones, also with
vibrations are recorded up to a distance of 100 m from the track. sensitivity 28.8 V/m/s were used.
All sites are found to consist of similar soil characteristics as During post-processing, for each velocity time history recorded,
determined through Multichannel Analysis of Surface Waves the low frequency content was amplified by multiplying it by
(MASW) testing, thus allowing for a general comparison between the inverse of the geophone response curve. This ‘corrected’
vibration characteristics. In addition to earthwork profile condi- the geophone response which otherwise would have inaccurately
tions, the effect of train type, horizontal vibration and abutment recorded frequency content below 4.5 Hz.
presence are investigated. A key aim of this paper is to provide a
series of vibration records that researchers can use for further

Fig. 1. At-grade track section. Fig. 3. Ballasted track configuration (all sites).

Fig. 2. At-grade track section geophone configuration.

104 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Table 1
Three component geophone distances.

Distance from rail (m) 3 Component measurements

9 11 15 19 23 27 31 35

Component(s) measureda H1, H2, V1 H1, H2, V1 H1, H2, V1 H1, H2, V1 H1, H2, V1 H1, H2, V1 H1, H2, V1 H1, H2, V1

a
H1 ¼ Horizontal component, H2 ¼ horizontal component, V1 ¼ vertical component.

Table 2
One component geophone distances.

1 Component measurements

Distance from rail (m) 9 11 13 15 21 25 29 33 37 41 45 49

Component(s) measureda V1 V1 V1 V1 V1 V1 V1 V1 V1 V1 V1 V1
Distance from rail (m) 53 57 61 69 73 77 81 85 89 93 97 100
Component(s) measureda V1 V1 V1 V1 V1 V1 V1 V1 V1 V1 V1 V1

a
V1 ¼ vertical component.

2.1.4. Site 4—Abutment

Site 4 was located approximately 100 m East of site 2 and thus
the track components were identical that of test site 2. The
embankment was also identical to site 2 except that there was a
concrete under-pass passing through the embankment and
beneath the track.
This under-pass served as a minor road for car passage and is
shown in Fig. 9. At this site a hybrid geophone setup was deployed,
combining aspects of both of the previously described setups.

2.2. Train characteristics

Four train set configurations were recorded across all sites

during the measurement campaign: TGV, Eurostar, Thalys and
Fig. 4. In-field deployment of a three component geophone. double-Thalys. A brief description of each train follows, with the
majority of train properties obtained from Ref. [35]. Sample time
histories from the passage of a Thalys train are shown in Fig. 14.

2.2.1. TGV Reseau (TGV)

TGV train-sets are manufactured by Alstom and commenced
commercial operation in 1993. The TGV-R is the successor to the
TGV Atlantique. During testing, each train-set consisted of two
power cars at each end (Y230A), six passengers cars in the centre
(Y237A) and two lateral cars (Y237B) connecting the power and
passenger cars. Bogies were shared between passenger cars and
the power cars had two separate bogies each (Fig. 10). Table 3
shows the specification of the TGV train-set.

2.2.2. Thalys and Thalys double (Thalys)

Fig. 5. Embankment track configuration. Thalys high speed train sets commenced operation on Eur-
opean high speed lines in 1998 and have a maximum commercial
2.1.2. Site 2—Embankment speed of 300 km/h. They are derived from the TGV and manufac-
Site 2 was also located on the Paris–Brussels line, North-East of tured by Alstom. The total train length spans 200 m. Double Thalys
the town of Braffe. The track configuration consisted of an train sets use identical cars as the single Thalys, however there is
embankment 5.5 m high with a slope of 30 degrees (Figs. 5 and twice the number of passenger cars. The layout and configuration
6). The experimental methodology and geophone arrangement of the Thalys locomotives is shown in Figs. 11 and 12 and Table 4.
was consistent with site 1.
2.2.3. Eurostar TransManche (Eurostar)
The Eurostar was manufactured by Alstom and has been
2.1.3. Site 3—Cutting operational since 1993. Its length of 394 m makes it longer than
Site 3 was also located on the Paris–Brussels line, North-West both the Thalys and TGV and it is capable of holding 750
of the town of Braffe (Figs. 7 and 8). The track configuration passengers. In common with the Thalys and TGV train-sets, wheel
consisted of a cutting (excavated embankment), 7.2 m high at a spacing is identical and it consists of three car types: driving cars
gradient of 25 degrees. The track components were identical that at the ends, lateral cars next to the driving cars and passenger cars
of test site 1. in the middle. The entire train-set consists of 20 carriages. Wheel
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 105

Fig. 6. Embankment track section geophone configuration.

array was placed far enough away from the track to ensure the
results were not contaminated from potential artefacts close to
the line, but close enough to ensure that the soil properties were
representative of those beneath the track. No MASW measure-
ments were undertaken during train passage.
Geophone spacing was 1 m as recommended by [44] and each
sensor was coupled to the ground using 150 mm spikes [48].
Excitation was performed at 7 individual locations by striking an
embedded metal impact plate. All results were amplified using a
high gain setting and recorded using a Panasonic Toughbook in
SEG-2 format. The gain was removed during post-processing.

Fig. 7. Cutting track section.

2.3.2. Multichannel analysis of surface waves

layout is shown in Fig. 13 and the train-set specifications are The MASW results were analysed using Geopsy [54] and sub-
shown in Table 5. program Dinver [53]. Geopsy is a graphical user interface (GUI)
capable of generating dispersion curve plots from recorded signals.
2.2.4. Train speed calculation From these plots, the best fit dispersion curves were chosen
Approximate train speeds were obtained using information visually and exported for use in sub-program Dinver.
provided by the train operator, Infrabel. In an attempt to maximise To perform the inversions using Dinver, density was held con-
accuracy, train speeds were also determined independently using stant at 2000 kg/m3. Shear wave (S-wave) speed is highly indepen-
a newly developed calculation procedure [34]. This procedure used dent of density and therefore density is typically held constant to
a combination of cepstral analysis, dominant frequency analysis increase the accuracy and reliability of the MASW process. Inversion
and a regression analysis (based upon minimising the error was then used to calculate the layer depths and wave speeds of the
between experimental frequencies and an analytical quasi-static underlying soil. Compressional wave (P-wave) profiles were vali-
excitation solution). Although all three approaches varied in dated using a refraction analysis, performed using the commercial
nature, the underlying methodology was similar, i.e. to isolate seismic software package, SiesImager/2D. The sub-module PickWin
the key vehicle frequencies (Fig. 15) and use them to calculate the was used to identify first arrivals and sub-module Plotrefa was used
train speed. For all the high speed train speeds computed in this to calculate the P-wave velocity profile. Geopysy MASW results were
study it was found that although all three techniques worked well, found to be consistent with SeisImager results.
it was sufficient to focus on using cepstral analysis. If speed As an additional check, a desktop study was undertaken by
information was required for alternative train types (e.g. freight) comparing results to existing soil information. For sites 1–4,
then this may not have been the case. After analysis, it was found generalised soil maps were available describing the soil layer
that for all 56 recorded train passages, the minimum speed was permutations and composition of each layer. For all sites, the
280.1 km/h, the maximum speed was 303.6 km/h, and the average experimental findings were generally consistent with the existing
train speed was 294.7 km/h. records. Once the wave speeds had been determined with con-
fidence, the Young’s modulus was calculated using elementary
material property relationships.
2.3. Ground dynamic characterisation

To determine the material properties of the soils at each test 2.3.3. Classification of soil properties
site, MASW was used in conjunction with a desktop survey of Fig. 17 describes the soil properties associated with each test
existing soils data. site (for further details see Appendix). As test site 4 was in very
close proximity ( o100 m) to test site 2 (Belgian embankment
2.3.1. Experimental setup site), no MASW tests were undertaken and the soil properties
The MASW experimental setup is shown in Fig. 16. Excitation were assumed to be identical to site 2.
was provided using a 12lb PCB 086D50 impact hammer with on- The resulting soil properties were in good agreement with
board accelerometer. The accelerometer was connected to a data existing soil records from the area and were also similar to those
acquisition unit using a microdot connector. This allowed for presented by [36] for previous spectral analysis of surface waves
calculation of the input force exerted by each hammer blow. (SASW) tests undertaken on nearby soils. Fig. 17 shows that the
Twenty four low frequency (4.5 Hz), vertical component, SM-6 soil properties at all three sites were similar in regards to wave
geophones were placed parallel to the railway track, in the same speed profile and layer depth/orientation. The only inconsistency
line as the geophones used for recording train vibrations. The was at site 3 which was underpinned by a layer of clay that was
106 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Fig. 8. Cutting track section geophone setup.

investigation and for the validation of numerical prediction

models. One common assumption used for railway vibration
modelling is that it is valid to model the problem using a linear
system of equations. This reduces model complexity and is
assumed valid because railway vibrations are often considered to
generate low level strains. To investigate the validity of this
assumption, the shear strain levels were investigated at each site.
Shear strain levels were estimated from the experimental results
using the equation [46]:
PPVi
γi ¼
V si

where PPV was the peak particle velocity, γ was shear strain, Vs
was the shear wave velocity of the upper soil layer and i was the
Fig. 9. Abutment site.
measurement location.
Fig. 20 shows three best fit curves displaying how shear strain
stiffer than the other two sites. A comparison between experi- varied from the track. For each earthwork profile, the passage of
mental the theoretical dispersion curves is shown in Fig. 18. five trains was considered and smoothed using higher order
polynomials. It can be noticed that the shear strain is at a
2.3.4. Soil damping calculation maximum at the nearest location to the track and reduces rapidly
Attenuation of vibration is primarily caused by material damp- in the far field. This was expected as waves both loose energy due
ing and geometrical damping. Geometrical damping describes the to material damping and spread energy due to radiation damping.
spreading of wave energy and is a function of soil geometry, while Assuming the soil was a clay with plasticity index (PI¼ 30%), it can
material damping describes the energy dissipation within soil be assumed to behave linearly for shear strains less than
particles. It has been shown that damping is dependent on 5  10  3%. After this threshold, the soil will start to exhibit non-
excitation frequency [28], which can be described by hysteric linear behaviour and at approximately 5  10  2% it will become
damping using linear complex stiffness parameters. highly non-linear [55]. As the maximum shear strain experienced
Methods to assess damping profiles from experimental data at each site was much lower than this threshold (1.15  10  3%), it
include the half-power bandwidth method [4], phase and ampli- was evident that the soil at all measured locations was behaving in
tude regression in the frequency-space domain [38], and fre- a visco-elastic manner.
quency–wavenumber amplitude regression [45]. A challenge
with these methods is that they depend on a very high coherence 3.2. Train speed
between signals. Therefore, [16,14] proposed an alternative solu-
tion which minimises the experimental and theoretical mobility Fig. 21 shows the relationship between train speed and ‘peak
(i.e. the velocity transfer functions). This approach is well suited to particle velocity’ (PPV). Similarly, Fig. 22 shows the relationship
MASW testing and therefore was used (i.e. the damping calcula- between train speed and Velocity decibels (VdB). PPV was calcu-
tions were performed by post-processing the data recorded during lated as:
the hammer excitation rather than during train passage).  
PPV ¼ maxvðtÞ
Fig. 19 compares the experimental and theoretical vertical
mobilities for three different source-receiver positions using the where v(t) was the velocity time history. VdB was calculated as:
damping profile shown Fig. 17a. Similar results were also obtained νrms
for sites 2 and 3. The agreement between experimental and VdB ¼ 20log 10
ν0
theoretical results was found to be acceptable, despite small
discrepancies between results. These discrepancies may have been where vrms was the root mean square of the time averaged signal
caused by factors such as the anisotropic behavior of the soil, and (over a one second period), and v0 was the background level of
have also been encountered by Refs. [2,16,41]. vibration (assumed to be 2.5  10  8 m/s, [22]).
For each metric the response at both the near and far receivers
was plotted along with a best fit line. It was found that there was a
3. Results and discussion reasonably large scatter, particularly for the train passages on the
near track. The standard deviation of PPV was 6.4  10  4 and
3.1. General remarks 4.4  10  4 m/s, for the near and far tracks, respectively. Similarly,
standard deviation of VdB was 2 and 1.5 dB, for the near and far
One of the aims of the experimental testing was to provide a tracks, respectively. It should be noted that a proportion of this
series of vibration records that researchers could use for further may have been caused by differences in train load (passenger
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 107

Fig. 10. TGV configuration.

Fig. 11. Thalys configuration.

For the bottom right figure, it was found then when all sites
and all trains were plotted together there was a large level of
scatter, however the best fit curves showed that the PPV levels
were similar. This was true for the PPV levels at all distances from
the track. Similarly, the best fit curve for all points closely followed
those exhibited by the Eurostar, TGV and Thalys trains.

3.4. Three component vibration levels

Figs. 24–26 show the variation in PPV levels for the 3 earthwork
profile configurations. For each figure, on the left, the average PPV
levels were plotted and on the right, the PPV levels from an
individual Thalys train were shown. Regarding the averaged levels,
as train type was found to be non-influential on vibration levels
(as noted elsewhere in this work) and the deviation between all
recorded train speeds was low, this allowed for all train passages
Fig. 12. Thalys passage at site 2 (embankment).
to be averaged.
To eliminate bias, for each case, the vibration levels were
averaged using only train passages occurring on either the near
numbers) and earthwork profile. Despite this, a tentative best fit or far track. This prevented skewing results due to the 4.5 m offset
line between train speed and PPV is shown in each diagram (black between tracks. As mentioned, no distinction was made between
line). Therefore, although all four lines showed a positive trend the passage of TGV, Thalys and Eurostar trains.
between vibration level and speed, the scatter was too large to It was noticed for all figures, especially when averaged, that at
conclude a definitive correlation. This is consistent with the locations near the track, vertical vibration levels were dominant,
findings by Ref. [18]. particularly for the embankment and cutting cases. This was
consistent with the results presented by Ref. [32]. Despite this,
as the distance from the track increased, the vertical vibration
3.3. Train type comparison levels decreased rapidly and became comparable with the hor-
izontal PPV levels. This effect was very clear for the at-grade and
Fig. 23 shows a comparison of VdB levels between all 3 train cutting cases where at 30–35 m from the track, unexpectedly,
types. For the at-grade site, all train speeds were within a range of the vertical vibration levels were significantly lower than the
16.7 km/h, for the embankment site they were within 15.6 km/h horizontal ones.
and for the cutting site, train speeds were within 6.2 km/h. The
individual PPV records for each train are shown along with a best
fit curve, which helped to remove some of the uncertainty
associated with possible variances in train weight (e.g. due to 3.5. The effect of earthwork profiles
changes in passenger numbers). To maximise the number of
records, trains from both the near and far tracks were plotted, Fig. 27 shows the effect of earthwork profile configuration on
albeit with a 4.5 m offset to account for railway track spacing. vertical vibration levels for both the near and far tracks. Consider-
For all three sites it was found that the PPV levels were similar ing both tracks, 29 train passages were analysed and both the
for all three train types, at all receivers, irrespective of distance individual vibration levels and averaged levels are plotted in
from the track. This was possibly because all trains were con- Fig. 27. The at-grade and embankment cases generated similar
structed by the same manufacturer and had similar weight and levels of vibration, with the embankment case generating slightly
suspension characteristics. For the embankment and cutting sites lower levels. On the other hand the cutting generated higher
the peak particle velocities associated with each train were very amplitude vibrations in all 3 component directions. This finding is
similar. Despite this, for the Eurostar passage at the at-grade site, consistent with historical French records which suggest that
the PPV values were slightly lower. This was attributed to the fact cuttings cause more ground vibration problems in comparison to
that only one Eurostar passage was recorded at the site and embankments [49]. Despite this, it is in contrast to the empirical
therefore the curve fitting approach was more susceptible to skew relationships presented in [22] which suggests that a cutting “may
(e.g. a train with a low number of passengers). reduce the vibration levels slightly”.
108 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Fig. 13. Eurostar configuration.

Fig. 14. Thalys vibration time histories left: Thalys single, right: Thalys double.

Table 3
TGV properties.

TGV

Driving Passenger cars

Car body mass (kg) 55,790 24,000

Bogie mass (kg) 2,380 3,040
Wheelset mass (kg) 2,048 2,003
Primary suspension stiffness (MN/m) 2.45 1.4
Primary suspension damping (kNs/m) 20 120
Secondary suspension stiffness (MN/m) 2.45 0.45
Secondary suspension damping (kNs/m) 40 40

Table 4
Thalys properties.

Thalys

Bogie Y230A (driving car) Bogie Y237A (lateral car) Bogie Y237B (passenger car)

Car body mass (kg) 53,442 34,676 14,250

Bogie mass (kg) 3,261 8,156 1,400
Wheelset mass (kg) 2,009 2009 2,050
Primary suspension stiffness (MN/m) 2.09 2.09 1.63
Primary suspension damping (kNs/m) 40 40 40
Secondary suspension stiffness (MN/m) 2.45 2.45 0.93
Secondary suspension damping (kNs/m) 40 40 40

3.6. Near vs far tracks generated the highest. It is observed that the average vibration
levels were similar for train passages on both the near and far
Fig. 28 compares average PPV vertical vibration levels for all tracks, however it was clear that for all three tracks, in the near
trains passing on either the near or far tracks. The PPV levels from field (o15 m from the track) the far track vibration was lower
the far track were normalised by adding a 4.5 m offset to the than the near track vibration. The opposite was true as the
receiver distances. This enabled a direct comparison between all distance from the track increased, with the far track showing
29 train passages. elevated average PPV levels in comparison to the near track. This
In a similar manner to Fig. 27, it was found that the embank- effect was particularly evident for the cutting earthworks profile.
ment case generated the lowest PPV levels while the cutting case The cause of this rise in PPV was unknown.
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 109

Table 5
Eurostar properties.

Eurostar

Bogie Y230A (driving car) Bogie Y237A (lateral car) Bogie Y237B (passenger car)

Car body mass (kg) 54,166 33,854 27,083

Bogie mass (kg) 3,075 9,440 2,360
Wheelset mass (kg) 2,046 2,046 2,046
Primary suspension stiffness (MN/m) 2.63 2.20 2.07
Primary suspension damping (kNs/m) 12 12 12
Secondary suspension stiffness (MN/m) 3.26 0.91 0.61
Secondary suspension damping (kNs/m) 90 2 4

Fig. 15. Train, track and soil excitation mechanisms.

Fig. 16. MASW geophone configuration.

3.7. Far field vibration vs near field vibration 15 and 30 Hz, with additional pronounced peaks at 27–31 Hz.
In the far field the dominant frequency range was generally still
It is observed from Fig. 28 that vertical vibration levels decayed located between 15 and 30 Hz although much less pronounced.
with distance from the excitation. This was as expected and was A small resonant frequency at 8.8 Hz was visible in the near field
due to material and geometrical damping [5]. For the purpose of and was greatly magnified in the far field.
comparing near and far field vibration characteristics, Figs. 29-31 For the embankment case (Fig. 30) in the near field the
show how the normalised amplitude frequency content of vertical frequency range was much broader, and generally higher than
railway vibration varied from near to the far field (with 1/3 octave the far field, with a key resonant frequency appearing at 141 Hz.
band histograms). For the at-grade case (Fig. 29) in the near field The majority of near field frequency content was located below
the frequency of propagating waves was predominantly between this peak, and similarly to the at-grade case there was a large
110 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Damping ξ (−) Damping ξ (−)

0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1
0 0

5 5
Soil depth (m)

Soil depth (m)

10 10

Damping Damping
P wave P wave
S wave S wave
15 15
0 500 1000 1500 0 500 1000 1500
Wave velocity (m/s) Wave velocity (m/s)

Damping ξ (−)
0 0.02 0.04 0.06 0.08 0.1
0
P wave
S wave
Damping

5
Soil depth (m)

10

15
0 500 1000 1500
Wave velocity (m/s)

Fig. 17. Test site soil properties (left: at-grade, right: embankment, lower: cutting).

volume of waves propagating in the 15–30 Hz range. Additional were of higher frequency in comparison to at-grade tracks. The
zones of frequency content were also visible at 50–65 and 80– frequency content of cuttings was also higher than the at-grade
95 Hz. For the far field, a large percentage of this high frequency case but less so than the embankment case. It was also concluded
content had dissipated and the frequency content was located that the high frequency vibrations generated by the track were
between 5 and 30 Hz. The main peak at 141 Hz had disappeared damped rapidly as they propagated through the soil. This was
and three main peaks appeared at 8.6, 17.5 and 22 Hz. because the frequency characteristics of soil typically prevent the
These higher frequencies inside the embankment were in propagation of high frequency vibration (Fig. 15). Instead, only the
agreement with numerical results presented by Ref. [20]. It was lower frequency waves, partly due to their longer wavelengths
postulated by [20] that these frequencies were generated due to were able to propagate to larger distances.
the propagating waves reflecting off the edges of the embankment
structure and a proportion of them becoming trapped within the 3.8. Scattering due to abutments
embankment, in a similar manner to how guided waves behave.
Similar conclusions were made regarding the experimental results Fig. 32 compares the variation in vibration levels with increas-
in this study. ing distance from the track for both the abutment and non-
For the near field cutting case (Fig. 31) the frequency content abutment cases. At distances close to the track there was a large
also exhibited a greater spread in comparison to the at-grade case. discrepancy between the vibration levels, however as the distance
The first major zone of frequency content was between 17 and was increased to 35 m from the track, responses became similar.
35 Hz, followed by another peak at 52 Hz and another smaller Although it cannot not be proved, it is postulated that this ‘shadow
region of frequency content around 85 Hz. In comparison, a large zone’ occurred because the ground vibrations could not pass
percentage of the frequency content present in the near field was directly from the track into the ground due to the presence of
not visible in the far field results. The lower frequency content was the abutment. Instead the waves were forced to pass around the
bound in the region 8–35 Hz, with a significant eigenfrequency at abutment before reaching the receivers. This travel path was
17 Hz. A low amplitude region of high frequency content was also longer thus causing the waves to lose a greater percentage of
visible around 130 Hz. their energy due to damping.
It was concluded from the frequency results that the near field Fig. 33 shows the difference in frequency content between the
vibration levels generated due the presence of an embankment abutment and non-abutment cases. Although both responses were
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 111

Wavenumber (/rad)
3

0
0 10 20 30 40 50 60 70 80
Frequency (Hz)
5

Wavenumber (/rad)
3

0
0 10 20 30 40 50 60 70 80
Frequency (Hz)

4
Wavenumebr (/rad)

0
0 10 20 30 40 50 60 70 80
Frequency (Hz)
Fig. 18. Left: experimental dispersion curves, right: theoretical dispersion curves (black solid lines) vs experimental (grey circles), top: at-grade site, middle: embankment
site, bottom: cutting site.

similar, the frequency spectrum for the abutment case was wider 3.9. Discussion
and a greater number of peaks were present. This occurred due to
the complex wave scattering process induced by the abutment The results presented in this work are useful for environmental
dimensions. When the waves generated by train passed through consultants and modellers, railway constructors, railway opera-
the track they were scattered due to the complex geometry of the tors, real estate owners, asset managers and academic researchers
abutment, thus generating a wider frequency spectrum. (e.g. Universities and research institutes). They will allow for the
Fig. 19. Vertical mobility of site 1 for two different distances between source-receiver, left: 15 m, right: 20 m.

Fig. 20. Vertical shear strain variation with distance from track.

Fig. 21. The effect of train speed on PPV, left: near track, right, far track.
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 113

Fig. 22. The effect of train speed on VdB, left: near track, right, far track.

Fig. 23. Vertical vibration levels for various train types, top left: at-grade, top right: embankment, bottom left: cutting, bottom right: all sites.

validation of new and existing vibration prediction models and In particular, a key finding was that the repeatability of
provide interesting insights into vibration propagation character- experiments was lower than anticipated and trains travelling at
istics within different earthwork profiles. the same speed, on the same track generated variances in
114 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Fig. 24. Three component PPV levels for at-grade track, left: mean PPV for all trains on near track, right: individual Thalys passage at 299 km/h on near track.

Fig. 25. Three component PPV levels for embankment track, left: mean levels for all trains on far track, right: Individual Thalys passage at 299 km/h on near track.

Fig. 26. Three component PPV levels for cutting track, left: mean levels for all trains on near track, right: individual Thalys passage at 297 km/h on near track.
Fig. 27. Earthworks profile effects in the vertical direction, top: near track, bottom: far track.

Fig. 28. Vertical vibration levels generated due to near and far train passages.

Fig. 29. Frequency spectrum at at-grade site (left: near, right: far).
116 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

Fig. 30. Frequency spectrum at embankment (left: near, right: far).

Fig. 31. Frequency spectrum at cutting (left: near, right: far).

effective, it is important that there are trends and a degree of

repeatability between results. As expected, trends were found in
the experimental analysis performed in this work, thus justifying
the application of numerical modelling for railway vibration
problems. Despite this, although justified, due to ground vibration
uncertainty, in practice prediction is undertaken using a conser-
vative approach, thus highlighting future research needs in this
area to better quantify the expected levels of risk for individual
projects.
Another notable finding was that vertical vibrations were not
always the most dominant form of vibration. Instead, horizontal
vibrations were found to be just as important. Once again, this is
relevant for numerical modelling which typically relies on solely
predicting vertical vibrations.
Lastly, PPV levels were found to be similar for all train types.
This was likely because the characteristics of three train types
measured (Eurostar, TGV and Thalys) were similar (e.g. wheel
spacings, weights, configuration…etc.). Despite this, for alternative
metrics (rather than PPV) that account for the time duration of a
Fig. 32. Vertical vibration variation (abutment). signal (e.g. KBfmax—[19]) this may not have been the case. This is
likely to be because trains such as the Eurostar commonly have
vibration levels. This is pertinent for numerical modelling, which twice number of carriages as Thalys, thus generating a longer
irrespective of discipline, commonly has to deal with innately duration of vibration and possibly causing an increase in these
large scatter and uncertainty. For numerical modelling to be alternative metrics.
 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118 117

iv. The embankment site caused the generation of higher fre-

quency content in comparison to at-grade track. The cutting
also generated higher frequency content than at-grade sec-
tions, albeit less than embankments.
v. The higher frequency components generated on all tracks was
damped rapidly as the waves propagate through the soil. Lower
frequency components attenuated less quickly.
vi. Thalys, TGV and Eurostar trains generate similar levels of
ground vibration.

Acknowledgements

The authors wish to thank the railway operator Infrabel, and

especially M. Debruxelles and M. Demaret, for their support during
the experimental investigations, and all data provided related to
the train/track characteristics. Acknowledgement also goes to the
University of Edinburgh, the Université de Mons, Heriot Watt
University and the University of Porto for the support and
resources provided for the undertaking of this research. Addition-
ally, the funding provided by Engineering and Physical Sciences
Research Council UK (EP/H029397/1), Natural Environment
Research Council (UK) and “FCT—Fundação para a Ciência e
Tecnologia”, Portugal, (research project—PTDC/ECM/114505/2009),
is also greatly appreciated, without which, this research could not
have been undertaken. Lastly, thanks is given to M. Percival and
M. Scoubeau who provided vital experimental assistance. The data
associated with this research can be found via [9].

Appendix

See Tables 6 and 7.

Table 6
Fig. 33. Frequency spectrum comparison (top: no abutment, bottom: abutment). Soil wave speeds.

Site 1 Site 2 Site 3

h (m) Vs (m/s) Vp (m/s) h (m) Vs (m/s) Vp (m/s) h (m) Vs (m/s) Vp (m/

4. Conclusions s)

Field experiments were undertaken at 4 railway sites across 1.5 175 270 1.3 142 280 1.35 160 270
1 120 270 1.3 162 280 1.35 171 270
Belgium for the purpose of providing researchers with a freely
1.7 202 550 1.2 157 280 3.1 223 410
available dataset for modelling validation, and to provide new 2.5 300 550 2.85 280 520 3.1 260 410
insights into railway ground vibrations. The experiments consisted Inf. 450 900 2.85 330 520 Inf. 798 1460
of ground vibration monitoring to assess vibration levels due to Inf. 598 940
train passage, and MASW tests to determine the underlying soil
properties. MASW tests were used to determine S-wave and P-
wave velocities and the results were validated using refraction
analysis. Train passage data can be found in the download section Table 7
of: www.davidpconnolly.com/. Soil damping.
Analysis of the field results revealed that:
Site 1 Site 2 Site 3

i. Vertical component vibration levels from high speed trains Layer Damping Layer Damping Layer Damping
were of higher amplitude than horizontal vibration levels at thickness () thickness () thickness ()
locations close to the track. However, as distance increased, the (m) (m) (m)
horizontal vibration levels were similar in magnitude and in
0.8 0.105 1.3 0.074 1.35 0.0775
some cases were more dominant. 1.5 0.0742 2.5 0.07 1.35 0.07
ii. There was a large scatter between train speed and vibration 1.5 0.09 2.85 0.05 3.1 0.0309
level data. A low, but positive correlation between variables 1.6 0.08 2.85 0.0344 3.1 0.05
was tentatively proposed. 1.5 0.07 Inf. 0.02 Inf. 0.03
5 0.04
iii. The cutting site generated elevated levels of ground vibration 0.01
in comparison to at-grade and fill embankment track sections.
118 D.P. Connolly et al. / Soil Dynamics and Earthquake Engineering 67 (2014) 102–118

References Earthquake Eng 2010;30(3):82–97. http://dx.doi.org/10.1016/j.soildyn.2009.

09.004.
[28] Hardin B. The nature of damping in sands. J Soil Mech Found Div 1965;91:
[1] Andersen L, Jones CJC. Coupled boundary and finite element analysis of
63–97.
vibration from railway tunnels—a comparison of two- and three-
[29] Hussein MFM, Hunt HEM. A numerical model for calculating vibration due to a
dimensional models. J Sound Vib 2006;293(3–5):611–25. http://dx.doi.org/
harmonic moving load on a floating-slab track with discontinuous slabs in an
10.1016/j.jsv.2005.08.044.
underground railway tunnel. J Sound Vib 2009;321(1–2):363–74. http://dx.
[2] Auersch L. Wave propagation in layered soils: theoretical solution in wave-
doi.org/10.1016/j.jsv.2008.09.023.
number domain and experimental results of hammer and railway traffic
[30] Hussein M, Hunt HEM, Kuo K, Costa PA, Barbosa J. The use of sub-modelling
excitation. J Sound Vib 1994;173(2):233–64.
[3] Auersch L. Train induced ground vibrations: different amplitude-speed rela- technique to calculate vibration in buildings from underground railways. Proc
tions for two layered soils. Proc Inst Mech Eng Part F J Rail Rapid Transit Inst Mech Eng Part F J Rail Rapid Transit 2013. http://dx.doi.org/10.1177/
2012;226(5):469–88. http://dx.doi.org/10.1177/0954409712437305. 0954409713511449.
[4] Badsar Sa, Schevenels M, Haegeman W, Degrande G. Determination of the [31] Kogut, J, Degrande, G, Haegeman, W, Karl, L. Free field vibration due to
material damping ratio in the soil from SASW tests using the half-power the passage of an IC train and a Thalys HST on the high speed track L2
bandwidth method. Geophys J Int 2010;182(3):1493–508. http://dx.doi.org/ Brussels-Koln. IABSE symposium report; 2002. http://dx.doi.org/10.2749/
10.1111/j.1365-246X.2010.04690.x. 222137803796330176.
[5] Barkan M. Dynamics of bases and foundations. New York, US: McGraw-Hill [32] Kouroussis G. Mesures de vibrations induites sur l’environnement par le
Inc.; 1962; 1–496. passage de trains à grande vitesse. Faculte Polytechnique De Mons 2005:1–84
[6] Campos J, de Rus G. Some stylized facts about high-speed rail: a review of HSR (Technical report).
experiences around the world. Transp Policy 2009;16(1):19–28. http://dx.doi. [33] Kouroussis, G, Connolly, DP, Forde, MC, Verlinden, O. An experimental study of
org/10.1016/j.tranpol.2009.02.008. embankment conditions on high-speed railway ground vibrations. In: Inter-
[7] Connolly D, Giannopoulos A, Fan W, Woodward PK, Forde MC. Optimising low national congress on sound and vibration (ICSV20). Bangkok, Thailand; 2013.
acoustic impedance back-fill material wave barrier dimensions to shield p. 1–8.
structures from ground borne high speed rail vibrations. Constr Build Mater [34] Kouroussis G, Connolly DP, Forde MC, Verlinden O. Train speed calculation
2013;44:557–64. http://dx.doi.org/10.1016/j.conbuildmat.2013.03.034. using ground vibrations. Proc Inst Mech Eng Part F J Rail Rapid Transit 2014;0
[8] Connolly D, Giannopoulos A, Forde MC. Numerical modelling of ground borne (0):1–18. http://dx.doi.org/10.1177/0954409713515649.
vibrations from high speed rail lines on embankments. Soil Dyn Earthquake [35] Kouroussis G, Connolly DP, Verlinden O. Railway induced ground vibrations—a
Eng 2013;46:13–9. http://dx.doi.org/10.1016/j.soildyn.2012.12.003. review of vehicle effects. Int J Rail Transp 2014;2(2):69–110. http://dx.doi.org/
[9] Connolly, DP. High speed rail testing—experimental vibration data (davidp- 10.1080/23248378.2014.897791.
connolly.com). 〈http://www.davidpconnolly.com/〉; 2014. [36] Kouroussis G, Verlinden O, Conti C. Free field vibrations caused by high-speed
[10] Connolly, DP, Kouroussis, G, Fan, W, Percival, M, Giannopoulos, A, Woodward, lines: measurement and time domain simulation. Soil Dyn Earthquake Eng
PK,et al. An experimental analysis of embankment vibrations due to high
2011;31(4):692–707. http://dx.doi.org/10.1016/j.soildyn.2010.11.012.
speed rail. In: Proceedings of the 12th international railway engineering
[37] Krylov V. Generation of ground vibrations by superfast trains. Appl Acoust
conference; 2013.
1995;44(2):149–64. http://dx.doi.org/10.1016/0003-682X(95)91370-I.
[11] Connolly DP, Kouroussis G, Giannopoulos A, Verlinden O, Woodward PK, Forde
[38] Lai C. Simultaneous measurement and inversion of surface wave dispersion
MC. Assessment of railway vibrations using an efficient scoping model. Soil
and attenuation curves. Soil Dyn Earthquake Eng 2002;22(9–12):923–30.
Dyn Earthquake Eng 2014;58:37–47. http://dx.doi.org/10.1016/j.soildyn.
http://dx.doi.org/10.1016/S0267-7261(02)00116-1.
2013.12.003.
[39] Ling X-Z, Chen S-J, Zhu Z-Y, Zhang F, Wang L-N, Zou Z-Y. Field monitoring on
[12] Connolly DP, Kouroussis G, Laghrouche O, Ho CL, Forde MC. Benchmarking
railway vibrations—track, vehicle, ground and building effects. Constr Build the train-induced vibration response of track structure in the Beiluhe
Mater 2014. http://dx.doi.org/10.1016/j.conbuildmat.2014.07.042. permafrost region along Qinghai–Tibet railway in China. Cold Reg Sci Technol
[13] Connolly DP, Kouroussis G, Woodward PK, Giannopoulos A, Verlinden O, Forde 2010;60(1):75–83. http://dx.doi.org/10.1016/j.coldregions.2009.08.005.
MC. Scoping prediction of re-radiated ground-borne noise and vibration near [40] Lombaert G, Degrande G. Ground-borne vibration due to static and dynamic
high speed rail lines with variable soils. Soil Dyn Earthquake Eng 2014;66 axle loads of InterCity and high-speed trains. J Sound Vib 2009;319:1036–66.
(2014):78–88. http://dx.doi.org/10.1016/j.soildyn.2014.06.021. http://dx.doi.org/10.1016/j.jsv.2008.07.003.
[14] Costa PA, Calcada R, Cardoso A. Vibrations induced by railway trafic: predic- [41] Lombaert G, Degrande G, Kogut J, Francois S. The experimental validation of a
tion, measurement and mitigation. Traffic induced environmental vibrations numerical model for the prediction of railway induced vibrations. J Sound Vib
and controls: theory and application 2013:49–89. 2006;297(3–5):512–35. http://dx.doi.org/10.1016/j.jsv.2006.03.048.
[15] Costa PA, Calcada R, Cardoso AS. Influence of train dynamic modelling strategy [42] Lopes P, Costa PA, Ferraz M, Calcada R, Cardoso AS. Numerical modelling of
on the prediction of track-ground vibrations induced by railway traffic. Proc vibrations induced by railway traffic in tunnels: from the source to the nearby
Inst Mech Eng Part F J Rail Rapid Transit 2012;226(4):434–50. http://dx.doi. buildings. Soil Dyn Earthquake Eng 2014;61–62:269–85. http://dx.doi.org/
org/10.1177/0954409711433686. 10.1016/j.soildyn.2014.02.013.
[16] Costa PA, Calcada R, Cardoso AS. Track–ground vibrations induced by railway [43] Madshus C. High-speed railway lines on soft ground: dynamic behaviour at
traffic: in-situ measurements and validation of a 2.5D FEM-BEM model. Soil critical train speed. J Sound Vib 2000;231(3):689–701. http://dx.doi.org/
Dyn Earthquake Eng 2012;32(1):111–28. 10.1006/jsvi.1999.2647.
[17] Costa PA, Calcada R, Cardoso AS, Bodare A. Influence of soil non-linearity on [44] Park Seismic. . Multichannel analysis of surface waves. Retrieved from www.
the dynamic response of high-speed railway tracks. Soil Dyn Earthquake Eng masw.com; 2013.
2010;30:221–35. http://dx.doi.org/10.1016/j.soildyn.2009.11.002. [45] Petyt M, Thompson DJ, Jones CJC. A model for ground vibration from railway
[18] Degrande G, Schillemans L. Free field vibrations during the passage of a Thalys tunnels. Proc Inst Civ Eng Transp 2002;153(2):121–9. http://dx.doi.org/
high-speed train at variable speed. J Sound Vib 2001;247(1):131–44. http://dx. 10.1680/tran.2002.153.2.121.
doi.org/10.1006/jsvi.2001.3718. [46] Rix G, Lai C, Wesley-Spang A. In situ measurement of damping ratio using
[19] Deutsches Institut fur Normung. DIN 4150-2—human exposure to vibration in surface waves. J Geotech Geoenviron Eng 2000;126(5):472–80.
buildings. International Standards Organisation 1999:1–63. [47] Sheng X, Jones CJC, Petyt M. Ground vibration generated by a harmonic load
[20] Ditzel A, Herman G. The influence of a rail embankment on the vibrations
acting on a railway track. J Sound Vib 1999;225(1):3–28. http://dx.doi.org/
generated by moving trains. J Sound Vib 2004;271:937–57. http://dx.doi.org/
10.1006/jsvi.1999.2232.
10.1016/S0022-460X(03)00772-7.
[48] Stiebel D. Protocol for free field measurement of mitigation effects (RIVAS—
[21] El Kacimi A, Woodward PK, Laghrouche O, Medero G. Time domain 3D finite
Deliverable D1.2) 2011:1–25.
element modelling of train-induced vibration at high speed. Comput Struct
[49] Stiebel D, Muller R, Bongini E, Ekbald A, Coquel G, Alguacil AA. Definition of
2013;118:66–73. http://dx.doi.org/10.1016/j.compstruc.2012.07.011.
reference cases typical for hot-spots in Europe with existing vibration
[22] Federal Railroad Administration. High-speed ground transportation noise and
vibration impact assessment. U.S. Department of Transportation 2012:1–248. problems (RIVAS—Deliverable D1.5) 2012:1–16.
[23] Ferrara, R, Leonardi, G, Jourdan, F. A two-dimensional numerical model to [50] Thompson D. Railway noise and vibration. Mechanisms, modelling and means
analyze the critical velocity of high speed infrastructure the 2D numerical of control. Oxford, UK: Elsevier Ltd.; 2009; 1–532.
model. In: Proceedings of the 14th international conference on civil, structural [51] Varandas JN, Hölscher P, Silva MaG. Dynamic behaviour of railway tracks on
and environmental engineering computing; 2013. p. 1–16. transitions zones. Comput Struct 2011;89(13–14):1468–79. http://dx.doi.org/
[24] Fryba L. Vibration of solids and structures under moving loads. Groningen: 10.1016/j.compstruc.2011.02.013.
The Netherlands: Noordhoff International Publishing; 1972; 1–524. [52] Wang J, Jin X, Cao Y. High speed Maglev train-guideway–tunnel–soil model-
[25] Galvin P, Domínguez J. Experimental and numerical analyses of vibrations ling of ground vibration. J Rail Rapid Transit 2011. http://dx.doi.org/10.1177/
induced by high-speed trains on the Córdoba–Málaga line. Soil Dyn Earthquake 0954409711424823.
Eng 2009;29(4):641–57. http://dx.doi.org/10.1016/j.soildyn.2008.07.001. [53] Wathelet M. An improved neighborhood algorithm: parameter conditions and
[26] Galvin P, Romero A, Domínguez J. Fully three-dimensional analysis of high- dynamic scaling. Geophys Res Lett 2008;35(9):1–5. http://dx.doi.org/10.1029/
speed train–track–soil-structure dynamic interaction. J Sound Vib 2010;329 2008GL033256.
(24):5147–63. http://dx.doi.org/10.1016/j.jsv.2010.06.016. [54] Wathelet, M. www.Geopsy.org; 2008b.
[27] Gupta S, Van den Berghe H, Lombaert G, Degrande G. Numerical modelling of [55] Wood DM. Soil behaviour and critical state soil mechanics. Cambridge, UK:
vibrations from a Thalys high speed train in the Groene Hart tunnel. Soil Dyn Cambridge University Press; 1990; 462.