Professional Documents
Culture Documents
Soil Dynamics and Earthquake Engineering: Alexandros Lyratzakis, Yiannis Tsompanakis, Prodromos N. Psarropoulos
Soil Dynamics and Earthquake Engineering: Alexandros Lyratzakis, Yiannis Tsompanakis, Prodromos N. Psarropoulos
A R T I C L E I N F O A B S T R A C T
Keywords: Ground-borne vibrations due to high-speed trains passage strongly depend, apart from the speed of the train, on
High-speed trains the geometry of the railways as well as the properties of the underlying soil layer(s). The main aim of this study is
Train-induced vibrations to investigate the effectiveness of expanded polystyrene (EPS) blocks in mitigating soil vibrations induced on
Railway embankments
railway embankments for different subsoil and railway embankment material conditions. The EPS blocks are
Mitigation measures
Expanded polystyrene (EPS)
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.
* Corresponding author.
E-mail addresses: alyratzakis@gmail.com (A. Lyratzakis), jt@science.tuc.gr (Y. Tsompanakis), prod@central.ntua.gr (P.N. Psarropoulos).
https://doi.org/10.1016/j.soildyn.2020.106482
Received 10 March 2020; Received in revised form 29 September 2020; Accepted 21 October 2020
0267-7261/© 2020 Elsevier Ltd. All rights reserved.
Please cite this article as: Alexandros Lyratzakis, Soil Dynamics and Earthquake Engineering, https://doi.org/10.1016/j.soildyn.2020.106482
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
2
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Table 2
Parameters of the examined four soil types.
Soil type Density Young’s Poisson’s Damping ξ
(kg/m3) Modulus ratio (-) (%)
Es (MPa)
Rail – 210 0.25 7900 As it has been earlier mentioned, several mitigation techniques have
Sleepers – 30 0.4 2400 been investigated in order to reduce the developed vibrations due to
Ballast 0.3 0.1 0.35 1800 HST passage. In the present study, the focus is given on the application
Subballast 0.2 0.3 0.35 2200 of EPS blocks in three different configurations (Models 2, 3 and 4, as
Subgrade 0.5 0.127 0.35 2100
described in Fig. 4), which are compared with the initial railway
3
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 3. Thalys train axle loads (adopted from Refs. [43]) and bogies dimensions (adopted from Ref. [44]).
Fig. 4. HSR embankment cross-section: (a) Model 1 refers to the soil embankment, (b) Model 2 refers to the mitigated embankment with EPS blocks at the slopes, (c)
Model 3 refers to the mitigated embankment with an EPS-filled trench, (d) Model 4 refers to the embankment with both mitigation measures.
embankment (Model 1). number of EPS46 blocks have been placed at the sides of the railway
Over the last decades EPS has been applied in many engineering embankment that were covered with a very thin layer of soil. The
applications, including normal speed railway embankments [47,48]. optimal configuration of EPS blocks has been investigated by the authors
Due to its low weight, geofoam fill can be easily implemented without in recent papers [35,36] Accordingly, a few EPS blocks (with dimensions
the use of large mechanical equipment at zones with difficult access, equal to 1 m) have been placed as illustrated in Fig. 4b. The efficiency of
reducing the construction time and cost. Apart from its practicality and this mitigation measure has been investigated by the authors for a real
ease of use, it should also be mentioned that EPS geofoam is a relatively embankment that is founded on silty soil layers [35]. Note that alter
low-cost material. Nevertheless, the cost of EPS material varies, native EPS configurations, i.e., to use an EPS embankment (fully or
depending on material type, dimensions of the blocks, required quan partially made by EPS blocks) had been examined [35]. Although higher
tities, transportation costs, etc. In addition, covering soil can be placed reductions on HST induced vibrations were achieved, these EPS schemes
directly on the geofoam blocks, eliminating the need for expensive and are not suggested, since apart from their increased cost, they also
time-consuming forming and casting. Lastly, if it is well-constructed and resulted in higher displacements that can cause derailments, since HST
protected EPS has durability and low maintenance cost. are much more sensitive to settlements compared to normal speed
A dense EPS material, i.e., EPS46, has been used, which is the stiffer trains. This cost-efficient technique has managed to reduce substantially
material among the seven types of EPS described in ASTM D6817 [49]), the vibration levels by altering the wave propagation within the em
since it was proven more efficient compared to softer EPS materials for bankment’s body due to the contrast between the properties of the two
mitigating HST vibrations [35]. EPS46 is a low-density material (45.7 materials (soil and EPS) which leads to high values of reflection coeffi
kg/m3) with relatively high Young’s Modulus (12800 kPa) and cient (RC), which strongly affects the vibration patterns. This is analo
extremely low Poisson ratio (0.05). gous with the discussion in Section 5 related to the wave propagation
According to the first mitigation scenario (Model 2), a limited under the embankment. The placement of the EPS blocks at the slopes
4
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
actually creates a waveguide, which leads the waves to a certain path 3.1. Soft clay
and does not allow them to be freely transmitted along the slopes and
further. In addition, the EPS blocks are slightly buried under 0.2 m of A soft soil, classified as soft clay, is the first scenario that has been
soil, hence, this soil cover may play a role in wave propagation at the examined in the present study. The implementation of a limited number
surface similar to the case of a buried trench [14]. of EPS blocks (Model 2) in such soils plays a very beneficial role in the
Secondly, the use of a trench with EPS blocks as filling material has mitigation of the increased HST vibrations. More specifically, the
also been examined (Model 3). As presented in the introductory section, maximum vertical velocity has been decreased from 0.8 mm/s to 0.35
trenches are the most commonly used mitigation measure for HST mm/s at the near-field location (at 15 m). The same trend is observed for
induced vibrations. According to several authors (e.g. Ref. [50]), the Model 3, where the maximum vertical velocity has been decreased from
most effective type of trench, is the open-air trench, i.e., without any 0.8 mm/s to 0.5 mm/s (the reductions is less than the corresponding one
filling material. On the other hand, problems related to stabilization and for Model 2). Lastly, as it can be noticed from Fig. 5d, the decrease of the
maintenance of open-air trenches, has led to the application of several vibrations levels is even higher in the case of Model 4.
filling materials, as discussed in the literature review. In the present The most critical vibrations are the vibrations in the low-frequency
study, the use of EPS46 blocks as trench filling material has also been range [2]. Accordingly, Fig. 6 presents the impact of the examined
examined. After a preliminary investigation regarding the optimal di mitigation measures on the velocity spectra at the low frequency range
mensions and location, a single trench has been placed at 14 m from the (0–60 Hz). In the case of Model 1, three frequencies (21.4 Hz, 25.4 Hz
middle of the railway embankment, having 50 cm width and 3 m depth and 29.3 Hz) can be spotted in Fig. 6a. The plot in Fig. 6b displays the
(see Fig. 4c). Finally, it was assumed that using EPS blocks at the slopes improvement of the soil response in the whole low-frequency range in
of the embankment and as filling material in the trench could further the case of Model 2. For example, the peak of the most dominant fre
decrease the HST induced vibrations. Hence, the hybrid mitigation quency (29.3 Hz) is reduced from 0.2 mm/s to 0.0 mm/s. In contrast,
scenario, depicted in Fig. 4d (Model 4), has also been studied. Fig. 6c illustrates that the implementation of the EPS-filled trench
(Model 3) has led to a slight reduction of the vibrations’ peaks in the
3. Numerical results whole examined frequency range (0–40 Hz). It is obvious from Fig. 6d
that regarding near-field vibrations (i.e., at 15 m), Model 4 has led to a
In the sequence, a comparison between the conventional and the remarkable improvement of the soil response. The same observation can
three mitigated embankments is presented for the four hypothetical soil be drawn at 35 m, where the examined measures mitigate effectively the
scenarios in terms of velocity time-histories and Fourier spectra. Note HST vibrations.
that the scales in the plots are kept constant in order to illustrate that -as
expected-the HST vibrations are increasing as underlying soil gets softer.
The results are given for two characteristic locations, i.e., d = 15 m and 3.2. Stiff clay
d = 35 m from the track, to illustrate the impact of the proposed miti
gation measures both near-field and far-field from the track. These lo As shown in Fig. 7, the beneficial influence of the mitigation with
cations are often used in field measurements [2,45] and were also used EPS46 blocks is also significant for stiff clay. In this case the Young’s
in the validation of the applied numerical modelling approach [35]. modulus of the soil is 3.5 times higher than in the previous scenario. This
results in lower vertical velocities compared to soft clay, e.g., at 15 m the
Fig. 5. Comparison of soil and EPS-retrofitted embankments on soft clay in terms of vertical velocity time-histories: (a) Model 1, (b) Model 2, (c) Model 3, (d)
Model 4.
5
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 6. Comparison of soil and EPS-retrofitted embankments on soft clay in terms of vertical velocity Fourier spectra: (a) Model 1, (b) Model 2, (c) Model 3, (d)
Model 4.
Fig. 7. Comparison of soil and EPS-retrofitted embankments on stiff clay in terms of vertical velocity time-histories: (a) Model 1, (b) Model 2, (c) Model 3, (d)
Model 4.
maximum vertical velocity is equal to 0.4 mm/s (see Fig. 7a), almost half observations can be made at 35 m, in this case all the examined miti
than previously (see Fig. 5a). Fig. 7b depicts the significant vibrations gation measures contribute to about 50% reduction of the vertical ve
mitigation in the case of Model 2. More specifically, the maximum locity. For example, the maximum vertical velocity is decreased from
vertical velocity is close to 0.19 mm/s, 50% lower than for Model 1. 0.11 mm/s to 0.06 mm/s after the implementation of the EPS-filled
Similarly, the vertical velocities have also been reduced in Model 3, in trench (Model 2).
which the maximum vertical velocity is 0.18 mm/s at 15 m. Τhe same Fig. 8 displays the comparison of the Fourier spectra for the four
6
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 8. Comparison of soil and EPS-retrofitted embankments on stiff clay in terms of vertical velocity Fourier spectra: (a) Model 1, (b) Model 2, (c) Model 3, (d)
Model 4.
examined models on stiff clay. As it is clearly illustrated, the most critical while Model 2 reduces significantly the vibrations in the lower fre
frequencies remain the same as in the case of soft clay, while the peak quency range (0–40 Hz). Fig. 8d illustrates that Model 4 leads to the
values are decreased. For instance, the peak at the frequency of 25.4 Hz maximum mitigation of the vibrations. Similarly, at 35 m, the proposed
is equal to 0.12 mm/s at 15 m, about 50% the peak at the same fre mitigation measures contribute to the spectral velocity reduction in the
quency for soft clay. The proposed measures are efficiently mitigating whole low frequency range.
the HST vibrations within the low frequency range. It can be noticed that
Model 3 is more effective in the frequency range from 40 Hz to 60 Hz,
Fig. 9. Comparison of soil and EPS-retrofitted embankments on dense sand with gravels in terms of vertical velocity time-histories: (a) Model 1, (b) Model 2, (c)
Model 3, (d) Model 4.
7
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
3.3. Dense sand with gravels rock and its properties are shown in Table 2. In this case, the vibrations
induced by Thalys HST are limited to very low levels (i.e., less than 0.06
Fig. 9 illustrates the vertical velocity time-histories when the mm/s) and they are not causing any problems to the train and the
embankment is located on dense sand with gravels, where there is a embankment. Therefore, the application of any mitigation measure is
significant decrease on the vibrations compared to stiff clay, while the not required/recommended and the focus of the parametric study is
beneficial impact of EPS is not so pronounced as in the previous cases of concentrated on the other three cases.
softer soils. More specifically, the maximum vertical velocity of the soil
embankment model at 15 m is equal to 0.2 mm/s. Due to the higher 4. Mitigation measures efficiency
weight of the first and the last locomotive bogies, higher near-field soil
response has been observed when these carriages are passing. As it is The efficiency of the examined mitigation measures, i.e., Models 2, 3
clearly depicted in Fig. 9b, Model 2 has slightly reduced the vibrations and 4, which are compared with the reference case of the initial railway
induced by Thalys HST at both examined locations. Furthermore, Model embankment (Model 1), is summarized in Fig. 11. More specifically, the
3 results in a higher decrease of the developed vibrations. Lastly, Model attenuation in terms of peak particle velocity (PPV) for dense sand with
4 leads to the optimum mitigation of the vibrations, since the maximum gravels, stiff and soft clay at 15 m is illustrated in this chart. In all soil
vertical velocity is less than 0.16 mm/s (see Fig. 9d). types the proposed mitigation measures have led to the reduction of the
Fig. 10 presents the comparison of the Fourier spectra for the four vibrations induced by Thalys HST passage. It is clearly depicted that the
embankments on dense sand with gravels. The main vibrations peaks at most effective mitigation measure is Model 4, since the attenuation is
25.4 Hz and 29.3 Hz in the case of the soil embankment (Model 1) at 15 close to 65% for the soft soil case, where it is noteworthy to observe that
m are equal to 0.06 mm/s and 0.08 mm/s, respectively. Furthermore, the side-fill (Model 2) is much more efficient compared to EPS-filled
two lower peaks have been observed at 21.4 Hz and 33.8 Hz. Model 2 trench (Model 3).
has contributed to the reduction of the most dominant peak at 29.3 Hz Fig. 12 illustrates the effectiveness of the examined mitigation ap
from 0.08 mm/s to 0.03 mm/s. In addition, the vibrations peak at 25.4 proaches in terms of PPV, which is calculated as follows:
Hz has been reduced to 0.04 mm/s. On the other hand, the peak at 21.4 ⃒√⃒̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
⃒ ⃒
Hz has been slightly increased. Furthermore, in Model 3, the vibrations PPV = max⃒ v2x (t) + v2y (t) + v2z (t)⃒ (4)
peaks at 15 m are slightly reduced. Model 4 exhibits the higher reduction
of the vibrations peak at 29.3 Hz, although the reduction of the vibra where vx(t), vy(t) and vz(t) are the longitudinal, the transversal and the
tions level at the other vibrations peaks is marginal. Similar observations vertical velocity time-histories, while the vertical velocity is the domi
are made for the examined models at the far-field location (35 m). nant component of PPV. More specifically, the PPV in the modelling
scenario for dense sand with gravels without any mitigation measure
(Model 1) has been compared with the Model 4 of the mitigated
3.4. Rock embankment laid on soft clay. As it is clearly shown in the plot, the PPV
at varying distances (from 15 m to 35 m) in the case of soft clay after the
In the last soil scenario, the passage of Thalys HST over a very stiff combined EPS mitigation (Model 4) is comparable with the
soil layer has been examined. The examined soil has been classified as
Fig. 10. Comparison of soil and EPS-retrofitted embankments on dense sand with gravels in terms of vertical velocity Fourier spectra: (a) Model 1, (b) Model 2, (c)
Model 3, (d) Model 4.
8
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 12. Comparison of PPV for the soil embankment (Model 1) on dense sand with gravels with the retrofitted embankment (Model 4) on soft clay.
corresponding ones of the soil embankment (Model 1) that is located on associated unit is m/s. This indicator provides an alternative way of
dense sand with gravels. In other words, by adopting this cost-effective assessing the effect of HST induced vibrations. Vehicle vibratory
intervention, the geotechnical conditions are notably upgraded and the discomfort is then assessed by comparing the maximum level of KBF,max
HST can operate in a better and safer manner, even when it has to cross with a threshold limit for residential areas, which is set equal to 15
areas with soft soil layers. mm/s, while KBF,max is derived as a function of PPV [52]:
Since the vibration is often non-stationary, the German Institute for √̅̅̅
Standardization (DIN) proposed the implementation of a running root 2 PPV
KBF,max = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( )2̅ (7)
mean square to the velocity signal. In particular, according to DIN 4150- 2
5.6
1+ f
2 [51], a weighted time-averaged signal (KBF) is computed using the
following expression [2,50]:
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ Consequently, Fig. 13 depicts the maximum level of KBF at six
√ ∫τ
√1 equally spaced observation points (15 m–35 m) from the track. It is
√
(5) evident that KBF,max has been decreased with distance from the track. In
t− ξ
KBF (t) = √ KB2 (ξ)e− τ dξ
τ
O addition, the KBF,max levels have been reduced after the implementation
of all the examined mitigation measures for all soil scenarios. More
where the weighted velocity signal KB(t) is derived by flowing the initial specifically, as it is clearly illustrated in Fig. 13a, the hybrid mitigation
velocity signal via the high-pass filter [2,50]: technique is the most efficient one for all soil types.
As displayed in Fig. 13c, in the case of dense sand with gravels, the
1
HKB (f ) = √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
( )̅ (6) KBF,max of Model 1 is lower than the threshold value of 15 mm/s for
distances between 17 m and 35 m. Hence, it can be deduced that the
2
5.6
1+ f
implementation of any mitigation is not necessary for such soil condi
tions, although the implementation of all the examined mitigation
while the integration time, τ, is set equal to 0.125s [2,50]. It should be measures contributes to a substantial reduction of KBF,max. The benefi
mentioned that while no specific unit is defined in DIN standards, the cial effect of the proposed mitigation approaches is more evident in the
9
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 13. Comparison of soil and EPS-retrofitted embankments at increasing distance from the track in terms of KBF,max: (a) soft clay, (b) stiff clay, (c) dense sand
with gravels.
case of soil stiff clay, where initially KBF,max values are higher than the mitigation approaches in Fig. 14, The calculations have been carried out
DIN limit for Model 1 for distances between 15 m and 25 m. In contrast, at the centre frequency of each one-third octave from 1.25 to 40 Hz and
the implementation of all EPS-based mitigation types has substantially the results were averaged to one-third octave bands. The insertion loss
decreased KBF,max values below the limit of 15 mm/s for almost all the (IL) is computed using the following expression [53]:
examined distances.
Vrms,soil
Finally, as it was expected, the highest values of KBF,max have been IL = 20log 10 (8)
Vrms,0,mit
observed for soft soil soft clay, where KBF,max values are much higher (up
to 4 times at near-field distances) than the DIN limit for all locations. where:
Model 3 has reduced the values of KBF,max, but they still remain above vrms, soil: root mean square of the spectral velocity of each 1/3 octave
15 mm/s for most of the examined points (<29 m from the track). On the band of Model 1,
other hand, Model 2 (and subsequently Model 4) has managed to reduce vrms,mit: root mean square of the spectral velocity of each 1/3 octave
effectively the KBF,max levels in all distances to acceptable levels, apart band for each of the examined mitigation approaches shown in Fig. 4.
from near-field (15 m), where it remains slightly higher than the limit. As it is illustrated in the plots of Fig. 14, the insertion loss is higher
Furthermore, the insertion loss is illustrated for the three examined
10
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 14. Comparison of insertion loss at 15 m from the track for the three mitigation approaches for: (a) soft clay, (b) stiff clay, (c) dense sand with gravels.
for the softer soils for all the examined mitigation scenarios in the fre case of dense sand with gravels (Fig. 15c), the higher IL is again observed
quency range of interest (20–35 Hz). It should be noted that according to at the octave band with centre frequency 20 Hz.
Fourier spectral analysis, the dominant frequencies for all the examined
soil layers are located at the 1/3 octave bands with centre frequencies at 5. Impact of embankment fill material
25 Hz and 31.5 Hz, in which the insertion loss curve is always positive.
In particular, when the embankment is constructed on soft clay Initially the embankment has been constructed by the same silty soil
(Fig. 14a), Model 4 reduces the vibration levels up to 5 db for the whole -as described in Section 2.2- for all the examined soil scenarios in order
examined frequency range, while Model 2 and Model 3 are effective only to investigate the subsoil response for the same conditions. However, in
for frequencies higher than 10 Hz. When the embankment is located on engineering practice, the surface soil is commonly used as railway
stiff clay (Fig. 14b), Model 2 exhibits a better performance in most fre embankment fill material. Hence, the vibrations during the passage of
quencies. On the other hand, in the case of the stiffer soil (dense sand Thalys have been calculated when the subsoil (dense sand with gravels,
with gravels) the role of all mitigation measures is not so beneficial. It is stiff or soft clay) has been used as embankment fill material and the
evident from Fig. 14c that the addition of the trench (i.e., Model 3), corresponding models are presented in Table 3.
increases the vibration levels at the low octave bands from 1.25 Hz to 8 Fig. 16a illustrates the vertical velocity Fourier spectra of Model 1B.
Hz. In contrast, Model 4, reduces the vibration levels for all the exam In this case, the implementation of the hybrid mitigation (Model 4B) has
ined centre frequencies, except from 20 Hz, where the insertion loss is led to the reduction of the vibrations’ peaks within the whole frequency
negative both for Model 2 and Model 4. range (0–60 Hz). More specifically, the two dominant peaks at 29.3 Hz
It should be mentioned that in existing HSR lines, residential build and 33 Hz have been reduced from 0.13 mm/s and 0.1 mm/s to 0.06
ings are commonly located at distances greater than 15 m from the track, mm/s and 0.04 mm/s, respectively.
thus, it is reasonable to examine the IL curves at distances where The same observation can be made when the embankment has been
buildings are usually cited. Nonetheless, in the future upgrading of constructed with soil obtained from the underlying stiff clay (see
existing normal speed railways, which pass through urban environ Fig. 16b). The most dominant peak is at 29.3 Hz and it is decreased from
ments, the disturbance of adjacent buildings in much lower distances 0.18 mm/s (Model 1C) to 0.06 mm/s (Model 4C). Furthermore, all the
will become a common problem. Fig. 15 depicts the IL curves at 35 m
from the track, where the combined approach (Model 4) is clearly the
most effective mitigation scheme. More specifically, this scheme led to Table 3
the reduction of the vibrations level from 3 to 16 dB for all subsoil Models with different embankment soil material.
conditions within the whole examined frequency range. Furthermore, in Embankment fill Regular Embankment with hybrid EPS
the case of soft clay (Fig. 15a), the higher IL is observed at the octave material embankment mitigation
bands with centre frequencies higher than 20 Hz, which are the most Dense sand with Model 1B Model 4B
dominant ones according to the Fourier spectra (see Fig. 6). Similarly, gravels
Stiff clay Model 1C Model 4C
the IL at the octave band with centre frequency 20 Hz is higher than 15
Soft clay Model 1D Model 4D
dB when the embankment is laid on stiff clay (Fig. 15b). Lastly, in the
Fig. 15. Comparison of insertion loss at 35 m from the track for the three mitigation approaches for: (a) soft clay, (b) stiff clay, (c) dense sand with gravels.
11
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 16. Comparison of soil and EPS-retrofitted embankments in terms of vertical velocity Fourier spectra for: (a) dense sand with gravels, (b) stiff clay, (b) soft clay.
12
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
Fig. 18. Comparison of KBF,max for different embankent filling material at increasing distance from the track for: (a) dense sand with gravels, (b) stiff clay, (b)
soft clay.
where ρS is the density of the subsoil, vS,S is the shear wave velocity of back into the embankment. In other words, the wave energy is trapped
the subsoil, ρE is the density of the embankment fill material, and vS,E is within the geostructure body, i.e., the embankment acts as a waveguide
the shear wave velocity of the embankment. [3]. On the other hand, the value of RC is significantly lower for stiff
As presented in Table 4, when the silty embankment (i.e., Model 1, clay, as it is only 26%. Lastly, the RC in the case of soft clay is very low
with ρE = 200 kg/m3 and vS,E = 142 m/s) is founded on stiff soil, i.e., (8%) due to the close shear wave velocity of the silty embankment and
dense sand with gravels, then the interface has a high RC value equal to the underlying soil materials, thus, the HST vibrations are more easily
62%. This causes wave energy to be reflected from the subsoil surface propagated from the embankment to the surrounding soil.
13
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
14
A. Lyratzakis et al. Soil Dynamics and Earthquake Engineering xxx (xxxx) xxx
[25] Coulier P, François S, Degrande G, Lombaert G. Subgrade stiffening next to the [39] Feng SJ, Zhang XL, Zheng QT, Wang L. Simulation and mitigation analysis of
track as a wave impeding barrier for railway induced vibrations. Soil Dynam ground vibrations induced by high-speed train with three dimensional FEM. Soil
Earthq Eng 2013;48:119–31. https://doi.org/10.1016/j.soildyn.2012.12.009. Dynam Earthq Eng 2017;94:204–14. https://doi.org/10.1016/j.
[26] Coulier P, Cuéllar V, Degrande G, Lombaert G. Experimental and numerical soildyn.2017.01.022.
evaluation of the effectiveness of a stiff wave barrier in the soil. Soil Dynam Earthq [40] ABAQUS. Analysis user’s manual Ver, 6.14. Providence, RI, USA: Simulia Corp.;
Eng 2015;77:238–53. https://doi.org/10.1016/j.soildyn.2015.04.007. 2014.
[27] With C, Bahrekazemi M, Bodare A. Wave barrier of lime–cement columns against [41] International union of Railways. UIC code 719R: earthworks and trackbed layers
train-induced ground-borne vibrations. Soil Dynam Earthq Eng 2009;29:1027–33. for railway lines. France: UIC. Paris; 1994.
https://doi.org/10.1016/j.soildyn.2008.12.005. [42] Nakamura N. Time history response analysis using extended Rayleigh damping
[28] Andersen L, Nielsen SRK. Reduction of ground vibration by means of barriers or model. Procedia Eng 2017;199:1472–7. https://doi.org/10.1016/j.
soil improvement along a railway track. Soil Dynam Earthq Eng 2005;25:701–16. proeng.2017.09.408.
https://doi.org/10.1016/j.soildyn.2005.04.007. [43] Kouroussis G, Verlinden O. Prediction of railway induced ground vibration through
[29] Pflanz G, Hashimoto K, Chouw N. Reduction of structural vibrations induced by a multibody and finite element modelling. Mech Sci 2013;4:167–83. https://doi.org/
moving load. J Appl Mech 2002;5:555–63. https://doi.org/10.2208/ 10.5194/ms-4-167-2013.
journalam.5.555. [44] Degrande G, Schillemans L. Free field vibrations during the passage of a thalys
[30] Dijckmans A, Ekblad A, Smekal A, Degrande G, Lombaert G. Efficacy of a sheet pile high-speed train at variable speed. J Sound Vib 2001;247:131–44. https://doi.org/
wall as a wave barrier for railway induced ground vibration. Soil Dynam Earthq 10.1006/jsvi.2001.3718.
Eng 2016;84:55–69. https://doi.org/10.1016/j.soildyn.2016.02.001. [45] Connolly DP, Kouroussis G, Woodward PK, Alves Costa P, Verlinden O, Forde MC.
[31] Thompson DJ, Jiang J, Toward MGR, Hussein MFM, Dijckmans A, Coulier P, et al. Field testing and analysis of high speed rail vibrations. Soil Dynam Earthq Eng
Mitigation of railway-induced vibration by using subgrade stiffening. Soil Dynam 2014;67:102–18. https://doi.org/10.1016/j.soildyn.2014.08.013.
Earthq Eng 2015;79:89–103. https://doi.org/10.1016/j.soildyn.2015.09.005. [46] Kouroussis G, Connolly DP, Olivier B, Laghrouche O, Costa PA. Railway cuttings
[32] Gao G, Li N, Gu X. Field experiment and numerical study on active vibration and embankments: experimental and numerical studies of ground vibration. Sci
isolation by horizontal blocks in layered ground under vertical loading. Soil Dynam Total Environ 2016;557–558:110–22. https://doi.org/10.1016/j.
Earthq Eng 2015;69:251–61. https://doi.org/10.1016/j.soildyn.2014.11.006. scitotenv.2016.03.016.
[33] Antes H, von Estorff O. Dynamic response of 2D and 3D block foundations on a [47] Neupane R. Expanded polystyrene geofoam embankment for support of railways
halfspace with inclusions. Soil Dynam Earthq Eng 1994;13:305–11. https://doi. and bridges. PhD Thesis. Utah: The University of Utah; 2015.
org/10.1016/0267-7261(94)90022-1. [48] Li S. Evaluation and numerical modeling of deflections and vertical displacement
[34] Takemiya H. Field vibration mitigation by honeycomb WIB for pile foundations of of rail systems supported by EPS geofoam embankments. Utah: MSc Thesis, The
a high-speed train viaduct. Soil Dynam Earthq Eng 2004;24:69–87. https://doi. University of Utah; 2014.
org/10.1016/j.soildyn.2003.07.005. [49] ASTM D6817/D6817M-17. Standard specification for rigid cellular polystyrene
[35] Lyratzakis A, Tsompanakis Y, Psarropoulos PN. Efficient mitigation of high-speed geofoam. West Conshohocken, PA: ASTM International; 2017. www.astm.org.
trains induced vibrations of railway embankments using expanded polystyrene [50] Yang Y-B, Hung H-H. A parametric study of wave barriers for reduction of train-
blocks. Transp Geotech 2020;22:100312. https://doi.org/10.1016/j. induced vibrations. Int J Numer Methods Eng 1997;40:3729–47. https://doi.org/
trgeo.2019.100312. 10.1002/(SICI)1097-0207(19971030)40:20<3729::AID-NME236>3.0.CO;2-8.
[36] Lyratzakis A, Tsompanakis Y, Psarropoulos PN. Mitigation of HSΤ-induced [51] Deutsches Institut fur Normung. Din 4150-2: structural vibrations – Part 2: Human
vibrations by EPS blocks applied in railway embankments. Invited paper for special exposure to vibration in buildings, Germany. 1999.
issue on: noise and vibration from transportation. J Zhejiang Univ Sci A 2020. [52] Kouroussis G, Conti C, Verlinden O. Building vibrations induced by human
https://doi.org/10.1631/jzus.A1900680. activities: a benchmark of existing standards. Mec Ind 2014;15:345–53. https://
[37] Jiang J, Toward MGR, Dijckmans A, Thompson DJ, Degrande G, Lombaert G, doi.org/10.1051/meca/2014041.
Hussein MFM. The influence of soil conditions on railway induced ground-borne [53] Coulier P, Dijckmans A, Jiang J, Thompson DJ, Degrande G, Lombaert G. Stiff wave
vibration and relevant mitigation measures. In: The 21st International Congress on barriers for the mitigation of railway induced vibrations. Noise and Vibration
Sound and Vibration, Beijing, China, 13-17 July; 2014. Mitigation for Rail Transportation Systems. In: Nielsen J, et al., editors. Stiff wave
[38] Thompson DJ, Jiang J, Toward MGR, Hussein MFM, Ntotsios E, Dijckmans A, et al. barriers for the mitigation of railway induced vibrations, 126. Berlin, Heidelberg:
Reducing railway-induced ground-borne vibration by using open trenches and soft- Springer; 2013. p. 539546.
filled barriers. Soil Dynam Earthq Eng 2016;88:45–59. https://doi.org/10.1016/j.
soildyn.2016.05.009.
15