Content uploaded by Ferenc Deák
Author content
All content in this area was uploaded by Ferenc Deák on May 30, 2022
Content may be subject to copyright.
International Society for Rock Mechanics and Rock Engineering
Norwegian Group for Rock Mechanics
ISRM International Symposium
Eurock 2020 – Hard Rock Engineering
Trondheim, Norway, 14-19 June
Excavation damage zone behaviour under dynamic loading
F. Deák
Mecsekérc Ltd., Pécs and Budapest University of Technology and Economics, Department of Engineering
Geology and Geotechnics, Hungary
deakferenc@mecsekerc.hu (email of corresponding author)
M. A. Perras
Department of Civil Engineering, Lassonde School of Engineering, York University, Toronto, Canada
Á. Török
Budapest University of Technology and Economics, Department of Engineering Geology and Geotechnics,
Budapest, Hungary
Abstract
The program for the final disposal of low and intermediate level radioactive waste has been
established by the Paks Nuclear Power Plant, Hungary. The Central Nuclear Financial Fund and the
Public Limited Company for Radioactive Waste Management (PURAM) has been established to
coordinate organizations and activities for all tasks in connection with nuclear waste treatment. The
project was started with a geological screening in order to find the most suitable geological formation
for a L/ILW radioactive waste repository. The selected potential host rock is the Mórágy Granite
Formation in the south-western part of Hungary, close to the village of Bátaapáti.
Construction of the Low and Intermediate Level waste repository in Bátaapáti was done using drill
and blast excavation methods. The excavation process affects the rock mass by producing zones of
damage, collectively termed the Excavation Damage Zones (EDZs). Moving away from the
excavation, the EDZs can be distinguished as the Highly Damaged Zone (HDZ), the inner EDZ
(EDZi) and the outer EDZ (EDZo). Beyond the EDZ is stress and/or strain influence zone that involves
only elastic change, the Excavation Influence Zone (EIZ).
The aim of the present study is to examine the time series data of the blasting process to determine the
extent of the EDZs. In light of recent updates in Earthquake hazard assessment for the repository, the
long-term influence on the EDZs due to dynamic loading from an earthquake is also assessed. Both
the influence of blasting and the earthquake on the EDZs are modelled dynamically and their control
on the extent and geometry of the EDZs is discussed.
Keywords
Excavation damage zones, blasting, earthquake, vibration, dynamic behaviour
Eurock 2020 – Hard Rock Engineering
2
1 Introduction
Since the first excavation blast rounds in the Bátaapáti radwaste repository tunnels, minimizing the
blasting damage in the vicinity of the excavation surface has been a priority. During the blasting
design, one of the most important points was the prediction of blast induced vibration (Lu et al., 2011).
It is generally understood that the rock vibration during blasting arises from explosion-induced seismic
waves, and blast loading is the excitation source of rock vibration (Yang et al. 2016). When the
explosives are detonated, shock waves are generated and propagated from the source. The dissipated
energy content of the elastic waves is adsorbed by the rock mass as a function of the distance from the
blast holes.
The characteristic shock wave propagating in a rock mass can be estimated from attenuation
relationships or during numerical modelling. There are numerous attenuation relationships, relating the
peak particle velocity (PPV) with scaled distance. The scaled distance is defined as the ratio of the
distance from the charge point to the square root of the charge mass, expressed in TNT net equivalent
charge weight and are most often used to predict the amplitude of vibration (Dowding, 1984 and
Kumar et al., 2016). Typical vibration measurements are carried out far from the blast source, hence
the rock mass inside the zone of measurement are not well defined in terms of attenuation, for instance
near field vibration is more difficult to characterized. In addition, the attenuation curves used in
practice cannot be used to predict the vibrations close to the detonation.
During rock blasting excavation, only a small portion of the energy released in the explosion is
utilized directly for rock fragmentation, but most of the explosion energy is dissipated in the form of
ground vibration, air blasts, flying rocks and noise (Hagan 1977). Responses and damage to structures
subjected to vibration depend not only on PPV, but also on vibration frequency content and therefore
there is a need to pay more attention to the vibration frequency mechanism and behaviour of the rock.
The in situ stress redistribution is a dynamic process that starts from the transient release of stress on
excavation boundaries and reaches the final static secondary stress after excavation (Yang et al. 2017).
Dynamic properties of rocks include Young’s modulus, Poisson’s ratio and a damping ratio. These
properties of the rock mass are not well documented in the scientific literature and are typically
selected based on experience. However, the lack of data is even more sparse for a jointed rock mass,
with the damping ratio having the highest uncertainty (Ahn et al. 2016).
In this study, we simulate the effect of a tunnel face blast, with an advance of 2.0 m in length using a
21 m2 tunnel cross section. 2D finite element modelling was performed with the RS2 Rocscience
software by applying both the static loads and the dynamic loads. During the 2D plane strain
modelling it is difficult to determine the damping ratio, because it is a 3D problem. The damping ratio
has to be additionally adjusted for the different attenuation characteristics of a 2D analysis since the
spherically induced blast load propagates through the 3D environment (Deák et al. 2018).
For the blastholes detonated in deeper rock masses, the presence of in situ stress suppresses the
development of blast-induced tensile cracks around blastholes, and fractures are mainly aligned along
the maximum in situ stress orientation perpendicular to the blasthole axis (Kutter and Fairhurst 1971;
Donze et al. 1997; Yilmaz and Unlu 2013). In an underground excavation with different blasting
sequences, the maximum principal stress exerted on the rock mass around the blastholes is generally
parallel to the connecting line of blastholes due to the in situ stress redistribution after blasts of the
former delays. Therefore, in most underground blasts, the blast-induced rock cracks are initiated and
propagated preferentially along the connecting line of adjacent blastholes (Yang et al. 2016).
Many studies indicate that the in situ stress release due to blasting in very short duration will trigger
stress waves in surrounding rock masses, and the stress redistribution in the vicinity of the excavation
walls due to blasting excavation is a dynamic process that starts from the transient release of in situ
stress to the final static stress state after vibrations cease (Miklowitz 1960; Carter and Booker 1990).
Rock masses contain intact rocks and discontinuities of different sizes ranging from microcracks to
faults, and the presence of these discontinuities contributes to the propagation of low-frequency
vibration by filtering the higher frequencies (Park et al. 2009; Li et al. 2011). Hence it is a top priority
to use hybrid and discrete numerical models not only continuum solutions. This consideration is
worth following in the case of earthquakes too.
Eurock 2020 – Hard Rock Engineering
3
It is known that the underground excavations are generally resistant to earthquake-induced motions
(Dowding et al., 1978). By combining the static and dynamic loads of an earthquake, the earthquake
resistance is becoming more complicated. In this study, the blasting and earthquake dynamic loading
on the EDZs is examined in one model at the same time.
2 The selected time histories for the numerical modelling
2.1 Passive seismoacoustic measurements
During the excavation of the inclined twin access tunnels of the repository, numerous seismoacoustic
measurements were carried out to determine the blasting effects on the neighboring tunnel. Two
measurement locations were selected for analysis in this paper where the actual tunnel face coincided
with one of the measurement points. The blasts were executed in the Western access tunnel (Blast 1. at
CH 357 m; Blast 2. at CH 422 m), while the measurement points were situated in the Eastern tunnel
(the pillar is ~26 m). At the measured tunnel sections, the depth was ~70 m from the ground surface.
At both measurement locations 6 accelerometers were mounted on the top of the 1 m long rock bolts
inside the rock mass which were installed at a height of 1-2 m from the tunnel floor into the side wall.
The direction of sensitivity of the accelerometers was perpendicular to the tunnel wall. In the Eastern
access tunnel, point 3. mp (Blast 1. in a rock mass with GSI = 40-50) and 4. mp (Blast 2. in a rock
mass with GSI = 60-70) were the only sensors which measured the acceleration in one direction
(horizontal) in line with the Western access blasted tunnel faces. During the vibration measurements
the conventional work was interrupted in both tunnels, the ventilation system was turned off and the
overall conditions could be considered “seismically silent”.
The main technical parameters of the monitoring setup were as follows:
• sampling density of 3000 samples / second / channel
• frequency range of 0-1200 Hz
• minimal measurable value (basic sensitivity) of 0.001 m/s2
• maximum measurable value of 100 m/s2
By using a high sampling density, it was also possible to examine the whole blasting process and to
record the high frequency components. Different delays can be easily identified in the data series, as
seen in the recorded accelerograms in Fig. 1. Taking these advantageous positions into consideration,
the results are presented, followed by back analysis modelling of the repository tunnel excavation
blasting and in addition earthquake acceleration is also considered.
2.2 Selected earthquake time series
A complex earthquake hazard evaluation study was carried out by Deák et al. (2019) to investigate the
influence on the underground disposal facilities. Based on the probabilistic seismic hazard evaluation
the exceedance curves and the uniform hazard response spectrums were calculated for the bedrock
with the following return periods: 100, 600, 10^4, 10^5 and 10^6 years.
From the Pacific Earthquake Engineering Research Center (PEER) database the real measured time
histories (in X, Y, Z directions) were extracted and matched to the calculated target spectrums with the
previously mentioned return periods. The matching and final time history selection procedures were
verified by using different artificial and synthetic seismograms. Fig. 2 shows the selected
accelerograms for input into the numerical model.
3 Numerical analysis
The numerical investigation was carried out by using the RS2 (Rocscience) software, taking advantage
of its dynamic capabilities.
The amplitude of vibration induced by blasting decreases with increasing distance from the source.
Due to attenuation, the decay of vibration is produced by two phenomena, which are geometrical
spreading and material damping (Dowding, 1996). The geometrical damping is caused by the
Eurock 2020 – Hard Rock Engineering
4
expansion of the surface over which the vibration energy is transmitted. The material damping is
caused by the nonlinear hysteretic behaviour of the geologic media (Park et al. 2015).
Fig. 1 Recorded accelerogram at the measurement point CH 357 m 3. mp (A.) and CH 422 m 4. mp (B.) – delays are
assigned by red points; at right figures are shown the time-varying frequency components (spectrogram) of the presented
accelerograms.
Fig. 2 Earthquake accelerograms in horizontal direction – 100 years return period (A.) and 600 years return period (B.)
During the preliminary work simplifications were made, such as the equations only described the
amplitude and not the time series of the propagated motion. The registered amplitudes were amplified
and then the dynamic load normal to the tunnel wall was defined. In this preliminary study, no
distinction between space and time was made for different delays. As a further investigation, it’s not
difficult to specify the accurate time delays, because we have the registered times and the used delay
times during the blasting process, accordingly, in this case, the question can be solved with simple
time correction.
For the attenuation, the following equation was used (Kim and Lee, 2000), see Eq. 1.:
A.
B.
Eurock 2020 – Hard Rock Engineering
5
𝐴2= 𝐴1(𝑟1
𝑟2)𝑚
(1)
Where A1 Is the amplitude of vibration at distance
r2 From the source
A2 Is the amplitude of vibration at distance
r2 From the source
m Is a geometric coefficient (for underground blasting, m=1.0)
From the previous investigations (Deák et al. 2018, 2019) and from the scientific literature (Lu et al.
2011, Kim et al. 2000, Ahn et al. 2016), it can be assumed that blasting creates a fractured zone around
the blast holes and the elastic waves are propagated beyond this crushed zone. This zone was assigned
to be a circle with a radius 1 m (starting from the tunnel wall, around the tunnel section. During the
amplification r2 = 1 m was used and the blast load was applied normal to the excavation surface. The
same solution was used in the current research.
Damping was solved by using Rayleigh damping. With this type of damping, the damping matrix that
relates the damping force and velocity of the system is expressed in terms of the stiffness and mass
matrix of the system. The damping becomes proportional to the mass and stiffness of the system via
Eq. 2.
[𝐶]= 𝛼[𝑀]+ 𝛽[𝐾]
(2)
Where [A] Is the damping matrix
[M] Is the mass matrix
[K] Is the stiffness matrix
α, β Are Rayleigh coefficients that determine the frequency dependence of the damping
formulation
The damping of the soils and rocks is independent of the loading frequency (Ahn et al. 2016). By
contrast, Rayleigh damping formulation is frequency dependent and adequate to solve numerical
damping. Use of Rayleigh damping in the numerical model allows the user to define the damping ratio
for two frequencies. Generally, the frequency range between the two defined threshold frequencies,
has a damping ratio lower than the specified damping ratios and frequencies outside this range are
damped more heavily. Numerous variations of the two frequencies and damping ratios were modelled
and a more appropriate result was set to the measured one. Based on the mentioned settings the
software automatically calculates the α and β values.
3.1 Input parameters and model construction
To reproduce the non-linear “S-shaped” spalling model, Diederichs (2007) and Perras et al. (2016)
presented a simplified constitutive model called the Damage Initiation and Spalling Limit (DISL)
approach, which can be used with different numerical modelling software to accommodate a strain
weakening constitutive model, using either the generalized Hoek-Brown model (Hoek et al. 2002) or
an equivalent Mohr-Coulomb approach.
It should be noted that using of generalized Hoek-Brown model gives more reliable results in case of
different rock mass qualities and different in situ stress conditions.
The DISL is capable to predict brittle rock spalling as a function of confinement, within underground
excavations. Simultaneously with the DISL (at a rock mass with GSI > 55) was used the generalized
Hoek-Brown criteria as a constitutive model (at rock mass with GSI < 55, in this case, the residual
strength parameters were determined by using GSI based on Cai et al. 2007 suggestions).
The input material parameters were generated by using the database from the geotechnical integration
report (Kovács et al. 2012) shown in Table 1. The excavated tunnel section is modelled as 21 m2 area.
The experimental modelling was carried out by using six stages: in the first stage the rock mass was
defined with no excavation, in the second stage the tunnel was excavated, in the third stage the
blasting dynamic load was applied (stage time till 1 s), the fourth stage continuing the blasting (till 3
s), in the fifth stage the earthquake dynamic load was applied (till 13 s), sixth stage (till 25 s). The
resulting zones became much larger in space and their shape was also compared to the static load
stages.
Eurock 2020 – Hard Rock Engineering
6
Table 1 Material properties
DISL Peak
DISL Residual
GSI Peak
GSI Residual
UCS (MPa)
129
129
m
4.30
10
2.515
0.712
s
0.020
0.001
0.0038
0.0003
a
0.502
0.75
0.506
0.529
The following figure shows the dynamic boundary conditions of the examined models (Fig. 3).
Fig. 3 Dynamic boundary conditions of the models and a simple sketch on the calibration procedure of the real measured
blasting accelerogram.
3.1.1 Stresses and depth
Each of the various calibration models in case of measured blasting time histories were assessed at a
depth of 70 m below ground surface. In case of the EDZs investigation the models are moved to into
deeper areas, where are positioned the repository chambers with ~270 m overburden.
Major and minor principal stress ratios in the model plane (KHh) of 1.0 at the blasting monitoring
points and at the chambers area 1.35 values were used.
4 Modelling results
In this work, just one series of modelling results are presented. This is a model by using the
generalized Hoek-Brown constitutive model at the repository chambers area (270 m overburden and
KHh of 1.35). Even at 270 m overburden, the DISL approach did not yield a plastic yield zone and as
such the Hoek-Brown approach was used. To this model dynamic loads, the blasting (CH 422 m 4. mp
– Fig.1/B.) and the earthquake (600 years return period) time histories were added (Fig. 2/B).
Based on the presented observations and using the work of Perras et al. (2010, 2012, 2016), the
yielded elements, volumetric strain, and principal stress concentrations were found to be the best
indicators for determining the depth of different EDZs. These values are plotted against the distance
from the excavation surface at the roof of the modelled horseshoe tunnel section (Fig. 4). The values
were measured along a line which passes from the centre of the excavation through to the deepest
yielded element away from the excavation. The contours of the yielded zones are shown for each mean
model case as an inset image in Fig. 4. Plastic yielding indicates that the peak elastic properties have
been exceeded, which results in the onset of distributed damage in the rock mass.
The extensile of the yielding zone increasing significantly when the dynamic loading from the blast is
activated (Fig. 4/B.) and followed by another increase when the earthquake load is added (Fig. 4/D.)
The extension volumetric strain still increasing drastically after the static loads by the initiation of the
dynamic loading from the blast. The outer limit of plastic yielding, therefore, corresponds to the outer
limit of the EDZo. The EDZo zone is one of which extension decreased during the initiation of the
dynamic loads (Fig. 4/B., C., D.).
Eurock 2020 – Hard Rock Engineering
7
Fig. 4 EDZ development in the tunnel roof area; A. without dynamic loads; B. blast dynamic load – stage time 1 s; C. blast
dynamic load – stage time 3 s; D. earthquake dynamic load – stage time 13 s; E. earthquake dynamic load – stage time 25 s;
Inset model plots in each case are figures corresponding to the yielded elements.
The start of extensile volumetric strain, used as the indicator for the start of the EDZi. The volumetric
strain reversal point is also consistent with a decrease in the confining stresses and the steepest slope
of the distance versus the maximum shear strain. As the extensile volumetric strain continues to
increase, it reaches a maximum value, which coincides with minimum principal stresses and continued
increase in shear strains. This rapid expansion of a true brittle material would result in visible
fractures. The HDZ limit is selected as the first point where σ3 begins to increase from the level at the
excavation boundary. The extension of HDZ is increasing with a similar trend as the plastic zone
increases. This indicates that the rock mass is beginning to be able to carry some load and therefore
macro fractures would be limited in length due to increasing confinement moving away from the
excavation boundary (Fig. 4/A.).
During the modelling process, several simplifications had to be introduced that call into question the
results of the HDZ. The whole blasting time histories were added normal to the tunnel contour, as
there was no opportunity to use a sequential full-face millisecond delay blasting solution (Yang et al.
2017). It was not possible to combine separate blast hole mechanical damage resulting from the high-
pressure gases with the dynamic loading due to the shock waves. The combined effects would
influence the HDZ significantly.
In summary, there is a clear increase in the extent of the EDZs due to the dynamic effects. Further
investigations should focus on hybrid or discrete element modelling approaches to examine circular
Yielded Elements (%)
B.
A.
C.
E.
D.
Eurock 2020 – Hard Rock Engineering
8
shape tunnels and real constructed tunnel shapes with different rock mass properties, depths and in situ
stress conditions, as well as the blast hole specific influences.
5 Conclusions
The work presented here highlighted the importance of complex numerical investigations for drilling
and blasting excavation and earthquake dynamic effects on the rock mass in the vicinity of an
underground excavation. A correct method of examination of the EDZs is required to use static and
dynamic approaches at the same time and to incorporate these experiences into the design.
The model was calibrated to existing acceleration measurements and therefore it was possible to
define the site-specific attenuations. It was shown that the near field vibration can be accurately
captured by the models, by using a very dense mesh, while the accuracy of the far-field attenuation is
influenced by the Rayleigh damping.
In underground excavations under high in situ stress the static component of the dynamic stress
redistribution becomes the main contributor in the EDZs development.
This work shows the sensitivity of the previously induced excavation damage zone depths using the
method for brittle rocks (Perras et al. 2016) due to the dynamic loads. Further investigations are
needed to compare with different numerical modelling methods and to produce calibrations with in
situ blasting PPV or acceleration measurements. This will improve our understanding of the EDZs
dynamic behaviour. Another issue is the control and reliable application of the proper pairing of the
blasting and earthquake dynamic loads within a numerical model.
During the characterization of the rock mass damage zones at the vicinity of the underground caverns,
the presented approach has a great significance in the design of the engineered barriers for radioactive
waste repositories and in the scaling of the pillars for tunnelling and mining.
References
Ahn J-K, Park D, Yoo J-K (2016) Estimation of damping ratio of rock mass for numerical simulation of blast
induced vibration propagation. Japanese Geotechnical Society Special Publication, vol. 2: 1589-1592.
https://doi.org/10.3208/jgssp.kor-34.
Cai M, Kaiser PK, Tasaka Y, Minami M (2007) Determination of residual strength parameters of jointed rock
masses using the GSI system. International Journal of Rock mechanics & Mining Sciences 44 (2007):
247-265. https://doi.org/10.1016/j.ijrmms.2006.07.005.
Carter JP, Booker JR (1990) Sudden excavation of a long circular tunnel in elastic ground. Int J Rock Mech Min
Sci Geomech Abstr 27(2):129–132. https://doi.org/10.1016/0148-9062(90)91464-i.
Deák F, Grabarics J, Lipovics T, Sándor Cs (2019) Earthquake hazard evaluation of National Radioactive Waste
Repository (NRWR) underground facilities. PURAM Report, RHK-K-009/19 (in Hungarian language),
299 p.
Deák F, Szűcs I (2018) Dynamic investigations of EDZs from Bátaapáti radwaste repository based on passive
seismoacoustic measurements. EUROCK2018, Geomechanics and Geodynamics of Rock Masses –
Litvinenko (Ed.), Taylor & Francis Group, London: 491-498.
https://doi.org/10.1016/j.proeng.2017.05.264.
Diederichs MS (2007) The 2003 Canadian Colloquium: Geotechnical Mechanistic interpretation and practical
application of damage. Canadian Geotechnical Journal, 2007, 44(9): 1082-1116.
https://doi.org/10.1139/t07-033.
Donze FV, Bouchez J, Magnier SA (1997) Modeling fractures in rock blasting. Int J Rock Mech Min 34: 1153–
1163. https://doi.org/10.1016/s1365-1609(97)80068-8.
Dowding C (1996) Construction vibrations, Prentice Hall.
Dowding CH, Rozen A (1978) Damage to rock tunnels from earthquake shaking. J Geotech Eng. Div. 1978;
104(GT2): 175–91. https://doi.org/10.1016/0148-9062(78)91355-4.
Hagan TN (1977) Rock breakage by explosives. Acta Astronaut 6(3–4):329–340. https://doi.org/10.1016/0094-
5765(79)90102-4.
Hoek E, Carranza-Torres C, Corkum B (2002) Hoek-Brown failure criterion 2002 edition. In: Proceedings of
NARMS-TAC Conference, Toronto, Canada, 1: 267-273.
Eurock 2020 – Hard Rock Engineering
9
Kim DS, Lee JS (2000) Propagation and attenuation characteristics of various ground vibrations. Soil Dynamics
and Earthquake Engineering, 19(2): 115-126. https://doi.org/10.1016/s0267-7261(00)00002-6.
Kovács L, Deák F, Somodi G, Mészáros E, Máté K, Jakab A, Vásárhelyi B, Geiger J, Dankó Gy, Korpai F,
Mező Gy, Ván P, Fülöp T, Asszonyi Cs (2012) Review and extension of the Geotechnical Integration
Report. RHK-K-032/12 – PURAM Report (in Hungarian language), 2012 December, 312 p.
Kumar R, Choudhury D, Bhargava K (2016) Determination of blast-induced ground vibration equations for rock
using mechanical and geological properties. Journal of Rock Mechanics and Geotechnical Engineering,
8 (2016): 341-349. https://doi.org/10.1016/j.jrmge.2015.10.009.
Kutter H.K., Fairhurst C. (1971): On the fracture process in blasting. Int. J. Rock Mechanics Mining Sci.
Geomechanics Abstracts 8 (3), 181–202.
Li HB, Xia X, Li JC, Zhao J, Liu B, Liu YQ (2011) Rock damage control in bedrock blasting excavation for a
nuclear power plant. Int J Rock Mech Min Sci 48(2):210–218.
https://doi.org/10.1016/j.ijrmms.2010.11.016.
Lu W, Yang J, Chen M, Zhou C (2011) An equivalent method for blasting vibration simulation. Simulation
Modelling Practice and Theory, 19(9): 2051-2062. https://doi.org/10.1016/j.simpat.2011.05.012.
Miklowitz J (1960) Plane-stress unloading waves emanating from a suddenly punched hole in a stretched elastic
plate. J Appl Mech 27(1):165–171. https://doi.org/10.1115/1.3643892.
Park D, Ahn J-K (2015) Estimation of Attenuation Relationship Compatible with Damping Ratio of Rock Mass
from Numerical Simulation. Journal of the Korean Geotechnical Society, vol. 3: 23-26.
https://doi.org/10.7843/kgs.2015.31.4.45.
Park D, Jeon B, Jeon S (2009) A numerical study on the screening of blast-induced waves for reducing ground
vibration. Rock Mech Rock Eng 42(3):449–473. https://doi.org/10.1007/s00603-008-0016-y.
Perras MA, Diederichs MS, Lam T, Kwok C, Armstrong R, Henderson J (2010) A review of excavation damage
zone in sedimentary rocks with emphasis on numerical modelling for EDZ definition. In: Proceedings of
the 63rd Canadian Geotechnical Conference, Calgary, Canada; 2010: 742-750.
Perras MA, Langford C, Ghazvinian E, Diederichs MS (2012) Numerical delineation of the excavation damage
zones: from rock properties to statistical distribution of the dimensions. In: Proceedings of Eurock2012,
Stockholm, Sweden.
Perras MA, Diederichs MS (2016) Predicting excavation damage zone depths in brittle rocks. Journal of Rock
Mechanics and Geotechnical Engineering, 8(1): 60-74. https://doi.org/10.1016/j.jrmge.2015.11.004.
Yang JH, Yao C, Jiang QH, Lu WB, Jiang SH (2017) 2D numerical analysis of rock damage induced by
dynamic in-situ stress redistribution and blast loading in underground blasting excavation. Tunnelling
and Underground Space Technology 70(2017): 221-232. https://doi.org/10.1016/j.tust.2017.08.007.
Yang JH, Lu WB, Jiang QH, Yao C, Jiang SH, Tian L (2016) A study on the vibration frequency of blasting
excavation in highly stressed rock masses. RockMech. Rock Eng. 49 (7): 2825–2843.
https://doi.org/10.1007/s00603-016-0964-6.
Yilmaz O, Unlu T (2013) Three dimensional numerical rock damage analysis under blasting load. Tunn.
Undergr. Space Technol. 38: 266–278. https://doi.org/10.1016/j.tust.2013.07.007.
