Tunnelling and Underground Space Technology 43 (2014) 59–66

Contents lists available at ScienceDirect

Tunnelling and Underground Space Technology

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

Evaluation of influence of vibrations generated by blasting construction

on an existing tunnel in soft soils
Haitao Yu a,b,c, Yong Yuan d,⇑, Guangxi Yu e, Xian Liu a,b
a
Key Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University, Shanghai 200092, China
b
Department of Geotechnical Engineering, Tongji University, Shanghai 200092, China
c
Department of Civil Engineering, Shanghai Jiaotong University, Shanghai 200030, China
d
State Key Laboratory for Disaster Reduction in Civil Engineering, Tongji University, Shanghai 200092, China
e
Shanghai Pudong Project Development Co., Ltd, Shanghai 201210, China

a r t i c l e i n f o a b s t r a c t

Article history: The controlled blasting technique is employed, for the first time, on the diaphragm wall of an existing
Received 20 June 2013 road tunnel in soft soils in the city of Shanghai, and therefore, safety demands on evaluation of influence
Received in revised form 28 February 2014 of the blasting vibrations on the tunnel are highly required. Based on the practical blasting scheme of four
Accepted 4 April 2014
sections of diaphragm wall of the tunnel to be exploded in three groups, the field monitoring of blasting
Available online 9 May 2014
vibrations is carried out on the bottom plate of the tunnel and the blasting vibration effects on the tunnel
are measured throughout the blasting construction. The sequential characteristics of vibration velocities
Keywords:
and accelerations of the tunnel structure are obtained from the field blasting tests. In order to fully under-
Blasting construction
Tunnel structure
stand the safety criterion of the blasting vibrations on the tunnel, a 3D numerical model is established by
Field monitoring the finite element software ABAQUS. The numerical simulation results are confirmed by field monitoring
Numerical simulation data. Based on the monitoring data and numerical results of the tunnel subjected to blasting loads, results
Dynamic response show: (1) the blasting energy is transferred mainly along the vertical direction; (2) the basic frequency
domain of the tunnel responses caused by the explosion is relatively in a lower frequency range; (3) dis-
tinct vibration response spectra of the tunnel structure are observed along the horizontal and vertical
directions; (4) the performance of the tunnel subjected to the blasting construction is in a critical state
of safety and thus the blasting construction scheme should be implemented carefully. Finally, critical
issues on evaluation of the blasting construction are fully discussed and evaluated, which provide a ref-
erence for other similar projects.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction The influences of blasting vibration on underground structures

Tunnels constitute a major part of civil infrastructure and serve
as public transportation facilities, water passage, and other pur-
poses such as electricity or communication cable installation (Yu
et al., 2013a). With the rapid development and upgrade of infra-
structures in China, there is a rising number of blasting construc-
tion near or on existing tunnels (Jiang and Zhou, 2012). The
blasting vibration during operation can damage the neighboring
or existing tunnel structure and endanger the construction pro-
jects. Therefore, it is necessary to thoroughly evaluate the dynamic
responses of the tunnel subjected to blasting vibration to guaran-
tee the safety of the tunnel during construction.
⇑ Corresponding author. Address: Institute of Underground Engineering, Tongji
University, 1239 Siping Road, Shanghai 200092, China. Tel.: +86 21 65980234;
fax: +86 21 65984573.
E-mail address: yuany@tongji.edu.cn (Y. Yuan).
have been studied by a number of researchers using field experi-
ments (Ahmed and Ansell, 2011; Ansell, 2004; Nateghi, 2011;
Ozer, 2008) and numerical simulations (Feldgun et al., 2008; He
et al., 2011; Lu et al., 2011; Ma et al., 1998; Wu et al., 2004; Zhu
and Tang, 2006). In most of the studies, the blasting construction
is limited to rock blasting. The elements affecting rock blasting
are drilling length and diameter, over-burden, the explosive char-
acteristic, rock mass strength and characteristics. However, no
experimental or numerical study of the effects of blasting vibration
on underground structures in soil medium, especially in soft soils,
has been reported. The theoretical research on blasting vibration in
soft soils is vacant and the safety criterion of blasting vibration
velocity for adjacent tunnel is still undetermined.
In this paper, based on the blasting construction of the
diaphragm walls of an existing road tunnel in soft soils, the field
monitoring is required to obtain actual structural vibrations of

http://dx.doi.org/10.1016/j.tust.2014.04.005
0886-7798/Ó 2014 Elsevier Ltd. All rights reserved.
60 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66
the tunnel from the blasting operation. The vibration monitoring
program, including monitoring equipments, monitoring points,
vibration measurements and measurement techniques, is devel-
oped to fully satisfy requirements of the field monitoring. Accord-
ing to the monitoring data of the blasting vibrations, velocity
response, vibration attenuation, and acceleration response and its
dominant frequency of the tunnel structure are obtained and ana-
lyzed. A thorough understanding of the safety criterion of the
blasting vibrations for the tunnel is realized by establishing a 3D
numerical model using ABAQUS. The numerical simulation results
are compared with field monitoring data.
2. A brief description of the tunnel
The Fuxing East Road Tunnel is the first twin-line and two-layer
shield tunnel used for highway transportation in the city of Shang-
hai, China. The tunnel has a total length of 2785 m. Its layout can
be divided into three segments: cross-river segment, nearshore
buried segment and Puxi terminal segment. The tunnel liner of
the Puxi terminal segment is made of 600 mm thick concrete. It
is 13.95 m high, 19.5 m wide, and the shape of its cross-section is
rectangular. The elevation of the bottom plate of the Puxi terminal
segment is 8.641 m, and the diaphragm wall is adopted as the
retaining structure. Below the elevation of 13.675 m, the dia-
phragm wall made of C10 fiberglass concrete is reserved for the
subsequent shield tunneling of the other planned tunnel used for
rail transit. The C10 fiberglass concrete, i.e. fiberglass used as a
reinforcing agent for C10-grade concrete (the grade is consistent
with the Chinese code for design of concrete structures 2010), is
used instead of the traditional reinforced concrete, mainly due to
the fact that compared to the reinforcement the fiberglass is much
easier to be cut by the shield machine. The general plan and loca-
tion of both the Fuxing East Road Tunnel and the planned subway
tunnel are shown in Fig. 1. In order to guarantee shield tunneling
successfully and accelerate construction schedule, the blasting
construction scheme is employed on the diaphragm wall firstly,
and then followed by the penetration of shield tunneling through
the blasting zone. In the explosion, four sections of diaphragm wall
are expected to be exploded, including two sections for the uplink
and downlink of the planned tunnel respectively (Fig. 1). According
to the geological exploration data following the geological
exploration report of the Fuxing East Road Tunnel (2007), the soil
Fig. 1. Plane view of the Fuxing East Road Tunnel and the planned subway tunnel.
through which the planned subway tunnel is excavated is mainly
composed of the silty clay, as shown in Fig. 2.
3. Blasting construction and field monitoring
3.1. Blasting scheme and its implementation
In the blasting construction, four sections of diaphragm wall are
divided to be exploded in three groups, i.e. the first and second
blasting constructions for the east and west sections of diaphragm
wall close to the south-line tunnel, respectively, while the third
blasting construction for both sections of diaphragm wall close to
the north-line tunnel simultaneously, as shown in Fig. 1. In relation
to tunnel construction, Peck (1969) stated three issues, which are:
(1) maintaining stability and safety during construction; (2) mini-
mizing unfavorable impact on third parties; (3) performing the
intended function over the life of a project. Among the issues, the
first and second ones are directly related to the appropriate con-
struction of tunnels.
To maintain safety and minimize effects of blasting vibration on
the existing tunnel structure, the reasonable controlled blasting
technique is selected with consideration of the geological condi-
tions and the surrounding environment. The blasting scheme of
vertical drilling along the center line of the diaphragm wall is
adopted. The vertical drilling starts from the center of the top sur-
face of the diaphragm wall. The drilling depth and diameter are
27.2 m and 110 mm respectively, and the distance between each
two adjacent drilling holes is 0.8 m, which equals to the width of
the diaphragm wall. The vertical view of drilling holes in the dia-
phragm wall is shown in Fig. 3. The emulsion explosive equipped
with RDX detonator is employed in the explosion, and the millisec-
ond blasting technique is adopted as the detonation. The detailed
blasting scheme for each section of diaphragm wall is arranged
as followed: firstly, the No. 6 drilling hole (Fig. 3) in the middle
of each section of diaphragm wall is blasted; then the next explo-
sion is followed by the two adjacent drilling holes, i.e. Nos. 5 and 7
(Fig. 3); and the followed explosions spread continuously in the
same way and finally end with the drilling holes of Nos. 1 and
11. The time interval for each adjacent explosion of drilling holes
Fig. 2. Typical cross section of the Puxi terminal segment of the Fuxing East Road
Tunnel.
 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66
Fig. 3. Vertical view of drilling holes in the diaphragm wall.
is 0.1 s. The explosive charge for each drilling hole is chosen to be
15 kg and divided into five segments each 3 kg weight. In order to
further increase the fragmentation degree of concrete, the millisec-
ond blasting is also applied to charge segments of each drilling hole
from bottom to top, with the time interval of 0.1 s for each adjacent
explosion.
It is the first time for the construction technology of controlled
blasting to be employed in tunnel construction in Shanghai, and
the blasting construction site is located in the heavy-traffic road
intersection of Zhonghua Road and Fuxing East Road. Furthermore,
a large number of buildings and public pipelines can be found
around the construction site. Hence, safety demands on manage-
ment of the blasting construction are highly required. According
to the construction requirements, the implementation of the blast-
ing scheme is listed as follows:
 Expose the top surface of the diaphragm wall and calibrate
the design position of each drilling hole based on the posi-
tion measurement of the diaphragm wall.
 Drill holes using geotechnical drilling rigs, and monitor the
perpendicularity of drill rods using theodolites.
 Fill charges into drilling holes with the aid of locating
rods each 27.2 m long, i.e. fix the charges segmentally
at the design position and insulate each charge seg-
ment with sandbags; install detonators and then put
the assembled locating rods into the assigned drilling
holes each 27.2 m depth. This is done to prevent the
misplacement of the charges.
 Refill drilling holes with yellow sands and assign protective
coverings to ground holes, in order to prevent blasting frag-
ments ejecting from ground holes.
 Connect to the detonating network, the millisecond blast-
ing starts from the middle drilling hole (No. 6 in Fig. 3)
and spreads outwards continuously, and finally ends with
the drilling holes of Nos. 1 and 11.
 Set blasting alert lines around the construction site and
implement the blasting.
 Finish the whole blasting construction of the target dia-
phragm wall.
61
3.2. Monitoring method and procedure
The field monitoring of blasting vibrations is carried out on the
bottom plate of the Fuxing East Road Tunnel and the blasting
vibration effects on the tunnel are measured using the velocity
and acceleration transducers throughout the blasting construction.
Both the velocities and accelerations of the tunnel along the longi-
tudinal, transverse and vertical directions are recorded to fully
understand the vibration parameters of the tunnel including the
amplitude, frequency spectrum, and time duration, etc.
The goals for the layout of monitoring points and transducers are
listed as follows: obtain the location and direction of the maximum
vibration of the tunnel; track the blasting vibration effects on the
tunnel; and observe the attenuation rules of blasting vibrations.
According to the blasting scheme and actual conditions of the
Fuxing East Road Tunnel, three observation points are arranged
in the monitoring of blasting vibrations. The plane view and verti-
cal view of the location of the monitoring points are shown in
Figs. 1 and 2, respectively. Both the three monitoring points are
located on the bottom plate of the underlayer of the south-line
tunnel, marked as P1, P2 and P3 from west to east, with the adja-
cent distance of 36 m and 22 m respectively. Among the three
monitoring points, point P1 is located in the center of the explosion
of four sections of the diaphragm wall (Fig. 1), and close to the mid-
dle wall of the tunnel. Two velocity transducers are laid out at
point P1 in the vertical and transverse directions of the tunnel
respectively. Three velocity transducers, as well as three accelera-
tion transducers, are arranged at point P2 in the vertical, transverse
and longitudinal directions of the tunnel respectively. In consider-
ation of the limited numbers of transducers, a velocity transducer
is used to monitor the vertical vibration of the tunnel at point P3.
The measurement equipment used in the field monitoring is
mainly composed of the dynamic data acquisition and processing
system (Type CY-1), velocity transducers (Type 941), and accelera-
tion transducers (Type GT41). The dynamic data acquisition and
processing system is applied for data transmission and analysis
in the field monitoring of the blasting construction. The system is
equipped with 32 data acquisition channels which are available
for the connection with velocity and acceleration transducers.
The operation frequency of the system covers the range 0.1–
3000 Hz. Both the velocity and acceleration transducers used in
the field monitoring belong to the piezoelectric vibrometer. The
maximum measuring ranges of the velocity and acceleration trans-
ducers are 0.15 m/s and 25 m/s2 respectively. The operation fre-
quency of both the velocity and acceleration transducers covers
the range 0.06–2000 Hz. The sensors were calibrated before usage
and the proofed accuracies are 0.00005 m/s for the velocity trans-
ducers and 0.0001 m/s2 for the acceleration transducers. In the
field monitoring, three acceleration transducers are fixed longitu-
dinally, transversely or vertically on an iron cube, which is located
at point P2, to accurately obtain the vibration spectra of three-
component accelerations. Similar approaches are applied for the
velocity transducers to monitor the vibration spectra of velocities
at the monitoring points. When monitoring, the wires of transduc-
ers are connected to the channels of the dynamic data acquisition
and processing system in a specified sequence. The data acquisi-
tion frequency of the system is set as 2000 Hz to fully record the
potential frequency ranges of the tunnel vibration induced by the
blasting construction. The data acquisition takes approximately
one hour to complete during the three groups of explosion (Fig. 1).
3.3. Test results and analysis
To illustrate the blasting vibration effects on the tunnel struc-
ture and yet limit the amount of data included in the paper, the
62 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66

velocity time histories of point P1 recorded in the third group of compared to those along the vertical direction during the first
explosion are plotted in Fig. 4, respectively. and third groups of explosion, while for the second group of explo-
It can be seen from the figure that multiple wave crests and sion, both the transverse and vertical basic frequencies of the tun-
troughs are included in the velocity time histories of point P1. Sim- nel are relatively lower and with the same value of 24.94 Hz.
ilar plots are obtained for other monitoring points. This is mainly Based on the analysis of test results (Fig. 5 and Table 1), the
due to the fact that each group of explosion is composed of multi- vibration spectral characteristics of the tunnel structure influenced
ple milliseconds blasting with the time interval of 0.1 s and the by the three groups of explosion are summarized as follows:
reflection of explosion waves occurs at interfaces of different
media. According to the monitoring data, the maximum velocities  Distinct vibration spectra of the tunnel responses are
and accelerations of the three monitoring points during the three observed in the three groups of explosion. Compared to
groups of explosion are listed in Table 1. the first and second groups of explosion, the charge quan-
From Table 1, the blasting vibration effects on the tunnel struc- tities are larger in the third group of explosion, and thus,
ture are summarized as follows: among the three groups of explo- the explosive reaction lasts longer and the blasting energy
sion, the maximum vibration energy occurs in the third explosion, induced by air expansion is larger. Due to the prolonged
and then followed by the first and second explosions, because both reaction duration of both positive and negative pressures
sections of diaphragm wall close to the north-line tunnel are in charge chambers, the basic frequency domain of the tun-
exploded simultaneously in the third explosion (Fig. 1); for each nel responses caused by the explosion is relatively in a
group of explosion, the maximum vibration response (i.e. velocity lower frequency range.
and acceleration) occurs at point P1, and then followed by points  Distinct vibration response spectra of the tunnel structure
P2 and P3, due to the closest distance from the explosion center are observed along the horizontal and vertical directions.
at point P1 compared to other monitoring points; and for each For each group of explosion, the vertical vibration response
group of explosion, the vertical vibration response at each moni- of the tunnel structure is larger than that of the horizontal
toring point is approximately 3–5 times the horizontal vibration vibration, which indicates that the blasting energy of each
response, which indicates that the blasting energy of each group group of explosion is transferred mainly along the vertical
of explosion is transferred mainly along the vertical direction. direction. The basic frequencies of the tunnel vibration
Based on the safety criterion of blasting vibration velocity for along the vertical direction are lower than those along
different structures referred to the Chinese code for blasting the horizontal vibration except for the second group of
safety standards (1999), it can be found that the allowable vibra- explosion. The basic frequency domain of the tunnel vibra-
tion velocity is 10–20 cm/s for traffic tunnels. Based on Table 1, tion along the vertical direction is narrow for each of the
the maximum vibration velocity of the Fuxing East Road Tunnel three groups of explosion. The different vibration response
during the three groups of explosion is 9.7 cm/s, which is close to spectra of the tunnel structure observed along the horizon-
the allowable upper limits of 10 cm/s, and so, the performance of tal and vertical directions are caused by the difference
the tunnel subjected to the blasting construction is in a critical between horizontal and vertical soil stiffness, wave propa-
state of safety. gation direction, and superposition of the explosive waves
In general, regarding the characteristics of blasting loads, the through different soil layers.
following comments apply (Giorgio, 1994): the explosion time is
short, and the blasting vibration waves are aperiodic transient
waveforms with complex frequency components and wide fre- 4. Numerical simulation and analysis
quency band; and the vibration waveforms and frequency compo-
nents of tunnel structures are influenced by the charge quantities, 4.1. FEM model
blasting methods and geological conditions in different extent.
Hence, it is significant to analyze the vibration spectral character- A two-dimensional model (2D) is commonly adopted to clarify
istics of the tunnel structure influenced by the three groups of the dynamic behavior of underground structures subjected to
explosion. blasting loads. However blast-resistance design based on 2D anal-
Fig. 5 show the acceleration response spectra of the tunnel ysis may underestimate complex dynamic response of a tunnel. In
structure at point P2 during the third group of explosion. Based addition, localized three-dimensional (3D) effects such as load
on the analysis of the monitoring data, the dominant frequencies transfer or damage at a portion of tunnel in sharply changed stra-
of the tunnel responses at point P2 during the three groups of tum layers will not be captured by a 2D model. In view of this, a 3D
explosion are summarized in Table 1. model is established to perform dynamic analysis of the tunnel
It can be seen from Table 1 that the basic frequencies of the tun- under blasting construction. The numerical code used for the sim-
nel responses along the transverse direction are quite higher ulations is the finite element method, which comparatively is the

0.03 0.08
Testing results Testing results
0.06 Numerical results
0.02 Numerical results
Horizontal velocity (m/s)

Vertical velocity (m/s)

0.04
0.01
0.02

0.00 0.00

-0.02
-0.01
-0.04
-0.02
-0.06

-0.03 -0.08
0 1 2 3 4 5 6 7 8 0 1 2 3 4 5 6 7 8
(a) Time (s) (b) Time (s)

Fig. 4. Velocity time histories of monitoring point P1 during the third explosion: (a) in the transverse direction; (b) in the vertical direction.
 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66 63

Table 1
Maximum velocity and acceleration responses of the monitoring points.a

Monitoring points The 1st explosion The 2nd explosion The 3rd explosion
Field monitoring Field monitoring Field monitoring Numerical
simulation
Max. Vel Max. Acc Dominant Max. Vel Max. Acc Dominant Max. Vel Max. Acc. Dominant Max. Vel Max. Acc
(m/s) (m/s2) frequency (Hz) (m/s) (m/s2) frequency (Hz) (m/s) (m/s2) frequency (Hz) (m/s) (m/s2)
P1 Long. / / / / / / 0.010 0.92
Trans. 0.023 / 0.012 / 0.026 / 0.017 2.09
Vertical 0.097 / 0.052 / 0.061 / 0.073 6.50
P2 Long. 0.002 / 0.003 / 0.003 / 0.008 2.31
Trans. 0.004 1.944 195.61 0.002 1.199 24.94 0.006 4.907 136.74 0.006 5.10
Vertical 0.013 6.818 12.57 0.010 4.327 24.94 0.013 18.018 9.40 0.012 17.25
P3 Long. / / / / / / 0.002 0.45
Trans. / / / / / / 0.003 1.27
Vertical 0.006 / 0.005 / 0.007 / 0.006 3.18
a
Letter ‘/’ refers to no transducer or fault.

0.010 0.030
Transverse amplitude (g/Hz)

0.025

Vertical amplitude (g/Hz)

0.008
0.020
0.006
0.015
0.004
0.010

0.002 0.005

0.000 0.000
0 200 400 600 800 1000 0 200 400 600 800 1000
(a) Frequency (Hz) (b) Frequency (Hz)

Fig. 5. Acceleration response spectra of monitoring point P2 during the third explosion: (a) in the transverse direction; (b) in the vertical direction.

method used for the most of the dynamic analysis of tunnels (Liu,
2009). The code used in this work is ABAQUS (ABAQUS, 2008).
Fig. 6 shows the finite element model of the Puxi terminal seg-
ment of the Fuxing East Road Tunnel, including the tunnel linings
and surrounding diaphragm walls. The global model dimension is
65 m  27.5 m  19.5 m (length  height  width). The tunnel lin-
ing and diaphragm walls are discretized with four-node shell
elements. The mesh, which includes the tunnel lining and dia-
phragm walls, has been constructed according to the actual tunnel
length and geometric data following the structural design report of
the Fuxing East Road Tunnel (2007). For simplification, a two-node
spring element is used and placed in three directions, i.e. longitudi-
nal, transverse and vertical directions, to simulate the 3D soil-
structure interaction behavior. Note that as soil is poor in sustaining
tension, the behavior of spring element for soil may be taken as uni-
axial. That is, the force generated by the soil spring element is non-
zero if and only if the spring element is in compression. In the
numerical model, the control mesh size is set as 0.5 m. The number
of nodes and elements of the entire model reached 13,807 and
13,906, respectively. These element divisions are sufficient to obtain
accurate solutions as verified by preliminary analyses.
Fig. 6. Finite element model of the tunnel.
An elastic material model is assumed for both the tunnel struc-
ture and diaphragm walls. According to the laboratory testing from
the structural design report of the Fuxing East Road Tunnel (2007), are listed in Table 2 based on the geological exploration data and
the ratio of the dynamic to static Young’s modulus of concrete was laboratory tests following the geological exploration report of the
approximately 1.30 based on laboratory high-frequency pulse tests Fuxing East Road Tunnel (2007). The horizontal foundation modu-
on five concrete samples. Thus the parameters of the tunnel lining lus profiles obtained from laboratory triaxial tests in addition to
and diaphragm walls in the model can be set as follows: Young’s vertical foundation modulus profiles from in situ plate loading
modulus, 36.4 GPa; density, 2500 kg/m3; Poisson’s ratio, 0.2. The tests were used to assess the model parameters for each soil layer
parameters used in the spring elements for different soil layers based on the m method (Yang, 2002). It should be noted that the
64 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66

Table 2
Input parameters of soil layers used in the model (geological exploration report of the Fuxing East Road Tunnel (2007)).

Soil layer Thickness (m) Unite weight Compressive wave Vertical soil Longitudinal /transverse
(kN/m3) velocity (m/s) stiffness (kN/m) soil stiffness (kN/m)
Sandy silt 1.5 18.2 56.2 5500 6000
Mucky silty clay 3.2 17.4 61.1 6000 7000
Sandy silt 2.5 16.8 110.5 18,000 23,000
Mucky clay 3.0 17.6 54.6 5000 5500
Sandy silt 2.0 18.1 52.6 5000 5000
Silty clay 2.6 18.3 73.9 10,000 10,000
Silty clay 2.5 18.3 90.5 15,000 15,000
Silty sand 5.0 18.5 110.3 20,000 25,000
Silty fine sand 5.2 19.5 205.7 85,000 80,000

spatial characteristics of the soil deposits are nearly continuous
(Huang et al., 2005; Yu et al., 2013b), and therefore the soil layers
can be assumed to be horizontal.
Considering the limited computing power, only the Puxi termi-
nal segment is modeled in the analysis. To impose the boundary
conditions more realistically, non-reflecting boundaries are used
on the two truncated boundaries of the model to prevent the influ-
ence of wave reflection, which is discussed in detail in (Yu et al.,
2013a). The adequacy of the type of boundary and mesh size was
verified by running a number of preliminary numerical tests where
the two truncated boundaries of the discretization were placed at
different distances from the Puxi terminal segment.
The blast effect of an explosion is in the form of a shock wave
composed of a high-pressure shock front that expands outward
from the center of the detonation, with pressure intensity decaying
with distance (Balsara, 1970). As the wave front impinges on the
tunnel, a portion of the tunnel will be impacted by the shock pres-
sures. The magnitude and distribution of the blast load acting on
the tunnel then depends on the tunnel geometry and flexibility,
blasting pressure–time history, and the dynamic soil characteris-
tics (Balsara, 1970).
The blast loading may be characterized as a pulse with an expo-
nential-shape time history that attenuates rapidly in amplitude
and broadens as it propagates outward from the detonation center.
Thus it is also necessary to establish the variation and decay of the
incident pressure with time because the effects on the tunnel
structure depend not only on the peak pressure Pmax but also on
the pressure–time history of the blast loading. In general, for soft
soils, Pmax may be estimated from the following expression (U.S.
Dept of Army, 1986):
     n
c C R ing the initial stress state, the shell elements at the planned tunnel
Pmax ¼ 160   
g 144 W 1=3
where c is the unit weight of the soil (Table 2); C is the average
velocity of compressive wave of the soil (Table 2); R is the distance
from the explosion, depending on the location of charge (Fig. 3); W
is the charge weight, 15 kg weight for each drilling hole; and n is the
attenuation coefficient which is controlled by the irreversible crush-
ing of the void volume within a soil matrix by the passage of a stress
wave, in average, n = 2.5 for soils.
The rise time tr taken to reach Pmax may be estimated from (U.S.
Dept of Army, 1986)
R control sections B, D and E are selected in accordance to the mon-
t r ¼ 0:1ta ¼ 0:1
C itoring arrangement, corresponding to the locations of monitoring
where ta is the elapsed arrival time from the instant of detonation to
the time at which the shock arrives at a given point of the tunnel.
From Pmax, the shock wave decays monotonically to nearly zero
over a time period of about one to three times the values of ta in the
expression of the following equation:
Pt ¼ Pmax et=ta
where Pt is the blast pressure at any given time t. Note that the arri-
val time ta is inversely proportional to the seismic velocity, thus an
explosion in high-velocity media such as saturated clay will pro-
duce very short, high-frequency pulses with high accelerations
and low displacements. In contrast, detonations in dry, loose mate-
rials will produce ground motions of much longer duration and
lower frequency.
In order to simulate the blasting construction process more
realistically, the blasting load is applied on the circumferential
boundary of each drilling hole in the diaphragm wall and then
results in vibrations in the existing tunnel lining, which means
the blasting energy was assumed to be absorbed and dissipated
in the ground and lining. The blasting load was simplified as a
shock wave with an exponential-shape pulse, whose pressure peak
was calculated by using empirical equations and duration deter-
mined from the millisecond detonation parameters. The time
interval of each adjacent blasting load was 0.1 s which was deter-
mined according to the millisecond delay detonation parameters.
4.2. Results analysis and comparison
The analysis is implemented in two steps. The first step obtains
the initial stress state before explosion, and the second step ana-
lyzes the dynamic response under the blast loadings. In order to
obtain the initial static stress state, dynamic explicit analysis is
conducted with linearly increasing gravity load over a long period
of time to simulate a quasi-static condition. Static standard analy-
sis is also carried out to check whether the initial stress state
obtained from explicit dynamic analysis is similar to those from
static analysis. A difference of less than 10% is found. After achiev-
ð1Þ
locations are switched to null model to model tunnel excavation.
The numerical simulations of the three groups of explosion are
then conducted each for 10 s, with very small time-step of
1.0  103 ms. Test runs are carried out to determine the length
of analysis and time step and it is found that the above two param-
eters are adequate in capturing the major response.
The results obtained from the numerical model provide a direct
and global understanding of dynamic response characteristics of
the system. With the analysis of the results from the three groups
of blasting simulations, five control sections are set for the follow-
ing data analysis, i.e. sections A, B, C, D, and E in Fig. 6. Of which,
ð2Þ
points P1, P2 and P3 (Fig. 1), respectively. Three critical points in
each of vertical cross sections, which are points 1 and 3 close to
the middle wall and point 2 close to the side wall of the south-line
tunnel, are chosen to discuss the tunnel response, as shown in
Fig. 6. According to the monitoring results analysis, the maximum
vibration energy occurs in the third explosion due to the large
explosive power induced by the simultaneous initiation of the
ð3Þ two sections of diaphragm wall close to the north-line tunnel
 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66 65

(Fig. 1). Hence, in order to provide a global understanding and yet
limit the amount data included in the paper, dynamic responses of
the tunnel subjected to the third group of explosion are presented
and discussed in this section.
The numerical simulation based on the finite element model of
the tunnel is implemented and the results are compared with
those from the field monitoring data. Fig. 4 shows the transverse
and vertical velocity responses at monitoring point P1 (i.e. control
section B in Fig. 6) during the third group of explosion, respec-
tively. It can be seen from the figure that both the transverse and
vertical velocity responses of the tunnel structure from numerical
simulation have a similar shape and characteristic with those from
the field monitoring data, that is, the velocity responses increase
dramatically immediately after the blasting loads, and as time
approaches infinity, the velocity responses all diminish to zero.
The figure reveals that the maximum transverse and vertical veloc-
ities of the monitoring point P1 obtained from the numerical sim-
ulation are 1.7 cm/s and 7.3 cm/s respectively, which compare well
with the testing results.
To fully understand the dynamic behaviors of the tunnel struc-
ture under blasting loads, the maximum velocity and acceleration
responses at the five control sections (sections A, B, C, D, and E in
Fig. 6) of the tunnel during the third group of explosion are illus-
trated in Table 3. It should be noted that the maximum vibration
response for each control section basically occurs at the critical
point 3 in Fig. 6, i.e. close to the middle wall of the underlayer of
the south-line tunnel. It can be seen from the table that among
the five control sections, the maximum vibration response (i.e.
velocity and acceleration) occurs at section B (Fig. 6), due to its
closest distance from the explosion center compared to other con-
trol sections, and the vibration response attenuates rapidly in
amplitude with the increasing distance from the explosion center.
Furthermore, the vertical vibration response at each control sec-
tion is quite larger than the horizontal vibration responses. These
are consistent with the monitoring results as mentioned above.
Table 1 shows the comparison between monitoring data and
numerical results of the tunnel response (i.e. velocity and accelera-
tion) under the third group of explosion. It can be seen from the table
that both the velocity and acceleration responses of the tunnel struc-
ture obtained from the numerical simulation are in accordance with
those from the field monitoring data, even if the difference of
response amplitude exists. A couple of comments are given to
explain the difference: the parameters used in the numerical analy-
sis of the tunnel structure subjected to the blasting loads may be
divided into two major groups: soil parameters and blasting charac-
teristics; the soil parameters are determinable or controllable
whereas the blasting characteristics are most likely undeterminable
or uncontrollable. However, the difference from Table 1, if it exists, is
extremely small. Thus the validation of the numerical simulation of
the tunnel responses under blasting loads is verified.
For structural lining design, provisions in the Chinese code for
design of concrete structures 2010 may be adopted. The lining is
considered as a ‘‘wall’’ for calculation of its capacity. The capacity
for a 0.6 m thick lining with 22 numbers of D30 high yield bars,
in the form of force-moment interaction diagram, is presented in
Fig. 7. The values of the lining thrust and moment obtained from
the above blasting analyses are also shown in this figure so as to
evaluate the performance of the lining, which is safe for all the
three groups of explosion.
5. Discussions
Based on the monitoring data and numerical results of the Fux-
ing East Road Tunnel subjected to blasting loads, the maximum
vertical and horizontal velocities of the tunnel structure during
the three groups of explosion are 9.7 cm/s and 2.6 cm/s respec-
tively. The former value is quite close to the allowable upper limits
of 10–20 cm/s based on the Chinese code for blasting safety stan-
dards (1999), and so the performance of the tunnel structure under
the blasting construction is in a critical state of safety. It can be
taken as an indication that the blasting construction scheme
should be implemented carefully.
Compared to the traditional shield tunneling, the parameters of
the shield tunneling through the blasting zone are different in the
following issues:
 The total thrust of shield tunneling was quite higher. The
total thrust in the process of shield tunneling was actually
maintained around 3000 ton;
 The torque of shield cutter was also higher. In the process
of shield tunneling, the used torque of shield cutter was
around 70–80% of the total capacity;
 The shield tunneling speed should be appropriately slowed
down to prevent the shield cutter from being stuck.
The shield tunneling speed was actually controlled within
2–3 mm/min;
The blasting construction of the Fuxing East Road Tunnel is sig-
nificant due to it is the first time for the construction technology of
controlled blasting to be employed in tunnel construction in
Shanghai. The shield construction of the planned subway tunnel
directly depends on whether the blasting construction is successful

Table 3 600
Maximum velocity and acceleration responses of the tunnel during the third Bearing capacity
explosion. Numerical results
500
Control sections Max. velocity (m/s) Max. acceleration (m/s2)
A Vertical 0.052 8.30 400
Moment (kN.m)

Transverse 0.013 2.80

Longitudinal 0.005 1.32
300
B Vertical 0.073 17.25
Transverse 0.017 5.10
Longitudinal 0.010 2.31 200
C Vertical 0.032 15.63
Transverse 0.018 4.20
100
Longitudinal 0.009 1.25
D Vertical 0.012 6.50
Transverse 0.0062 2.09 0
Longitudinal 0.008 0.92 0 2000 4000 6000 8000 10000 12000
E Vertical 0.0062 3.18 Axial force (kN)
Transverse 0.0025 1.27
Longitudinal 0.0020 0.45 Fig. 7. Force–moment interaction diagram for a 0.6 m thick side wall with 22
numbers of D30 bars per meter.
66 H. Yu et al. / Tunnelling and Underground Space Technology 43 (2014) 59–66
or not. Therefore, the blasting scheme and its implementation
should be feasible and reliable in the blasting construction. For this
purpose, the following practical recommendations are provided:
(1) High-precision drilling is required as the basis of the blasting
implementation. In the blasting construction, two theodo-
lites are employed to alternately monitor the perpendicular-
ity of drill rods, and the diameter of drill rod is increased to
improve its stiffness. The blasting scheme of underground
diaphragm walls with such high depth is infeasible and the
size of blasting fragment cannot be guaranteed, if without
the required high-precision drilling.
(2) The size of blasting remnants after explosion should be small
enough to ensure the shield machine passing through the
blasting zone successfully. For this reason, three RDX deto-
nators are adopted to initiate the explosion of each single
explosive and the multiple initiating networks are estab-
lished and employed on the ground.
After the blasting construction, the shield tunneling passed
through each section of target diaphragm wall depending on the
excavation capacity of the shield itself. Each penetration took
about 24 h to complete. The shield cutter, the circumferential cut-
ter and the arrow cutter originally equipped in the shield cutting
wheel were well preserved in the whole process of shield tunnel-
ing through the blasting zone. However, big blasting fragments
did appear occasionally in the process of penetration and the open-
ing of auger conveyor had to be dismantled to benefit slagging. It is
therefore suggested that a reserved embedded explosive hole
should be considered in the future design of retaining structures
instead of vertical drilling to further improve the blasting accuracy
and increase the fragmentation degree of blasting remnants.
6. Conclusions
To maintain safety and minimize effects of blasting vibration on
the existing tunnel structure, the reasonable controlled blasting
technique is selected with consideration of the geological condi-
tions and the surrounding environment, and four sections of dia-
phragm wall are divided to be exploded in three groups. The
field monitoring of blasting vibrations is carried out on the bottom
plate of the tunnel and the blasting vibration effects on the tunnel
are measured using the velocity and acceleration transducers
throughout the blasting construction. A 3D numerical model
including geological data and tunnel geometry is established to
calculate and analyze the dynamic responses of the tunnel struc-
ture subjected to blasting vibrations. The validation of numerical
results is verified by comparison with field monitoring data.
Based on the monitoring data and numerical results of the Fux-
ing East Road Tunnel subjected to blasting loads, the following can
be concluded:
(1) For each group of explosion, the vertical vibration response
at each monitoring point is approximately 3–5 times the
horizontal vibration response, which indicates that the
blasting energy of each group of explosion is transferred
mainly along the vertical direction.
(2) Distinct vibration spectra of the tunnel responses are
observed in the three groups of explosion. Due to the pro-
longed reaction duration of both positive and negative pres-
sures in charge chambers, the basic frequency domain of the
tunnel responses caused by the explosion is relatively in a
lower frequency range.
(3) Distinct vibration response spectra of the tunnel structure
are observed along the horizontal and vertical directions.
The basic frequency domain of the tunnel vibration along
the vertical direction is narrow for each of the three groups
of explosion.
(4) The maximum vibration velocity of the tunnel during the
three groups of explosion is 9.7 cm/s, which is close to the
allowable upper limits of 10 cm/s, and so, the performance
of the tunnel subjected to the blasting construction is in a
critical state of safety. Hence, the blasting construction
scheme should be implemented carefully.
Acknowledgements
The research is mainly supported by the National Natural Sci-
ence Foundation of China (51208296), the Fundamental Research
Funds for the Central Universities (2013KJ095), and Shanghai
Educational Development Foundation. The writers would also
gratitude financial resources from the National Key Technology
R&D Programs (2011BAG07B01 and 2012BAK24B04) and PCSIRT
(IRT1029). We would also like to acknowledge the anonymous
reviewers who played a significant role in shaping and improving
the manuscript.
References
ABAQUS/explicit user’s manuals, version 6.8, 2008. Hibbit, Karlsson and Sorensen
Inc.
Ahmed, L., Ansell, A., 2011. Structural dynamic and stress wave models for the
analysis of shotcrete on rock exposed to blasting. Eng. Struct. 35, 11–17.
Ansell, A., 2004. In situ testing of young shotcrete subjected to vibration from
blasting. Tunn. Undergr. Space Technol. 19, 587–593.
Balsara, J.P., 1970. Blast loaded buried arches. J. Eng. Mech. Div., ASCE 96, 1–16.
Feldgun, V.R., Kochetkov, A.V., Karinski, Y.S., Yankelevsky, D.Z., 2008. Internal blast
loading in a buried lined tunnel. Int. J. Impact Eng. 35, 172–183.
GB50010-2010, 2010. Chinese Code for Design of Concrete Structures. China
Architecture & Building Press, Beijing (in Chinese).
GB6722-2003, 1999. Chinese Code for Blasting Safety Standards. China Standard
Publishing Company, Beijing (in Chinese).
Giorgio, B., 1994. Blasting-induced vibration in tunnelling. Tunn. Undergr. Space
Technol. 9, 175–187.
He, W., Chen, J.Y., Guo, J., 2011. Dynamic analysis of subway station subjected to
internal blast loading. J. Central South Univ. Technol. 18, 917–924.
Huang, Y., Ye, W.M., Tang, Y.Q., Chen, T.L., 2005. Characteristic analysis for seismic
ground response of soft soils in Shanghai. Chin. J. Undergr. Space Eng. 1, 773–
778 (in Chinese).
Jiang, N., Zhou, C.B., 2012. Blasting vibration safety criterion for a tunnel liner
structure. Tunn. Undergr. Space Technol. 32, 52–57.
K200705-C, 2007. Geological Exploration Report of the Fuxing East Road Tunnel.
Shanghai Tunnel Engineering & Rail Transit Design and Research Institute,
Shanghai (in Chinese).
K200710-C, 2007. Structural Design Report of the Fuxing East Road Tunnel.
Shanghai Tunnel Engineering & Rail Transit Design and Research Institute,
Shanghai (in Chinese).
Liu, H., 2009. Dynamic analysis of subway structures under blast loading. Geotech.
Geolog. Eng. 27, 699–711.
Lu, W.B., Yang, J.H., Chen, M., Zhou, C.B., 2011. An equivalent method for blasting
vibration simulation. Sim. Model. Practice Theory 19, 2050–2062.
Ma, G.W., Hao, H., Zhou, Y.X., 1998. Modeling of wave propagation induced by
underground explosion. Comput. Geotech. 22, 283–303.
Nateghi, R., 2011. Prediction of ground vibration level induced by blasting at
different rock units. Int. J. Rock Mech. Mining Sci. 48, 899–908.
Ozer, U., 2008. Environmental impacts of ground vibration induced by blasting at
different rock units on the Kadikoy–Kartal metro tunnel. Eng. Geol. 100, 82–90.
Peck, RB., 1969. Deep excavations and tunneling in soft ground. In: Proceedings of
the 7th international conference on soil mechanics and foundation engineering,
Mexico City, pp. 225–290.
U.S. Department of Army. Fundamentals of protective design for conventional
weapons: TM 5–855-1, Headquarters of the U.S. Department of the Army,
Washington, D.C., 1986.
Wu, C.Q., Lu, Y., Hao, H., 2004. Numerical prediction of blast-induced stress wave
from large-scale underground explosion. Int. J. Num. Anal. Meth. Geomech. 28,
93–109.
Yang, LD., 2002. Research on Static and Dynamic Properties of Soft Soils in Shanghai,
Report No. K200293-S, Tongji University, Shanghai. (in Chinese).
Yu, H.T., Yuan, Y., Bobet, A., 2013a. Multi-scale method for long tunnels subjected to
seismic loading. J. Num. Anal. Meth. Geomech. 37, 374–398.
Yu, H.T., Yuan, Y., Qiao, Z.Z., Gu, Y., Yang, Z.H., Li, X.D., 2013b. Seismic analysis of a
long tunnel based on multi-scale method. Eng. Struct. 49, 572–587.
Zhu, W.C., Tang, C.A., 2006. Numerical simulation of Brazilian disk rock failure
under static and dynamic loading. Int. J. Rock Mech. Mining Sci. 43, 236–252.