Tectonophysics 637 (2014) 198–206

Contents lists available at ScienceDirect

Tectonophysics
journal homepage: www.elsevier.com/locate/tecto

A combined complex electrical impedance and acoustic emission study in

limestone samples under uniaxial loading
V. Saltas ⁎, I. Fitilis, F. Vallianatos
Laboratory of Geophysics and Seismology, Department of Environmental & Natural Resources Engineering, School of Applied Sciences, Technological Educational Institute of Crete, Greece

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

Article history: In the present work, complex electrical impedance measurements in the frequency range of 10 mHz to 1 MHz
Received 31 March 2014 were carried out in conjunction with acoustic emission monitoring in limestone samples subjected to linear
Received in revised form 3 October 2014 and stepped-like uniaxial loading, up to ultimate failure. Cole–Cole plots of the complex impedance during the
Accepted 8 October 2014
stepped loading of limestone have been used to discriminate the contributions of grains interior, grain bound-
Available online 18 October 2014
aries and electrode polarization effects to the overall electrical behavior. The latter is well-described with an
Keywords:
equivalent-circuit model which comprises components of constant phase elements and resistances in parallel
Electrical impedance spectroscopy connection. Electrical conductivity increases upon uniaxial loading giving rise to negative values of effective
Acoustic emissions activation volume. This is a strong experimental evidence for the generation of transient electric signals recorded
Electrical conductivity prior to seismic events and may be attributed to charge transfer (proton conduction) due to cracks generation
Pressure and propagation as a result of the applied stress. The time-series of ac-conductivity at two distinct frequencies
Limestone (10 kHz, 200 kHz) during linear loading of limestone samples exhibits a strong correlation with the acoustic
emission activity obeying the same general self-similar law for critical phenomena that has been reported for
the energy release before materials fracture.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction
The application of mechanical stress in rocks and in brittle materials
in general may be accompanied by various fractoemission phenomena
such as emission of charged particles, atoms, molecules, electromagnetic
radiation (including infrared radiation) and acoustic emission (AE) activ-
ity (Cress et al., 1987; Dickinson et al., 1981, 1991; Freund, 2002; Frid
et al., 2003; Lavrov, 2005; Lockner, 1993; Mori et al., 2009; Yoshida and
Ogawa, 2004). In each case, the underlying physical mechanisms which
are responsible for the occurrence of one or more of these fracture-
induced processes have been widely investigated, mainly due to the
fact that the plethora of all these phenomena constitutes the basis of
searching precursory signals in mechanical damage as well as in earth-
quake prediction (Bleier et al., 2010; Eberhardt et al., 1999; Freund
et al., 2006; Park, 1997; Turcotte et al., 2003; Varotsos, 2005).
For example, transient weak electric currents (known better as
pressure stimulated currents or PSCs) which have been experimentally
recorded before and during the failure of uniaxially-loaded rock speci-
mens have also been observed in field measurements and discussed in
theoretical models, i.e., seismic electric signals (SES) generated before
the occurrence of seismic events (Stavrakas et al., 2004; Vallianatos
⁎ Corresponding author at: 3 Romanou St, PO Box 89, KDS Chania GR 73 135, Greece.
Tel.: +30 28210 23061; fax: +30 28210 23003.
E-mail address: vsaltas@chania.teicrete.gr (V. Saltas).
et al., 2004; Varotsos and Alexopoulos, 1984a,b, 1987; Varotsos and
Lazaridou, 1991; Varotsos et al., 1986; Varotsos et al., 1991). Various
models have been proposed to explain the generation of these weak
electric signals in certain cases, including the piezoelectric effect in
rocks containing piezoelectric minerals, electrokinetic phenomena in
water-saturated rocks, the motion of charged edge dislocations (MCD
model), the presence of point defects including positive holes (defect
electrons in the O2− sublattice of silicate minerals), etc. (Enomoto,
2012; Freund, 2002; Vallianatos and Tzanis, 1998; Vallianatos et al.,
2004; Varotsos, 2005; Yoshida et al., 1998).
The electric transport which is responsible for the recorded transient
electric signals in geomaterials may be investigated by means of electri-
cal impedance spectroscopy (EIS) since the application of an ac-electric
field lead the pressure-induced charges and thus will cause a permanent
variation to the measured electrical properties. In such a way, the elec-
trical and dielectric properties may be analyzed in terms of the physical
parameters related to the damage evolution in the material under
investigation during its mechanical stress (Glover et al., 1997, 2000;
Gómez et al., 1997; Nover et al., 2000).
To the best of our knowledge, detailed studies on the frequency-
dependent electrical properties in conjunction with acoustic emissions
during the application of mechanical stress in materials are rather limit-
ed and are restricted only to the investigation of phase transitions at
high temperatures in relaxor ferroelectric compounds (Dul'kin et al.,
2009, 2011).

http://dx.doi.org/10.1016/j.tecto.2014.10.011
0040-1951/© 2014 Elsevier B.V. All rights reserved.
 V. Saltas et al. / Tectonophysics 637 (2014) 198–206 199

So, in the present work, measurements of complex electrical imped-

ance were carried out simultaneously with AE records during the uniax-
ial mechanical loading of limestone samples and different formalisms of
representation were investigated in order to reveal any possible interre-
lations between acoustic parameters and the electrical and mechanical
properties of samples.

2. Experimental setup

A light brown-beige limestone sample (polycrystalline CaCO3

mainly with crystal forms of calcite and aragonite) of Mesozoic age orig-
inated from Ioannina, Greece, was supplied from a marble company and
was sawed in disks of cylindrical shape (12–15 mm height and 33 mm
diameter). This length-to-diameter ratio of the limestone samples is
inevitable in order to compromise the requirement of a homogenous
electrical field in the sample for ac-electrical measurements with the
capacitance method and the necessity of enough adjacent space to
mount properly the AE sensors and the strain gauge (Jonscher, 1996;
Papathanassiou et al., 2010, 2011). In this way, AE activity which can
be recorded for any sample geometry cannot be correlated with pub-
lished data in the field of rock mechanics where length/diameter ratios
from 2:1 to 3:1 are used, but can be further correlated in perspective to
the measured electrical measurements. The samples dried in an oven at
105 °C for sufficiently long time in order to remove free water, until no
weight loss could be measured and the electrical impedance spectrum
remained unchanged. The high density limestone (2662 kg/m3 in
dried condition) has a Young's modulus of 37 GPa, its porosity is 1.1%
and its water content at saturation is 0.42% (values provided to supplier
by IGME, www.igme.gr/grlith.htm). The measurement of the longitudi-
nal ultrasonic velocity of the samples was measured with a pair of AE
sensors in pulse-receiver mode and gave an average value of 4.65 km/s.
The experimental setup is shown schematically in Fig. 1. The speci-
men was mounted between two parallel stainless-steel electrodes
forming a cylindrical capacitor. Conductive silver paint was applied at
both sides of the cylindrical sample in order to achieve a stable electrical Fig. 1. (a) Schematic representation of the experimental setup used for the simultaneous
surface contact with the stainless-steel electrodes during the uniaxial measurement of complex electrical impedance and acoustic emissions during the uniaxial
loading of limestone samples. (b) Photograph of the sample inside the load frame which is
loading. The electrodes were electrically isolated from the load platens
covered by copper sheets for EM shielding.
by thin Teflon plates (Stavrakas et al., 2004 and references there in). A
miniature strain gauge (1 mm length) was mounted on the lateral sur-
face of the specimen and a load cell was placed at the bottom side of the the sample (103–1011 Ω). Proper electromagnetic shielding consisting
capacitor for continuous stress–strain monitoring. Signals from both of a Faraday cage with copper sheets was implemented to the whole
load cell and strain gauge were recorded with a signal-conditioning apparatus, in order to diminish noise problems which are common
module (SCXI-1520 with SCXI-1314 terminal block, National Instru- at low frequencies (Fig. 1b). The WinDeta and WinFit software of
ments) and a DAQ card (PCI-6221, National Instruments), through an Novocontrol Technologies were used for acquisition and modeling of
appropriate program developed in LabVIEW graphical programming the dielectric experimental data.
environment. Assuming a parallel connection of the capacitance C(ω) and the re-
Acoustic emissions were detected through 4 miniature piezoelectric sistance R(ω) which are the output values of the dielectric analyzer,
sensors (PICO sensors, 200 kHz–1 MHz, MISTRAS Group, SA) mounted the complex impedance Z*(ω), the complex conductivity σ*(ω) and
on the sample's lateral surface and recorded in an integrated multi- the dielectric permittivity ε*(ω) are interrelated through the following
channel system of Physical Acoustics Corporation. A pre-amplification relations:
of 40 dB was used in each channel and the sampling frequency of signals
was 5MSPS. The threshold of detection was determined with the load 1 1
¼ þ jω  C ðωÞ ð1Þ
machine turned on and the specimen in contact with the load platens Z
ðωÞ RðωÞ
at a minimum load of 2–3 kN and thereby was settled at 40 dB in
order to eliminate the background noise. The AE activity is represented

0 C ðωÞ 1 j
in time series of detected signals (hits), events (single count for hits ε ðωÞ ¼ ε − jε″ ¼ −j ¼− ð2Þ
Co ω  C 0  RðωÞ ω  C 0  Z
ðωÞ
detected to different sensors into a predefined time window from a
hit) and amplitude (signal peak in dB).
Uniaxial stress was applied by a 3000 kN loading machine (ALPHA and
S-3000, Form + Test GmbH) equipped with a digital controller  
(Digimaxx-21) for the automated control of the servo-hydraulic system.
0

 εo C ðωÞ
σ ¼ σ − jσ ″ ¼ jωεo ε −1 ¼ − jωεo −1 ð3Þ
Complex conductivity measurements were carried out by means of a C o RðωÞ Co
high-resolution broadband spectrometer (Novocontrol Alpha-N Ana-
lyzer). The frequency of the applied ac electric field was varied between where C0 = εo ⋅ π ⋅ r2/d is the capacitance of the empty cylindrical capac-
10−2 Hz and 106 Hz. At this frequency range, the accuracy of impedance itor, with distance d between the electrodes and r their radius, ω = 2πf
measurements is 0.2–1%, depending on the measured resistance of is the angular frequency and εo is the permittivity of the vacuum.
200 V. Saltas et al. / Tectonophysics 637 (2014) 198–206

The reciprocal permittivity or electric modulus M⁎ representation

can be used alternatively to describe relaxation processes. It was first 8
introduced by MćCrum et al. (1967) and has been applied successfully
in cases where interfacial polarization causes large variations in dielec-
tric permittivity at the low frequency range (Molak et al., 2005; 6 uniaxial stress (MPa)

[ x10 Ω ]
Tsangaris et al., 1998; Wübbenhorst and Turnhout, 2002). Electric 6 82

7
modulus M⁎, which is an electrical analogue to the mechanical shear 12 94
4 24 106
modulus, is defined as:
35 118
47 129

1 0 ″ ε0 ε″ 59 141
M ðωÞ ¼
¼ M ðωÞ þ jM ðωÞ ¼ þj 2
ε ðω
2Þ ∞
02
ε3 þ ε ″ 2
ε þ ε″ 2
02 71 153
Z  
¼ M∞ 41− e
− jωt d ϕðt Þ
d t5 ð4Þ
(a)
dt 0
0

1.5
where Φ(t) describes the decay of the electric field within the sample

- Z'' [ x10 Ω ]
and M∞ = 1 / ε′∞ is the inverse of dielectric permittivity at high

7
1.0

frequencies.
The dried limestone samples were subjected to two different types 0.5

Z" [ x10 Ω ]
of uniaxial mechanical loading. In the first set of experiments, a f

7
0.0
stepped-like loading with a step height of 10 kN (11.5 MPa) was applied 1 0 2 4 6
7
to the limestone sample and the isobaric spectrum of dielectric response Z' [ x10 Ω ]
was recorded in each step, in conjunction with continuous monitoring
of the AE activity. Afterwards, different samples of the same geometry
were subjected to linear loading and time-series of AE activity were
recorded simultaneously with the ac-conductivity at two selected
frequencies (10 kHz and 200 kHz). The experiments were carried out (b)
several times to ensure the reproducibility of the results and the validity 0
of the derived correlations between electrical properties and acoustic 10
-2
10
-1
10
0
10
1
10
2
10
3
10
4
10
5
10
6

emissions.
frequency [Hz]
3. Experimental results and analysis
Fig. 2. (a) Real and (b) imaginary part of complex impedance as a function of frequency
during the uniaxial loading. Cole–Cole plots at a few representative loadings are also included.
3.1. Stepped-loading experiments The equivalent circuit is shown in the inset. Solid lines represent simulated data.

3.1.1. Complex impedance representation

Fig. 2 shows the real (Z′) and the imaginary (Z″) part of the complex
electrical impedance as a function of frequency, during the sequential circuit. Therefore, the total impedance Z of the system may be expressed
loading steps. The last spectrum at 153 MPa corresponds to the totally as:
fractured sample. Some representative Cole–Cole plots of the electrical
impedance (− Z″ versus Z′) are also included in the inset of Fig. 2b.
Each Cole–Cole graph comprises two superimposed depressed semicir- Z ¼ Z R1 jjCPE1 þ Z R2 jjCPE2 þ Z CPE3 ð5Þ
cles and a tail at the low frequency range. Furthermore, the impedance
values are reduced with increasing applied stress but the shape of the
where Z Ri jjCPEi denotes the impedance of CPEi in parallel connection
curves remains almost unchanged over the whole stress range except
with the resistance Ri (i = 1 for grains interior and i = 2 for grain
the shift of Z″ to higher frequency. The main complex impedance arc
boundaries).
in the Cole–Cole plot observed at high frequencies denotes the intrinsic
In general, the impedance of a CPE is given by the following complex
contribution of the grain interior of the limestone specimen while, the
equation (Huebner and Dillenburg, 1995):
arc at medium frequencies is due to the contribution of the grain bound-
aries. The low-frequency tail at the right side of the plots is due to the
undesirable result of the electrodes response, limiting the effective 1 1 h  π  πi
frequency range of impedance spectroscopy to about 5 × 10−2 Hz. Z CPE ¼ ¼ cos −n þ j sin −n ð6Þ
Q ð jωÞn Qωn 2 2
This discrimination of the aforementioned contributions to the electrical
conductivity is quite common in different kinds of materials such
as ionic glasses, disordered materials, and ion-conducting polymers where Q is a factor of proportionality taking numerical values and
(Dyre and Schrøder, 2000; Henn et al., 2007; Macdonald, 2005). the exponent n defines the phase shift and may vary from 0 for pure
In the present case, the contributions of grain interior and grain resistive to + 1 for pure capacitive behavior. Separating the real and
boundaries may be modeled via an equivalent circuit of two compo- imaginary part of the complex total impedance Z in Eq. (5), we finally
nents in series, where each one consists of a constant phase element obtain (Saltas et al., 2013):
(CPE) in parallel connection with a resistance R. Such kind of equivalent
circuit models have been used to describe the electrical behavior of  π
silicate minerals and rocks in general as a function of temperature, but X
2 Ri þ Q i R2i ωni cos ni
Z RE ¼  π 2
are also applicable in the case of pressure-induced variation of electrical
i¼1 1 þ 2Q i Ri ω i cos ni þ Q i 2 R2i ω2ni
n
properties (Huebner and Dillenburg, 1995; Li et al., 2013; Saltas et al., 2
1  π
2013). The low-frequency tail indicates the presence of another CPE þ n3 cos −n3 ð7aÞ
connected in series with the previous components of the equivalent Q 3ω 2
 V. Saltas et al. / Tectonophysics 637 (2014) 198–206 201

and

dc-conductivity, σ dc [S / cm]
-7
grains interior 10
π 
X2 Q i R2i ωni
sin ni -6
Z Im ¼−  π 2 10
i¼1 1 þ 2Q i Ri ω i cos ni þ Q i 2 R2i ω2ni
n 10
-8

  2
1 π grain boundaries
þ sin −n3 ð7bÞ
Q 3 ω n3 2 0 20 40 60 80 100 120 140 160

ac-conductivity, σ
-7 uniaxial stress [MPa]
10

where ω = 2πf is the angular frequency and Qi and ni refer to CPEi

uniaxial stress (MPa)
parameters. These parameters were estimated for each isobaric imped- 0 47 106
ance spectrum during the loading cycle and are plotted in Fig. 3. Obvi- 6 59 118
ously, the corresponding fitting curves (solid lines in Fig. 2) exhibit -8
10 12 71 129
very good correlation with the experimental data. We have to point 24 82 141
35 94 153
out that the effect of mechanical loading is more evident in the grain
boundaries where the electrical resistance varies more drastically than 10
-3
10
-2
10
-1
10
0
10
1
10
2
10
3
10
4
10
5
10
6

in grains interior which exhibits a small variation up to the fracture.

Furthermore, grain boundaries exhibit a more resistive behavior after
60 MPa in contrast to grain interior, as it is indicated from the variation
of the n parameter. This stress value corresponds to the departure of
the linear response of the sample, according to the stress–strain
measurements.
The charge transport mechanisms related to electrical behavior of
the limestone sample can be further analyzed in terms of the complex
conductivity function σ* and the electric modulus representation M*.
3.1.2. Complex conductivity representation
The real part of ac-conductivity σ′ac as a function of frequency during
the stepped loading of limestone sample up to the fracture is depicted
in Fig. 4. The surprising finding is the increase of the electrical conduc-
tivity upon pressure but apart from this, the recorded spectra exhibit a
typical behavior of σ′ac in a hopping model of charge carriers observed
8 ly modified by the geometrical factor and are illustrated in the inset
R1, Q1, n1
R [ x10 Ω ]
frequency [Hz]
Fig. 4. Real part of ac-conductivity variation (σac′) as a function of frequency during the
stepped-loading of the limestone samples. The inset demonstrates the dc-conductivity
of grains interior and grain boundaries contributions as a function of the applied stress.
in disordered materials (Dyre and Schrøder, 2000). Specifically, at
high frequencies a linear increase of ac-conductivity with increasing
frequency is recorded, namely, polarization conductivity due to the
localized hopping motion of charge carriers. At the low frequency
range the dc-plateau is well-developed enabling the estimation of
the corresponding dc-conductivity values. An abrupt step observed in
some cases (for example at 71 MPa) at the low frequency region of
the recorded spectra is related to high AE activity (i.e. significant crack
generation) during the acquisition process. The discrete contributions
of grain boundaries and grain interior to the total dc-conductivity σdc
have been estimated from the corresponding impedance values proper-
of Fig. 4. Considerable changes in each case are observed from 60 MPa

6
up to 100 MPa, where departure from linearity is observed in the
7

R2, Q2, n2 stress–strain curve, while out of this range dc-conductivity varies
4
more smoothly.
2 In the interpretation of conductivity and/or diffusivity measure-
ments in different kinds of materials, an important parameter is the
0 activation volume υact which describes the volume change in the
host material due to the formation of point defects and the lattice
-8
Q parameter distortion caused by their migration and is defined as the pressure
10
derivative of the activation Gibbs free energy which governs the trans-
port process (Papathanassiou and Grammatikakis, 1996; Varotsos and
10
-9 Alexopoulos, 1986). Positive values of υact indicate the lattice expansion
during charge transferring between two neighboring sites, while nega-
tive values which are less common, denote that the activated state
-10
10 where charge is transferred has a smaller volume than its ground state
(Fontanella et al., 1996). The sign of activation volume is a crucial
parameter for the theoretical justification of the generation mechanism
0.8
n parameter of seismic electric signals (SES) based on piezo-stimulated currents
(Varotsos, 2005; Varotsos and Alexopoulos, 1986; Varotsos and
Lazaridou, 1991; Varotsos et al., 1986). Negative activation volume
0.6 reveals a process in which the activation energy is reduced as pressure
(or applied stress) increases. Taking into account that in the case
of a non-hydrostatic pressure, stress states are characterized by
0.4 a stress tensor
(σ  ij), we may consider an effective hydrostatic pressure
P eff ¼ − 13 Tr σ i j and a corresponding effective activation volume, υacteff
0 20 40 60 80 100 120 140 160 (Hinterberg et al., 2013). More generally however, one can express
uniaxial stress [MPa] the influence of stress (σij) on conductivity in terms of activation vol-
ume tensor introduced by Aziz (1997) and applied to diffusion related
Fig. 3. Fitting parameters Ri, Qi, and ni according to the equivalent circuit which is repre- phenomena (Aziz et al., 2006). To a rough approximation for a uniaxial
sented by Eqs. (7a)–(7b). load where σ11 ≠ 0 and σij = 0 for i, j ≠ 1, the effective activation
202 V. Saltas et al. / Tectonophysics 637 (2014) 198–206

volume υact
eff may be obtained from the stress derivative of the dc-
conductivity according to the following relation:

act ∂ln σ dc
υeff ≈ −kB T
∂σ 11 T
where kB is the Boltzman's constant.
In each of the three regions of the applied pressure where the dc-
conductivity exhibits different variation, namely, at P b 60 MPa, between
60 and 100 MPa and at P N 100 MPa, values of υact eff were estimated to
be − 10 ± 1 cm3/mol, − 61 ± 3 cm3/mol and − 9 ± 1 cm3/mol,
respectively.
The above findings are in contradiction with the positive values
of activation volume that have been measured in geomaterials such as
alkaline earth carbonate salts (dolomite, magnesite, calcite) and silicate
materials (sandstone, pelite, olivine) (Papathanassiou, 1998, 2001,
2002; Papathanassiou and Grammatikakis, 1996; Xu et al., 2000). How-
ever, the increase of the electrical conductivity with pressure has been
occasionally observed in the case of hydrated rocks such as leukolite,
limestone, granodiorite and amphibolites from the KTB drilling, with
values of activation volume ranging from − 0.8 ± 0.1 to − 60 ±
15 cm3/mol (Papathanassiou et al., 2010, 2011, 2012; Sakellis et al.,
2014; Shankland et al., 1997). The last reported value for hydrate
leukolite is the same with our maximum estimated effective value.
Although the limestone samples were dried at 105 °C for a long time,
obviously, this temperature is not sufficient to remove bound water
located in grain boundaries. This strongly bound water might dissociate
upon pressure giving rise to enhanced proton conduction (Fontanella
et al., 1996). Conductivity measurements that were carried out in
dried polycrystalline calcite samples under vacuum, resulted in low
activation energy (b0.5 eV) during the initial stages of heating (Saltas
and Vallianatos, in preparation), suggesting the proton motion as a
possible conduction mechanism which is further enhanced due to the
dissociation of the bound water upon pressure.
3.1.3. Dielectric modulus representation
The imaginary part of the dielectric modulus M″ as a function of
frequency during the sequential loading steps of limestone sample up
to the fracture is depicted in Fig. 5. A broad relaxation peak is observed
which is gradually shifted to higher frequencies as uniaxial loading
increases indicating that the related relaxation mechanism becomes
faster. Furthermore, it is evident that the shift in frequency of the peak
is not uniform.
uniaxial stress (MPa)
0 47 106
15 6 59 118
12 71 129
24 82 141
Modulus [x 10 ]
The validity of the time-pressure superposition principle may be
checked by plotting the normalized dielectric modulus M″/M″max ver-
sus the normalized frequency f/fmax (Jonscher, 1996). The resulting
master plot (Fig. 6a) reveals that the spectral shape is independent
ð8Þ
of pressure over the whole pressure range suggesting that the basic
conduction mechanism of the dynamic process is stress independent,
regardless of the stress dependence of the frequency window of this
process. The full width at half-maximum (FWHM) of the master curve
is more than 2 orders of log units suggesting a broad distribution of
relaxation times, as compared to the 1.14 units of the simplified
Debye model with a single relaxation mechanism.
The master curve of the imaginary part of the dielectric modulus M″
may be well fitted by a single-term Havriliak–Negami (HN) dielectric
relaxation function, according to the following expression:

ΔM
M ðωÞ ¼ þ M∞ ð9Þ
ð1 þ ð jωτM Þα Þβ
where τM is the characteristic relaxation time and ΔM = Ms − M∞ is
the relaxation strength, with Ms, M∞ the values of M′ at zero and infinite
frequency, respectively. The shape parameters α and β describe the
broadness and the asymmetry of the relaxation peaks, respectively.
The estimated parameter values have been plotted in Fig. 6b–d as a
function of stress, up to the ultimate failure of the sample. A small
increase of ΔΜ with stress implies an increase of the available charge
carriers which could be protons generated from the dissociation of
bound water upon pressure, as suggested previously in the presentation
of the dc-conductivity results (see Fig. 4). Furthermore, the relaxation
time τΜ decreases with stress, implying faster conduction processes,
which may be explained by an increase in the mobility of charge carriers
upon pressure.
3.1.4. Correlation of complex impedance measurements with the acoustic
emission activity
The acoustic emission activity was monitored continuously during
the loading steps enabling the correlation of the recorded impedance
spectra with the damage evolution in the limestone sample. To the
best of our knowledge, such a simultaneous monitoring has not been
investigated so far and the effort has been focused in the dielectric prop-
erties of geomaterials under hydrostatic pressure (Papathanassiou et al.,
2010, 2011, 2012). Thus, an important parameter that has not yet been
taken into consideration although it should play a significant role in the
macro-scale of seismic activity is the generation and propagation of
cracks upon pressure in different dimension scales (micro- and meso-
scale). The evolution of cracks seems to contribute significantly to trans-
port properties, in conjunction with the dominant role of bound water
into the rock samples.
We have to note that each impedance spectrum was recorded only
after the hit rate in each loading step had been decreased considerably,
in order to avoid unstable conditions of the sample and thus fluctuations
to the measured conductivity values. However, the latter situation in

35 94 153
-3

certain cases was inevitable (see for example the lower frequency
10 region of the conductivity spectrum at 71 MPa in Fig. 4). The time evo-
lution of the hit rate and the corresponding distribution of amplitudes
are demonstrated as a 3D graph in Fig. 7. The cumulative hits and the
stress steps during uniaxial loading are also shown in the back plane
5 of Fig. 7. Microcracks are generated upon uniaxial loading as indicated
by the AE activity without leading in an expeditious failure of the
sample due to its special geometry which permits significant deforma-
tion before the ultimate failure. Actually, due to the low length-to-
0 diameter ratio of the tested samples, the condition of uniaxial stress is
10
1
10
2
10
3
10
4
10
5
10
6 not fulfilled as in the cases of 2:1 to 3:1 ratios, and the stress distribution
frequency [Hz] inside the limestone samples is not uniform (Singh and Ghose, 2006).
Obvious acoustic activity starts at the step of 35 MPa while a distinct
Fig. 5. Imaginary part of the dielectric modulus (M″) as a function of frequency during the behavior is observed in each of the following loading steps, up to the
stepped-loading of the limestone samples. final rupture of the sample. In other words, considerable hit rate is
 V. Saltas et al. / Tectonophysics 637 (2014) 198–206 203

4
-5
(a) τ , x10 s
1.0
(b) τM 3

2
τ max
0.8 1

-3
ΔM , x10
(c)
M'' / M'' max
0.6 70

65
0.4

(d) α
0.2 0.8

HN fitting curve
β 0.4
0.0
-3 -2 -1 0 1 2 3
10 10 10 10 10 10 10 0 20 40 60 80 100 120 140 160
f / f max uniaxial stress (MPa)

Fig. 6. (a) Normalized plots of M″/M″max versus log(f/fmax) of all the measured curves (master curve), where the modulus has been scaled to its maximum value M″max and the frequency
axis has been scaled to the frequency of the maximum in M″. The symbols denote the experimental data points of all curves, while the line represents the HN fitting curve of the master
curve. (b–d) Plots of the fitting parameters as a function of uniaxial stress during the stepped-loading.

observed during the initial application of each loading step (after
35 MPa) which however decreases with time, until the next loading
step. The amplitude of the acoustic hits also follows a monotonic de-
crease during the application of each loading step while, the majority
of the hits have amplitudes below 60 dB. A remarkable hit rate with a
fast decay is always observed during the period of each loading step
(~15 min) suggesting that microcracks are generated and propagated
not only during the stress increment but also during the constant stress
state of the sample in each loading step. The latter is attributed to the
fast loading rate in each step and as a consequence, the limestone
sample does not have enough time to deform and the deformation is de-
layed during the constant stress state. Furthermore, the highest acoustic
activity (more than 60%) is recorded from 71 to 106 MPa, suggesting
that most of the cracks are generated at this stress range (region II in
Fig. 7). Below 71 MPa (region I) AE is significant lower while, above
106 MPa (region III) the acoustic activity is also reduced. Obviously,
the large scale cracks (related to a high AE activity) which are observed
in the vertical direction with a size similar to the small height of the
sample cannot destroy the structural stability of the sample and at the
sequential ultimate failure the AE activity appears diminished. It has
been reported for brittle deformation and associated AE activity in
rocks that the rate of AE becomes much higher as stress increases
and the sample approaches the macroscopic failure, as a result of the
cracks interaction and cracks coalescence (Amitrano, 2003; Baud and

Fig. 7. 3D-plot of the acoustic emissions (hit rate and amplitude vs. time) during the stepped-loading of limestone sample. The pressure steps and the cumulative hit counts during the
loading are also shown at the back plane. In region I, low hit rate is recorded and is ascribed to cracks closure and the generation of new micro-cracks. In region II with the highest hit
rate, cracks are generated and propagated through the material while, in region III, the hit rate has been decreased in a partially damaged specimen.
204 V. Saltas et al. / Tectonophysics 637 (2014) 198–206

Meredith, 1997; Wassermann et al., 2009). In our case however, we did uniaxial stress [MPa]
not observe such acoustic evolution because brittle deformation is rath- 0 20 40 60 80 100 120
10
er discontinuous and inhomogeneous due to the sample's special geom-

normalized stress
0.8
etry, as it has been mentioned previously. A negligible AE activity below
0.6
35 MPa should be related to cracks closure of the limestone sample. The -6 8
10 0.4
initiation of considerable AE activity at high stress (N 35 MPa) could be
explained to some extent by the low porosity of the limestone samples
0.2
200 kHz 6
0.0
in perspective with high porosity materials where acoustic emissions 0.0
strain
2.0x10-4

start at lower stresses. Another important parameter that could also 4

affect the inception of the AE activity is the compressive strength that
depends on the shape of the specimen and increases as the length-to-

cumulative events [ x10 ]

4
2
diameter ratio decreases, resulting in a time-shifted acoustic activity

ac-conductivity [S / cm]
10 kHz sample A
(Tuncay and Hasançebi, 2009). -7
10 0
The separation of acoustic activity in these three stress regions 0 100 200 300
(Fig. 7) is in accordance with the corresponding variation of dc-
conductivity (Fig. 4). This observation leads us to investigate further -6
10
the correlation of AE activity with electrical conductivity and per-
4
form supplementary measurements of continuous monitoring of
ac-conductivity and acoustic emissions during the linear loading of
limestone, as it will be described in the next section. The generation 200 kHz
of cracks in the microscale regime, possibly in grain boundaries, may
create effective surface dipoles between the grains, i.e. conduction 2
pathways that enhance the transport of charges. The cracks evolu-
tion results to their interconnection in a greater scale (meso- or
macro-scale) according to percolation theory and finally to the ulti-
mate failure of the limestone sample (Guéguen et al., 1997). 10 kHz sample B
-7
10 0
3.2. Time-series of ac-conductivity and acoustic emissions during linear 0 100 200 300
loading time [s]

At the second part of loading experiments the time-series of acoustic
activity are investigated in conjunction with the variation of ac-
conductivity at two selected frequencies (10 kHz, 200 kHz) during a
linear uniaxial loading of limestone samples, in order to ascertain any
possible similarities of these two responses due to the dynamical pro-
cess of stress application.
The time-series of the cumulative events of acoustic activity and the
recorded ac-conductivity values, during the linear loading (0.3 kN/s) of
two limestone samples of the same geometry are depicted in Fig. 8. Part
of the stress–strain curve for sample A is also included in the inset of
Fig. 8. We have to note that, due to the small length-to-diameter ratio
of the sample and the miniature size of the strain gauge, the localization
of deformations caused by the creation of micro-cracks renders the
strain gauge unreliable long before the peak stress. For both samples,
ac-conductivity at each measured frequency increases with linear load-
ing, following a similar variation with that of the cumulative events. In
other words, a sudden or a gradual increase of acoustic activity is always
accompanied by an analogous increase of ac-conductivity, which is
more evident at the low measured frequency of 10 kHz. The small in-
crease of AE and ac-conductivity at the beginning of the load is ascribed
to closing of cracks, according to the non-linear behavior of axial strain
during the first stage of uniaxial compression (see inset of Fig. 8).
The observed correlation of AE activity and electrical ac-conductivity
suggests that the cracks generation and propagation in limestone is
strongly associated with the conduction mechanism which gives rise
either to excess charge carriers or their increased mobility, resulting in
the increase of conductivity during the loading.
Moura et al. (2005, 2006) have proposed a general law of energy
release during the mechanical loading of materials up to the critical
point of fracture, based on the self-similar approximation theory
(Yukalov et al., 2004). This law which has been applied successfully
to various stressed composite materials is described via the following
expression:
α 
α  a strong correlation of the generated cracks with the density of charge
f ðxÞ ¼ cos cx sin gðxÞexp hðxÞ exp cx cos gðxÞexp hðxÞ ð10Þ
Fig. 8. Time-series of AE activity (cumulative events) and ac-conductivity at 10 kHz and
200 kHz during linear loading of two limestone samples (A, B) of the same geometry
with a stress rate of 0.3 kN/s up to the failure. The normalized stress–strain curve for
sample A is shown in the inset.
where
gðxÞ ¼ a sin ðω ln xÞ þ b cos ðω ln xÞ;
hðxÞ ¼ a cos ðω ln xÞ−b sin ðω ln xÞ:
The variable x is the relative dimensionless time, x = (tC − t)/tC,
(0 ≤ x ≤ 1) where t varies from t = 0 at the starting point of load, up
to the critical time tC when catastrophic failure occurs. The independent
parameters c, α describe the rate of the overall increase of the released
energy while, the parameter ω refers to the log-periodic frequency of
oscillations and a and b to their amplitude.
In the present case, we applied this law to fit the time-series of ac-
conductivity during the linear loading of the limestone sample. The
fitting curve of the normalized ac-conductivity (measured at 10 kHz)
for sample A is depicted in Fig. 9, showing a satisfactory correlation, in
comparison to other reported results on AE and electromagnetic energy
release in composite and granitic samples (Mastrogiannis et al., 2011;
Yukalov et al., 2004).
4. Conclusions
In the present work a combined investigation of complex electrical
conductivity and acoustic emission activity was carried out in limestone
samples under linear and stepped-like uniaxial loading, in order to as-
certain a possible correlation of electrical properties and AE parameters
related to micro-cracks generation and propagation in the geomaterial.
A qualitative finding relates the variation of the electrical conductivity
with the AE events during the gradual loading of the specimen, implying
carriers of the responsible conduction mechanism. Furthermore, a
 V. Saltas et al. / Tectonophysics 637 (2014) 198–206
1.0
a = -1.02 ± 0.08 data points
normalized ac-conductivity (10kHz)
0.8 b = 1.04 ± 0.03 fitting curve
c = -5.21 ± 0.05
0.6 α = 1.60 ± 0.03
ω = 0.90 ± 0.01
0.4
0.2
0.0
0.0 0.2 0.4 0.6 0.8
normalized time (1-x)
Fig. 9. Normalized ac-conductivity at 10 kHz of sample A as a function of time and its
fitting curve (red line) according to Eq. (10) (see text). The fitting parameters and their
errors are also included.
quantitative correlation of ac-conductivity and AE activity has been
found, according to a general self-similar law for critical phenomena
which describes the energy release during materials fracture. The anal-
ysis of the complex impedance spectra in various representations indi-
cates the different contributions of the grain boundaries and intrinsic
mechanisms (grain interior) to the overall electrical behavior of
limestone. Proton conduction related with the dissociation of bound
water upon pressure may be proposed as the predominant conduction
mechanism. The finding of negative effective activation volumes is a
further evidence of analogous reported experimental observations in
various rock-forming minerals which explains the seismic electric sig-
nals recorded prior to earthquake events. However, further investiga-
tion should be focused to the influence of water content to the sign
and values of the activation volume, since, a critical concentration of
bound water might exists where the transition from positive to negative
activation volume takes place.
Acknowledgments
This work was supported by the THALES Program of the Ministry
of Education of Greece and the European Union in the framework of
the project entitled “Integrated understanding of Seismicity, using inno-
vative Methodologies of Fracture mechanics along with Earthquake
and non-extensive statistical physics — Application to the geodynamic
system of the Hellenic Arc. SEISMO FEAR HELLARC” (MIS 380208).
References
Amitrano, D., 2003. Brittle–ductile transition and associated seismicity: experimental and
numerical studies and relationship with the b-value. J. Geophys. Res. 108 (B1), 2044
(15 pp.).
Aziz, M.J., 1997. Thermodynamics of diffusion under pressure and stress: relation to point
defect mechanisms. Appl. Phys. Lett. 70, 2810–2812.
Aziz, M.J., Zhao, Y., Gossmann, Hans-J., Mitha, S., Smith, S.P., Schiferl, D., 2006. Pressure
and stress effects on the diffusion of B and Sb in Si and Si–Ge alloys. Phys. Rev. B 73
(054101-1), 054101–054120.
Baud, P., Meredith, P.G., 1997. Damage accumulation during triaxial creep of darley dale
sandstone from pore volumometry and acoustic emission. Int. J. Rock Mech. Min.
Sci. 34 (3–4), 371.
Bleier, T., Dunson, C., Alvarez, C., Freund, F., Dahlgren, R., 2010. Correlation of pre-
earthquake electromagnetic signals with laboratory and field work experiments.
Nat. Hazards Earth Syst. Sci. 10, 1965–1975.
Cress, G.O., Brady, B.T., Rowell, G.A., 1987. Sources of electromagnetic radiation from
fracture of rock samples in the laboratory. Geophys. Res. Lett. 14, 331–334.
Dickinson, J.T., Donaldson, E.E., Park, M.K., 1981. The emission of electrons and positive
ions from fracture of materials. J. Mater. Sci. 16, 2897–2908.
205
Dickinson, J.T., Jensen, L.C., Langford, S.C., Rosenberg, P.E., Blanchard, D.L., 1991. CO2 emis-
sion accompanying the fracture of calcite. Phys. Chem. Miner. 18, 320–325.
Dul'kin, E., Mojaev, E., Roth, M., Khamman, O., Tan, X., 2009. Acoustic emission and dielectric
studies of phase transitions within the morphotropic phase boundary of xPb(Zr1/2Ti1/2)
O3–(1 − x)Pb(Ni1/3Nb2/3)O3 relaxor ferroelectrics. Appl. Phys. Lett. 95, 252903-3.
Dul'kin, E., Kojima, S., Roth, M., 2011. Dielectric maximum temperature non-monotonic
behavior in unaxial Sr0.75Ba0.25Nb2O6 relaxor seen via acoustic emission. J. Appl.
Phys. 110, 044106.
Dyre, J.C., Schrøder, T.B., 2000. Universality of ac conduction in disordered solids. Rev.
Mod. Phys. 72, 873–892.
Eberhardt, E., Stead, D., Stimpson, B., 1999. Quantifying progressive pre-peak brittle fracture
damage in rock during uniaxial compression. Int. J. Rock Mech. Min. Sci. 36, 361–380.
Enomoto, Y., 2012. Coupled interaction of earthquake nucleation with deep Earth gases: a
possible mechanism for seismo-electromagnetic phenomena. Geophys. J. Int. 191,
1210–1214.
Fontanella, J.J., Edmondson, C.A., Wintersgill, M.C., Wu, Y., GreenBaum, S.G., 1996. High-
pressure electrical conductivity and NMR studies in variable equivalent weight
NAFION membranes. Macromolecules 29, 4944–4951.
Freund, F., 2002. Charge generation and propagation in igneous rocks. J. Geodyn. 33,
543–570.
Freund, F.T., Takeuchi, A., Lau, B.W., 2006. Electric currents streaming out of stressed igneous
1.0 rocks — a step towards understanding pre-earthquake low frequency EM emissions.
Phys. Chem. Earth 31, 389–396.
Frid, V., Rabinovitch, A., Bahat, D., 2003. Fracture induced electromagnetic radiation.
J. Phys. D. Appl. Phys. 36, 1620–1628.
Glover, P.W.J., Gómez, J.B., Meredith, P.G., Hayashi, K., Sammonds, P.R., Murrel, S.A.F.,
1997. Damage of saturated rocks undergoing triaxial deformation using complex
electrical conductivity measurements: experimental results. Phys. Chem. Earth 22,
57–61.
Glover, P.W.J., Gómez, J.B., Meredith, P.G., 2000. Fracturing in saturated rocks undergoing
triaxial deformation using complex electrical conductivity measurements: experi-
mental study. Earth Planet. Sci. Lett. 5621, 201–213.
Gómez, J.B., Glover, P.W.J., Meredith, P.G., Hayashi, 1997. Damage of saturated rocks un-
dergoing triaxial deformation using complex electrical conductivity measurements:
mechanical modeling. Phys. Chem. Earth 22, 63–68.
Guéguen, Y., Chelidze, T., Le Ravalec, M., 1997. Microstructures, percolation thresholds,
and rock physical properties. Tectonophysics 279, 23–35.
Henn, F., Durand, C., Cerepi, A., Brosse, E., Giuntini, J.C., 2007. DC conductivity, cationic
exchange capacity and specific surface area related to chemical composition of pore
lining chlorites. J. Colloid Interface Sci. 311, 571–578.
Hinterberg, J., Zacherle, T., De Souza, R.A., 2013. Activation volume tensor for oxygen-
vacancy migration in strained CeO2 electrolytes. Phys. Rev. Lett. 110 (205901-1),
205901–205905.
Huebner, J.S., Dillenburg, R.G., 1995. Impedance spectra of hot, dry silicate minerals and
rock: qualitative interpretation of spectra. Am. Mineral. 80, 46–64.
Jonscher, A.K., 1996. Universal Relaxation Law. Chelsea Dielectrics, London.
Lavrov, A., 2005. Fracture-induced physical phenomena and memory effects in rocks: a
review. Strain 41, 135–149.
Li, K., Zheng, H., Ivanov, I., Guthrie, M., Xiao, Y., Yang, W., Tulk, C., Zhao, Y., Mao, H., 2013.
K3Fe(CN)6: pressure-induced polymerization and enhanced conductivity. J. Phys.
Chem. C 117 (46), 24174–24180.
Lockner, D., 1993. The role of acoustic emission in the study of rock fracture. Int. J. Rock
Mech. Min. Sci. Geomech. Abstr. 30, 883–899.
Macdonald, J.R., 2005. Universality, the Barton Nakajima Namikawa relation, and scaling
for dispersive ionic materials. Phys. Rev. B71 (184307-1-12).
Mastrogiannis, D., Hadjicontis, V., Mavromatou, C., 2011. Experimental evidence of the
cumulative electromagnetic energy release data, with the hierarchical models for
the catastrophic fracturing process. Nat. Hazards Earth Syst. Sci. 11, 1605–1608.
MćCrum, N.G., Read, B.E., Williams, G., 1967. Anelastic and Dielectric Effects in Polymer
Solids. Wiley, London, p. 108.
Molak, A., Paluch, M., Pawlus, S., Klimontko, J., Ujma, Z., Gruszka, I., 2005. Electric modulus
approach to the analysis of electric relaxation in highly conducting (Na0.75Bi0.25)
(Mn0.25Nb0.75)O3. J. Phys. D. Appl. Phys. 38, 1450.
Mori, Y., Obata, Y., Sikula, J., 2009. Acoustic and electromagnetic emission from crack
created in rock sample under deformation. J. Acoust. Emission 27, 157–166.
Moura, A., Lei, X., Nishisawa, O., 2005. Prediction scheme for the catastrophic failure of
highly loaded brittle materials or rocks. J. Mech. Phys. Solids 53, 2435–2455.
Moura, A., Lei, X., Nishisawa, O., 2006. Self-similarity in rock cracking and related complex
critical exponents. J. Mech. Phys. Solids 54, 2544–2553.
Nover, G., Heikamp, S., Freund, D., 2000. Electrical impedance spectroscopy used as a tool
for the detection of fractures in rock samples exposed to either hydrostatic or triaxial
pressure conditions. Nat. Hazards 21, 317–330.
Papathanassiou, A.N., 1998. Effect of hydrostatic pressure on the electrical conductance of
polycrystalline magnesite (MgCO3). Phys. Rev. B 58, 4432–4437.
Papathanassiou, A.N., 2001. Pressure variation of the conductivity in single crystal calcite.
Phys. Status Solidi B 228, R6–R7.
Papathanassiou, A.N., 2002. Dependence of the electrical conductivity and the low-
frequency dielectric constant upon pressure in porous media containing a small
quantity of humidity. Electrochem. Acta 48, 235–239.
Papathanassiou, A.N., Grammatikakis, J., 1996. Pressure variation of the electrical conduc-
tivity of dolomite [CaMg(CO3)2]. Phys. Rev. B53, 16247–16251.
Papathanassiou, A.N., Sakellis, I., Grammatikakis, J., 2010. Negative activation volume for
dielectric relaxation in hydrated rocks. Tectonophysics 490, 307–309.
Papathanassiou, A.N., Sakellis, I., Grammatikakis, J., 2011. Dielectric properties of granodi-
orite partially saturated with water and its correlation to the detection of seismic
electric signals. Tectonophysics 511, 148–151.
206 V. Saltas et al. / Tectonophysics 637 (2014) 198–206
Papathanassiou, A.N., Sakellis, I., Grammatikakis, J., 2012. Dielectric relaxation under pres-
sure in granular dielectrics containing water: compensation rule for the activation
parameters. Solid State Ionics 209–210, 1–4.
Park, S.K., 1997. Electromagnetic precursors to earthquakes: a search for predictors. Sci.
Prog. 80, 65–82.
Sakellis, I., Papathanassiou, A.N., Grammatikakis, J., 2014. Measurements of the dielectric
properties of limestone under pressure and their importance for seismic electric
signals. J. Appl. Geophys. 102, 77–80.
Saltas, V., Vallianatos, F., 2014n. Complex electrical conductivity of polycrystalline calcite
at elevated temperatures, (in preparation).
Saltas, V., Vallianatos, F., Gidarakos, E., 2013. Charge transport in diatomaceous earth
studied by broadband dielectric spectroscopy. Appl. Clay Sci. 80–81, 226–235.
Shankland, T.J., Duba, A.G., Mathez, E.A., Peach, C.L., 1997. Increase of electrical conductiv-
ity with pressure as an indicator of conduction through a solid phase in midcrustal
rocks. J. Geophys. Res. 102, 14741–14750.
Singh, R.N., Ghose, A.K., 2006. Engineering Rock Structures in Mining and Civil Construc-
tion. Taylor & Francis/Balkena, London, UK, p. 35.
Stavrakas, I., Triantis, D., Agioutantis, Z., Maurigiannakis, S., Saltas, V., Vallianatos, F.,
Clarke, M., 2004. Pressure stimulated currents in rocks and their correlation with
mechanical properties. Nat. Hazards Earth Syst. Sci. 4, 563–567.
Tsangaris, G.M., Psarras, G.C., Kouloumbi, N., 1998. Electric modulus and interfacial polar-
ization in composite polymeric systems. J. Mater. Sci. 33, 2027.
Tuncay, E., Hasançebi, N., 2009. The effect of length to diameter ratio of test specimens on
uniaxial compressive strength of rock. Bull. Eng. Geol. Environ. 68 (4), 491–497.
Turcotte, D.L., Newman, W.I., Shcherbakov, R., 2003. Micro and macroscopic models of
rock fracture. Geophys. J. Int. 152, 718–728.
Vallianatos, F., Tzanis, A., 1998. Electric current generation associated with the deforma-
tion rate of a solid: preseismic and coseismic signals. Phys. Chem. Earth 23, 933–938.
Vallianatos, F., Triantis, D., Tzanis, A., Anastasiadis, C., Stavrakas, I., 2004. Electric earth-
quake precursors: from laboratory results to field observations. Phys. Chem. Earth
29, 339–351.
Varotsos, P.A., 2005. The Physics of Seismic Electric Signals. TERRAPUB, Tokyo.
Varotsos, P.A., Alexopoulos, K., 1984a. Physical properties of the variations of the electric
field of the earth preceding earthquakes, I. Tectonophysics 110, 73–98.
Varotsos, P.A., Alexopoulos, K., 1984b. Physical properties of the variations of the electric
field of the earth preceding earthquakes, determination of epicenter and magnitude,
II. Tectonophysics 110, 99–125.
Varotsos, P.A., Alexopoulos, K.D., 1986. Thermodynamics of Point Defects and Their
Relation With Bulk Properties. North-Holland, Amsterdam.
Varotsos, P., Alexopoulos, K., 1987. Physical properties of the variations in the electric field
of the earth preceding earthquakes, III. Tectonophysics 136, 335–339.
Varotsos, P., Lazaridou, M., 1991. Latest aspects of earthquake prediction in Greece based
on seismic electric signals. Tectonophysics 188, 321–347.
Varotsos, P., Alexopoulos, K., Nomicos, K., Lazaridou, M., 1986. Earthquake prediction and
electric signals. Nature 322, 120.
Varotsos, P., Alexopoulos, K., Lazaridou, M., 1991. On recent seismic electric signal activity
in northern Greece. Tectonophysics 188, 403–405.
Wassermann, J., Senfaute, G., Amitrano, D., Homand, F., 2009. Evidence of dilatant and
non-dilatant damage processes in oolitic iron ore: P-wave velocity and acoustic
emission analyses. Geophys. J. Int. 177 (3), 1343–1356.
Wübbenhorst, M., van-Turnhout, J., 2002. Analysis of complex dielectric spectra. I. One-
dimensional derivative techniques and three-dimensional modeling. J. Non-Cryst.
Solids 305, 40–49.
Xu, Y., Shankland, T.J., Duba, A.G., 2000. Pressure effect on electrical conductivity of
mantle olivine. Phys. Earth Planet. Inter. 118, 149–161.
Yoshida, S., Ogawa, T., 2004. Electromagnetic emissions from dry and wet granite associ-
ated with acoustic emissions. J. Geophys. Res. 109 (B09204), 1–11.
Yoshida, S., Clint, O.C., Sammonds, P.R., 1998. Electric potential changes prior to shear
fracture in dry and saturated rocks. Geophys. Res. Lett. 25, 1577–1580.
Yukalov, V.I., Moura, A., Nechad, H., 2004. Self-similar law of energy release before mate-
rials fracture. J. Mech. Phys. Solids 52, 453–465.