electromagnetic emission and ae kaiser effect for ......electromagnetic emission and ae kaiser...

– 1 –

ISSN 0386-1678

Report of the Research Institute of Industrial Technology, Nihon UniversityNumber 93, 2008

* Professor, Department of Mechanical Engineering, College of Industrial Technology, Nihon University** Associate Professor, Department of Mechanical Engineering, College of Industrial Technology, Nihon University

Electromagnetic Emission and AE Kaiser Effect for Estimating Rock In-situ Stress

Yasuhiko MORI* and Yoshihiko OBATA**

( Received April 8, 2008 )

Abstract

The Kaiser Effect in acoustic emission is often used for estimation of the current stress level to which rocks have been subjected. However, there are cases in which the Kaiser Effect is not clear, since the noises due to the contact and/or the frictional slip (stick slip) between the pre-induced fracture surfaces are measured during the reloading process. In such cases, estimation of initial stress is difficult by the conventional method which is based on the acoustic emission activity observed under reloading process. In the tests for the Kaiser Effect on rocks, therefore, the noises must be eliminated from the acoustic emission generated from newly created cracks during the second loading process. Such techniques as analysis of the difference between the acoustic emission activity observed in the first and second reloading and the analysis of the change in the slope of the acoustic emission amplitude distribution have been proposed. In this paper we present a new method by which the maximum previous stress in rocks can be directly estimated without any post signal analysis. In this new method, simultaneous measurement of acoustic and electromagnetic emission during loading test of rock sample is employed. The electromagnetic emission in the deformation of rock sample generates only when the fresh surfaces due to cracking are created in the materials, and the source of electromagnetic emission is the electrification between the fresh crack surfaces. This paper describes the simultaneous measurement of acoustic and electromagnetic emission being useful for estimating the rock in-situ stress.

Keywords: Electromagnetic emission, Acoustic emission, Rock, Kaiser Effect, Stress history

1. Introduction

The stress state of rock in the Earth’s crust is an important consideration in the oil, mining and tunneling industries. To estimate the stress state, a number of geo-

physical and geological methods have been applied1). In addition, the acoustic emission (AE) technique has also become a method for non-destructive rock in-situ stress estimation.

The Kaiser Effect 2) which shows a marked stress

Yasuhiko MORI and Yoshihiko OBATA

– 2 –

electromagnetic emission during loading test of rock sample is employed.

It has been known that the changes in geo-electric potential and the anomalous radiation of geo-electro-magnetic waves were observed before major earth-quakes9) - 12). These phenomena also have been observed in laboratory experiments on rock samples, and it was found that micro- and macro-cracking processes are often accompanied by AE and electromagnetic emission (EME)13) - 18).

By conducting the monotonously increasing com-pressive loading test and cyclic loading test on granite samples, we have investigated the EME during the deformation process and found that the EME generates only when the fresh surfaces due to cracking are created in the material. It is concluded that the source of elec-tromagnetic emission is the electrification between the fresh crack surfaces accompanied with whole sample mechanical vibration19) - 22). This paper describes the simultaneous measurement of AE and EME being useful for estimating the rock in-situ stress.

2. Characteristics of EME from rock under deformation

The measurements of EME were carried out under the monotonously increasing compressive stress in the laboratory. The cyclic loading tests of the rock samples were also conducted to characterize the AE and EME generations in detail19), 20). This chapter describes the characteristics of EME from rock sample under de-formation.

2.1 EME under uniaxial compression

Inada Granite was examined in the present experi-ment. The surface appearance of the granite tested is shown in Fig. 1. Rectangular block sample with dimen-sions of 10 mm × 10 mm square and height of 30 mm was deformed under the uniaxial compression stress. Schematic sample assembly is shown in Fig. 2. Both ends of the rock sample are attached to steel end pieces by epoxy resin, as shown in Fig. 2 (a). The thickness of the epoxy fillets gradually decrease from the end pieces towards the middle of the sample (see the part of cross-sectional view of the fillet), so that the stress

history effect has been observed in various industrial materials. If the stress is monotonically increased, the AE activity begins to increase markedly when the stress exceeds the previous maximum stress. This non-revers-ible phenomenon of AE generation is well known as the Kaiser Effect. Based on this effect, the previously applied maximum stress can be estimated by observ-ing the AE activity under monotonically increasing stress. The Kaiser Effect has also been found during the deformation of rocks 2), 3). However in some cases, the Kaiser Effect is not clearly observed in the rocks, where the AEs occur in the reloading process even at a stress level lower than the previously applied stress level 4). It can be considered that the source of such AE signals are the frictional noises, that is, the contact and/or the stick slip between the pre-induced fracture surfaces induced even at the previous low load level 5). In such cases, estimation of the previous stress is dif-ficult by the conventional method, in which any signal analysis are not made for the AE data obtained during reloading process. In the tests for the Kaiser Effect on rocks, therefore, the frictional noises must be eliminated from the AE activity generated due to crack nucleation during the loading process.

In order to discriminate the desired AE events from the frictional noises, Yoshikawa and Mogi proposed a method of subtracting the AE activity of the second reloading from that of the first by conducting repeated loading 6). This method is based on the assumption that the frictional noise mentioned above shows the repro-ducibility in its onset load level under cyclic loading in rocks.

On the other hand, Shiotani et al. proposed a method of utilizing the b-value, which is defined as a slop of the AE amplitude distribution7), 8). In this method, the range of the amplitude has been determined by introducing the statistical values of amplitude distribution, such as mean value and standard deviation. The b-value ob-tained by this method is referred to as improved b-value (Ib-value). With Ib-value, the fracture process in the material under stressing is evaluated by characteristic changes in the Ib-values.

In this paper we present a new method by which the maximum previous stress in rocks can be directly estimated without any post signal analysis. In the new method, simultaneous measurement of acoustic and


– 3 –

concentration at the rock to steel contact point would be reduced23). The bending in uniaxial compression was minimized by keeping the ends of sample in parallel and the load was applied at a small central area of the upper steel end piece by placing a small disc of stiff paper (Fig 2 (a)) to reduce a mismatch of loading and sample axes. A stiff paper sheet was also placed between the bottom steel piece and fixed jig and reduced mechanical noise, helping to achieve good contact.

Two strain gages were mounted on the sample sur-face as shown in Fig. 2 (b) to measure the axial and transversal strains during loading. The strain signals and applied load from the loading machine were digitized and stored in the data logger.

In the deforming tests, acoustic emission (AE) signals were simultaneously measured with the electromagnetic emission (EME) signals, so that the generation of EME

could be directly compared with the entire fracture process of the rock sample estimated by AE. AE trans-ducer with a frequency range of 200 kHz to 700 kHz was mounted on the sample surface, as shown in Fig. 2 (b), to detect the AE signals during the loading test. In the present experiment, EME signals from rock sample were detected as the electric potential change appeared between two electrodes A-A, which were formed by painting a conductive paste on the sample surfaces, as shown in Fig. 2 (b). Both the EME and AE signals were amplified by a 40-dB preamplifier and led into the input channels of the AE system used. When an AE signal was detected at AE transducer the AE and EME signals were digitized at 0.25 μs interval, and the waveforms of 1024 μs duration were stored on 32 bits/word memory in the AE system. The threshold level of 60 dB was used for AE event detection. To eliminate interference arising from ambience, both the sample assembly on the loading frame and the preamplifiers were shielded electromagnetically by using an aluminum chamber having the wall thickness of 25 mm.

In the uniaxial tests, the sample was loaded at the different displacement speeds ranged from 0.02 mm/min. to 0.5 mm/min., so that the effect of displacement speed on the generation behaviors of EME and AE were examined.

In order to recognize and discriminate the EME signals from the background noise, the simultaneously recorded AE and EME signal waveforms were visually Fig. 1 Surface appearance of the Inada granite tested.

steel end pieceepoxy fillet rock

sample

stiff paper

stiff paper

A-A:electrodes

A A

straingages

AEtransducer

Fig. 2 Sample assembly: (a) setup for uniaxial loading and (b) arrangement of AE transducer, electrodes for EME detection, and strain gages.

(a) (b)


– 4 –

observed. Fig. 3 shows the simultaneously recorded AE (top) and EME (bottom) signal waveforms detected in the compression test conducted at a displacement speed of 0.5 mm/min., as an example. The background noise of EME signal channel is larger than that of the AE channel. Though the rock sample and preamplifier used for EME signal detection were shielded by using a thick aluminum chamber during the test, the electrodes for EME detection forms a high-impedance circuit (see Fig. 2(b)), and the background noise of EME channel is higher. This results in difficulty of discriminating the small amplitude EME signals. In the case shown in Fig. 3, generations of three major AE event signals are clearly recognized. As the onset of AE signal indicated by an arrow on the EME signal waveform matches the EME signal onset, it can be concluded that this EME signal is associated with AE. Conversely speaking, the AE signal, which is associated with EME signal, is the AE generated due to the creation of new fresh surfaces, i.e., crack. In this manner, even small ampli-tude EME signals, such as the signal generated near 100 μs in Fig. 3, are detectable. In addition, when the EME waveform has a good signal-to-noise ratio, it is clearly observed that the EME signal generates several microseconds before AE generation. This delay in AE is expected from the travel of the elastic waves through the sample. It is expected that the time, at which the EME signal is detected, corresponds to the moment of

crack nucleation in the sample, since the EME signal which is the electromagnetic wave propagates with the light velocity.

The results of simultaneous measurements of AE and EME signals during monotonously increasing compres-sion tests of the granite samples conducted at three different displacement speeds of 0.02, 0.06 and 0.5 mm/min. as shown in Figs. 4 (a), (b) and (c), respec-tively. In Fig. 4, applied load (P) and dilatant strain of

Fig. 3 A set of simultaneous measurement of EME and AE signals recorded in the compression test conducted at a displacement speed of 0.5 mm/min.

2

0

1

0 100 200 300 400 500

EME

-2

-1

t / μs

VE

M EV /

(a) displacement speed = 0.02 mm/min.

Fig. 4 Results of simultaneous measurements of AE and EME signals during monotonously increasing compression tests of the granite samples con-ducted at three different displacement speeds of 0.02, 0.06 and 0.5 mm/min.

6

0 100 200 300 400 500

0

3AE

VEA

V/

-3

-6

t / μs

899.0

999.0

000.1

100.1

200.1

ε v

εv

0 005 0001 0051 0002 0052 0003 005321-

9-

6-

3-

0

t s/P

/ kN P

.S.H.C mm/min.20.0=

0 005 0001 0051 0002 0052 0003 00530

5.0

1

5.1

2

5.2[× 01 4 ]

EA

NAE

, NEM

E / c

ount

t s/

EME

.S.H.C mm/min.20.0=

0 005 0001 0051 0002 0052 0003 005306

07

08

09

001

V AEp

/ dB

t s/

.S.H.C mm/min.20.0=

0 005 0001 0051 0002 0052 0003 005305

06

07

08

09

001

V EMEp

/ dB

t s/

.S.H.C mm/min.20.0=


– 5 –

(b) displacement speed = 0.06 mm/min. (c) displacement speed = 0.5 mm/min.

the rock sample (εv), cumulative AE event count (NAE) and EME event count (NEME), and amplitude of AE event signal (VAEp) and EME event signal (VEMEp) are plotted as a function of elapsed time of the loading (t). Generation of AE starts at a level of dilatant strain of the sample, where the dilatant strain deviates from the linear trend, and increases until the sample failure takes place. Generation of EME events starts after AE generation and increases with increasing of the applied stress. It is

also observed that the generation of active EME events is peculiar to the stressing stage, at which the volume of sample changes from the contraction due to compres-sion to the dilatancy developed in a direction vertical to the sample axis stressed. This result suggests that the EME signals were emitted from the micro crack created in tensile. In comparison with AE, the number of EME events measured is lower, because of a low sensitivity of EME channel. It can be said that the displacement

699.0

899.0

000.1

200.1

400.1

600.1

ε v

0 005 0001 005121-

9-

6-

3-

0

t s/

P / k

N

P

εv

.S.H.C mm/min.60.0=

0 005 0001 00510

5.0

1

5.1

2

5.2[× 01 4 ]

EA

t s/

NAE

, NEM

E/ c

ount

.S.H.C mm/min.60.0=

EME

0 005 0001 005106

07

08

09

001

t s/

V AEp

/ dB

.S.H.C mm/min.60.0=

0 005 0001 005105

060708

09001

.S.H.C mm/min.60.0=

t s/

V EMEp

/ dB

0 05 001 051 00221-

9-

6-

3-

0

P / k

N2999.0

4999.0

6999.0

8999.0

0000.1

ε v

P

εv

t s/

.S.H.C mm/min.5.0=

0 05 001 051 0020

1

2

3

4

5[× 01 3 ]

t s/

NAE

,NEME

/ cou

nt

EA

EME

.S.H.C mm/min.5.0=

0 05 001 051 00206

07

08

09

001

t s/

V AEp

/ dB

.S.H.C mm/min.5.0=

0 05 001 051 002050607

08

09001

t s/

V EMEp

/ dB

.S.H.C mm/min.5.0=


– 6 –

speed has no effect on the above mentioned generation behavior of AE and EME during the stressing. However, there is a difference in the number of total AE event count among three displacement speeds. Though the total event counts measured in the tests conducted at the displacement speeds of 0.02 and 0.06 mm/min. show almost the same value of 20 000, while only one fifth of the value, i.e., 4 000 counts were measured in the test at the displacement speed of 0.5 mm/min. The reason for the difference in event counts observed will be explained as follows. That is, again, the waveforms shown in Fig. 3 were the simultaneously recorded AE and EME signal waveforms detected in the test con-ducted at a displacement speed of 0.5 mm/min. By the visual observation three major AE event signals are recognized in a period of 500 microseconds. In this case, a dead time for AE detection was set at 1 millisecond on the AE system used. Therefore, the AE system was triggered by the first hit signal generated at about 100 microseconds, and counted the signals as one event, even though three events were visually recognized. In general, it is expected that when the material is stressed, the time interval of micro-cracks nucleation in the mate-rial becomes short with increasing of the deformation rate of the stressing. Thus, when the frequency of the occurrence of micro-cracks exceeds the AE dead time, as seen in Fig. 3, the AE event counts measured by the AE system show the smaller value than those of the actual value.

Next, the plots of the signal amplitude of EME (VEMEp) and AE (VAEp) as a function of the elapsed time of loading, as shown in Fig. 4, clearly demonstrate that the EME and AE events having the larger signal amplitude generate with increasing of elapsed time i.e., with increasing of the dilatant strain of the sample. This nature is irrespective to the displacement speeds tested.

The EME signals detected in the loading tests, with-out exception, were accompanied with the generation of AE signals. Thus, the relationship between the signal amplitudes of EME and AE was analyzed. The results, which are summarized in Fig. 5, strongly suggest the existence of the positive correlation between AE and EME in the signal amplitude. That is, the EME signals with the larger amplitude associate with the AE signals of larger amplitude.

2.2 EME under cyclic stressing

Cyclic loading tests of the rock samples have been conducted to characterize the AE and EME genera-tions19), 20). In the test 20), a granite sample with dimen-sions of 55 mm × 35 mm square and length of 200 mm was cyclically stressed until the rupture in a four-point bending manner. Experimental setup used is shown in Fig. 6. The function of cyclic loading was a sinusoidal waveform at a frequency of 0.2 Hz with constant load amplitude.

Figs. 7 and 8 show the important results. Fig. 7 shows

Fig. 5 Relationship between EME and AE signal amplitude.

100908070

6050

60 70 80 90 100VAEp / dB

VpE

M EBd /

(a) displacement speed = 0.02 mm/min.

100908070

6050

60 70 80 90 100

Vp E

M EBd/

VAEp / dB

(b) displacement speed = 0.06 mm/min.

100908070

6050

60 70 80 90 100VAEp / dB

VE

MEBd/

(c) displacement speed = 0.5 mm/min.


– 7 –

AE transducer

5355

electrode

200

180

60

P

rock sample

unit: mm

Fig. 6 Experimental setup of four-point bending cyclic loading test, showing the location of AE transducers and electrodes for detection of EME signal.

the AE source location plots as a function of the number of cyclic load. In this figure, open circles indicate the generation of AE and solid circles indicate the AE as-sociated with the generation of EME signals, and the diameter of each circle is proportional to the number of events located at the point. Referring to the sketch of the fractured sample shown at the right side of the figure, both the generations of AE and the EME signals concentrate in the fractured location of the sample.

On the other hand, Fig. 8 shows the plot of load level at which each AE and EME event occurred as a function of the number of cyclic loading. In the figure, y-axis is the time in a load cycle for each loading cycles, and corresponding loading waveform is shown at the left side of the plot. The meaning of the plot symbol in the figure is the same to that of Fig. 7. From this result,

the synchronized occurrence of AE and EME mainly occur during the uploading process and at around the maximum load levels in a sinusoidal load. AE signals also occur at fairly low load levels in the load increasing and decreasing stages just before the sample rupture. It had been found that the contact and/or the stick slip be-tween the pre-induced fracture surfaces are the sources of those AE 5). It must be noted that two types of AE generate in the rock sample under cyclic loading: i.e., AE due to the formation of new micro-fractures and its expansion, and AE due to the frictional sliding along existing crack surfaces within a rock. Therefore, the result shown in Figs. 7 and 8 clearly demonstrates that the EME signals would be due to the nucleation of new micro crack and its extension.

30

erutcarf

AE

elpmas

kcor

AE with EME

0

-10

-20

10

20

E A(

EME

mm/

n oita colecru os )

-3088.2 8988.888.688.4

number of cycles, n / 103 cycles

88

,elcy cda ol

ani

emit

ts/

xam

n im

daol

AE with EME

88 88.2 88.4 88.6 88.8 89

number of cycles, n / 103 cycles

AE5

4

3

2

1

0

Fig. 7 Plot of AE and EME source location as a function of the number of cyclic loading. (Inada granite sample in four-point bending)

Fig. 8 Plot of AE and EME generation load level as a function of the number of cyclic loading. (Inada granite sample in four-point bending)


– 8 –

2.3 Characteristics of EME from rock under de-formation

The experimental results shown in the previous sec-tion 2.2 obviously pointed out that the EME generates only when the fresh surfaces due to cracking are created in the material. Several theoretical models have been proposed for the explanation of EME phenomena, such as the piezoelectric effect, the electro-kinetic effect, or the electrification between the newly created surfaces16). The present authors have proposed a model of EME generation, which is based on the movement of charged crack faces during the process of cracking 22). On the faces of a newly created crack, the electric charges constitute an electric dipole system and due to the crack wall vibration, an EME signal is emitted with a fre-quency corresponding to the crack length. The second source of EME signal, which is activated after the first source EME signal mentioned above with a delay of 10 to 20 micro-seconds, is due to the whole sample mechanical vibrations with a frequency given by the sample boundary conditions. The dependence of EME and AE events on mechanical strain was investigated using a ramp increasing pressure. Events frequency is very low in the range of Hooke’s law and reaches a high value near the sample breakdown.

On the other hand, the signal amplitude of EME would correspond to the amount of the electric charges re-distributed on the newly created crack surfaces. It is expected that the amount of electric charges would depend on the size of the crack surfaces created. The results, obtained form the monotonously increasing compressive tests discussed in Section 2.1, showed that both EME and AE signals are correlated and their amplitudes increase just before the sample rupture (Fig. 4). In addition, it was also found that there is a positive correlation between the amplitudes of EME and AE signals (Fig. 5). It is well known, in general, that the amplitude of AE signal depends on the scale of the fracture created in the material. That is, the larger AE signal is associated with the creation of the larger crack volume. Therefore, it can be said that the present ex-perimental results strongly support the proposed model of EME generation, i.e., the electrification between the newly created surfaces is the primary source of EME generation.

3. EME under Kaiser Effect testing

One of the important applications of AE technique in the field of civil engineering is the determination of stress history induced in the rocks by using Kaiser Effect. The AE activity under monotonically increasing stress begins to increase appreciably at the previously applied maximum stress. This is the Kaiser Effect, which is observed in rocks as well as other materials. In the metallic materials, in general, the value of Kaiser Effect ratio (also called Felicity ratio in the AE/compos-ite field), which is defined as the ratio of the stress at which AE activity appears in the second stressing to the previously applied peak stress, ranges from 0.9 to 1.024). A ratio of 1.0 indicates the intact materials condition, exhibiting Kaiser Effects holding. On the other hand, in the rocks, AEs occur in the reloading process even at a stress level lower than the previously applied stress level. It is known that the source of such AEs is fric-tional, that is, the contact and/or the stick slip between the pre-induced fracture surfaces, since the micro-cracks would have been induced even at small load levels in the rock during the previous loading process, as mentioned in Chapter 2 in this paper 5), 8).

In tests for the Kaiser Effect on rocks, therefore, the frictional AE must be eliminated from the AE measured during the second reloading process; that is, only the AE generated from newly created cracks should be measured for evaluating the stress history in the rocks. Several methods have been proposed to distinguish the frictional AE from the cracking AE; for example, the method of subtracting the AE activity of the second reloading from that of the first23), the method of utilizing the b-value, which is defined as a slop of the AE am-plitude distribution7), 8), and the method combining the deformation rate analysis and the Kaiser Effect 25), 26).

The EME signal appears only when the fractures oc-curred inside the material under stressing, as mentioned in the previous section. It is thus expected that EME signals would directly provide the information only on the creation of fracture event. No special equipment is needed to measure the EME signals. We can make do with usual AE system by allocating one channel for the input of EME signal. The sensing device of EME signals is a coil and/or an antenna or a couple of electrodes, which are placed near or on the sample under test. Such


– 9 –

sensing device is quite simple and can be formed easily. This is also the great advantage in utilizing the EME.

In the present study, Inada Granite was chosen for the experiment. Two types of rectangular block samples, of which dimensions were 20 mm × 20 mm × 80 mm and 10 mm × 10 mm × 40 mm, were stressed in repeated uniaxial compression load. The sample was pre-loaded up to a certain value and then unloaded. It is then loaded to a larger value and then unloaded again. This load-ing-unloading procedure was repeated until the sample rupture took place. Each of repeated loading-unloading procedures was conducted at a constant displacement speed of 0.5 mm/min. The sample assembly for the repeated loading test was basically the same as those shown in Fig. 2.

In the repeated loading test, the EME and AE signals were also simultaneously measured. The experimental set-up and condition for measuring EME and AE signal are shown in Fig. 9 and Table 1, respectively. AE trans-ducer with a frequency range of 200 kHz to 700 kHz was mounted on the sample surfaces to detect the AE

signals during the loading test. The EME signal was detected as the electric potential change appeared at a couple of electrodes, which were formed on the sample surfaces (Fig. 2). The EME signals were amplified by 20 dB using 3S Sedlak PA 21 ultra low nose preampli-fier. The internal noise level of PA21 preamplifier was 3 nV/sqrHz in the frequency range of 500 Hz – 10 MHz. The output signals of PA21 were further amplified by 40 dB with a PAC 1220A preamplifier. AE signals were amplified by 40 dB with a PAC 1220A preamplifier. Both the EME and AE signals were led into the input chanels of AE system of PAC Mistras 2001. The thresh-old levels were 70 dB for EME signal and 60 dB for AE signal. To eliminate noise, both the sample assembly on the loading frame and the preamplifiers were shielded electromagnetically by using an aluminum chamber having the wall thickness of 25 mm.

The result of simultaneous measurements of EME and AE signals during the repeated loading test con-ducted on the block sample of 20 mm × 20 mm × 80 mm is shown in Fig. 10. The loading history for the repeated

pre-amp. signal PA21 1220A thresholdEME 20 dB 40 dB 70 dB AE — 40 dB 60 dB

Fig. 9 Measuring set-up for repeated loading test.

Table 1 Conditions for EME and AE signal detection.

EMEAE

#3#4

#2#1

#0

#5#6

#7#8 #9

0 1000 2000

-100

-50

0

-150

3

2

1

0

30

20

10

0

t / s

σaP

M/

NEA

01×/3

tnuo c

NE

MEtnu oc/

Fig. 10 History of the repeated stressing test and the result of simultaneous measurement of EME and AE signals conducted on Inada granite sample of 20 mm × 20 mm × 80mm.


– 10 –

loading test is shown as the stress curve, σ. After an initial stress of 12.5 MPa, which corresponds to ap-proximately 10 % of compression strength of the rock tested, was applied, the maximum stress level of suc-cessive loading was increased by 12.5 MPa steps until the sample failure. In Fig. 10, EME and AE events are represented by the cumulative event counts measured for each stressing step. The EME signals are detected every stressing steps, with the exception of step #1. The EME signals are generated only in the stage where the applied stress increases and reaches at the maximum stress level for each step, whereas the generation of AE signals is observed not only in the stress increasing stage but also in the unloading stage. For example, the details of the stressing steps #3 and #8 are shown in Fig. 11. In the figures, the arrow head indicates the stress cor-responding to the maximum pre-stress for the each step. In the Fig. 11, open symbol plotted on the stress curve denotes the onset stress of active AE event generation during the stress increasing stage. Similarly, the solid symbol on the stress curve denotes the onset stress of EME signal. These onset stresses of AE and EME are referred as σAE and σEME, respectively. The values of the σAE and σEME for each repeated stressing step as shown in Fig. 10 and also for the experimental results obtained in the test conducted on the block sample of 10 mm × 10 mm × 40 mm were evaluated. In Fig. 12, the σAE and σEME obtained are plotted as a function of the pre-stress level σpre. In the figure, these stresses are normalized by the compression strength of the rock tested σb. The Kaiser Effect of acoustic emission is well recognized in the range of the pre-stress below 22 % of the σb, giv-ing the underestimation of pre-stress. Specifically, the amount of the underestimation is 10 % to 25 % of the actual pre-stress. This underestimation would be caused by that the AE due to the friction between the fracture surfaces induced during the previous loading process is measured in the early stage of each successive stressing step. However, it should be noted that in general, Kaiser Effect shall be discussed within the region of elastic or elasto-plastic deformation. In the case of the sample tested, it is expected that the corresponding stress range is 30 % of the maximum strength.

On the other hand, the onset stress level of EME signal, σEME, estimates the pre-stress level within the deviation of 10 %, over a wide range of the pre-stress

EME

AEσpre

σ

#8

t / s

σ/

aP

M

0

-40

-80

-1202000 2100 2200 2300

0

1

3

2

NEA

/01×

3tnu oc

0

15

10

5NE

M E/

tn uo c

Fig. 11 Expanded stressing steps #3 and #8 in the test procedure shown in Fig. 10.

1

0 0.2 0.4 0.6 0.8 10

0.2

0.4

0.6

0.8σΕΜΕσΑΕ

σpre /σb

σE

ME/σb,σ

EA/σb

Fig. 12 Plots of estimated stresses σEME and σAE as a function of the pre-stress σpre.

600 0

600 700

-10

-20

-30

-50

-40

500

400

300

0

200

100

2

5

4

3

1

0

NE

ME/

tn uoc

σaP

M/

NEA

/tn uoc

AE

#3

σEME

σpre

t / s

(a) Step #3

(b) Step #8


– 11 –

up to 90 % of the compression strength of the sample tested.

This result clearly suggests that the emission of EME signals is associated with the creation and/or extension of micro-cracks, and is non-reversible phenomena, simi-lar to the Kaiser Effect of acoustic emission. Therefore, it can be concluded that the EME signal measurement can estimate more accurately the current stress level to which the rock samples have been subjected than that of AE activity.

Next, the above mentioned method proposed by Yoshikawa and Mogi was investigated for the results shown in Fig. 10. Two examples of the results obtained for the steps #1-2 and steps #5-6 are shown in Fig. 13. To estimate the pre-stress, by conducting repeated re-loading test, it was proposed that if the acoustic emission activity observed in the second re-loading (steps #2 or #6 in Figs. 13 (a) and 13 (b), respectively) is subtracted from that of the first (steps #1 or #5), the difference between them begins to increase markedly at the previous maximum stress, as seen in the case of Fig. 13 (a). In the case of Fig. 13 (b), the pre-stress is not successfully estimated. Table 2 shows the summary of the estimated stress σAE and the pre-stress σpre examined for the result shown in Fig. 10. In the table, the estimated stress σAE is normalized by the pre-stress σpre, and σb is the fracture strength of the sample tested. In some cases, σAE is appreciably lower than σpre. However, it has been noted that the accuracy in the stress estimation of this method is satisfactory for a lower ratio of the pre-stress level to the rock strength. In the case of Inada granite tested, the stress can be estimated when the pre-stress is smaller than 40 % of the strength.

The method of analyzing the slop of AE amplitude distribution to estimate the stress was also examined for the results obtained in the present work. For the present test, the range of AE amplitude on the amplitude distribution was limited within μ–σS and μ+0.5 σS ac-cording to the process proposed by Shiotani et al.8) and the slope of the amplitude distribution for the range was calculated as the Ib-value. Here μ is a mean value of AE amplitude distribution of AE events measured and σS is its standard deviation. Two examples of the results obtained for the step #4 and step #7 are shown in Fig. 14. In Fig. 14, the elapsed time at which the applied stress reached the pre-stress is indicated by an arrow

σpre

#1

#2

300

100

200

0

σ / MPa 3020100

NE A

tnuoc /

#2σpre

#1

#0

Fig. 13 Estimation of pre-stress σpre by observing AE activity for the successive repeated re-loadings.

0 200

200

400

600

40 60 80σ / MPa

#6σpre

#5

#4#5

σpre

#6NEA

tnuoc/

σpre [MPa] ( pre/σσ b) σ σAE/ pre

12.4 (0.11) 0.97

24.8 (0.22) 0.81

37.0 (0.36) 0.91

49.6 (0.44) 0.51

62.3 (0.56) 0.70

73.9 (0.66) 0.73

87.1 (0.78) 0.63

Table 2 Pre-stress σpre and estimated stress σAE by repeated loading method.

(a) repeated re-loading steps #1 and #2

(b) repeated re-loading steps #5 and #6


– 12 –

and also the time of the generation of the first EME signal is shown by a triangle symbol. In this method, the fracture events occurred inside the material are evalu-ated by observing the changes in the Ib-value, on the assumption that the AE activities would become larger with approaching the final failure of the materials. In the cases shown in Fig. 14, the abrupt changes in the Ib-value are observed at the elapsed time at which the applied stress is approaching the pre-stress level. The appearance of the characteristic change in the Ib-value is also close to the EME signal generation. For the rock tested, it seems that the method would be applicable to estimate the pre-stress larger than 40 % of the strength. Though it is difficult to associate the onset of macro-fracturing, obtained from the drop of Ib-value, with estimation of initial stress with Kaiser Effect, Ib-value analysis would be useful for detecting the newly created fracture inside the material.

4. Summary

Simultaneous measurement of EME and AE during the monotonously increasing compressive loading test and the cyclic loading test has been conducted on the granite samples to characterize the EME generation in detail. It was found that the EME signal is accompanied with the generation of AE signal, and there is a positive correlation between the amplitude of EME and AE signal. It can be concluded that the EME generates

only when the fresh surfaces due to cracking are cre-ated in the materials, and the electrification between the newly created surfaces is the primary source of the EME generation.

Simultaneous measurement of EME and AE to es-timate the rock in-situ stress was proposed as a new method. Such techniques as the AE method based on the Kaiser Effect and the analysis of the change in the slope of AE amplitude distribution have a limitation in the pre-stress level to be estimated. A new method is applicable to estimate the pre-stress over the wide range of the stress up to 90 % of the rock strength. The greatest advantage of the new method is that the current stress level to which the rock samples have been subjected can be directly estimated without any post signal analysis. In addition, the detection of EME from rock sample under stressing is very simple and easy, and a usual AE system can be used to record and analyze the EME signals.

The simultaneous measurement of AE is important to help find the source position of the electromagnetic emission signal and to understand what type of fracture is created.

Acknowledgment

This research has been partially supported by JSPS Grant-in-Aid for Scientific Research in Category C of Scientific Research Grant of project No. 15560082.

bI

σpre

EME

700 800 900 1000t / s

0.11

0.1

0.09

0.08

0.07

0.06

0.05

Fig. 14 Estimation of pre-stress by the slop of AE amplitude distribution analysis.

1700 1800 19001600t / s

0.11

0.1

0.09

0.08

0.07

0.06

0.12

b I

σpre EME

(a) Step #4 (σpre = 0.44σb) (b) Step #7 (σpre = 0.78σb)


– 13 –

References

1) e.g., A. McGarr and N. C. Gay, “State of stresses in the earth’s crust”, Annual Review of Earth and Planet Sciences, Vol. 6, pp. 405-436 (1978).

2) J. Kaiser, “Erkentnisse und Folgerungen aus der Messung von Geraeuschen bei Zugbeanspuruchung von metallischen Werkstoffen”, Archiv fuer das Eisenhuettenwesen, 24, pp. 43-45 (1953).

3) R. E. Goodman, “Subaudible noise during com-pression of rocks”, Geological Society of America Bulletin, 74, pp. 487-490 (1963).

4) S. Yoshikawa and K. Mogi, “Kaiser effect of acoustic emission in rocks-influences of water and temperature disturbances”, Proc. 4th Acoustic Emission Symposium, Tokyo, pp. 22-40 (1978).

5) Y. Mori, K. Saruhashi and K. Mogi, “Acoustic emission from rock specimen under cyclic load-ing”, Progress in Acoustic Emission VII, pp. 173-178 (1994).

6) S. Yoshikawa, and K. Mogi, “A new method for estimation of the crustal stress from cored rocks samples – Laboratory study in the case of uniaxial compression”, Tectonophysics, 74, pp. 323-339 (1981).

7) T. Shiotani, K. Fujii, T. Aoki and K. Amou, “Evaluation of progressive failure using AE sources and improved b-value on slope model tests”, Progress in Acoustic Emission VII, pp. 529-534 (1994).

8) T. Shiotani, M. Ohtsu and K. Ikeda, “Detection and evaluation of AE waves due to rock deformation”, Construction and Building Materials, 15, pp. 235-246 (2001).

9) M.B. Gokhberg, V.A. Morgunov, T. Yoshino and I. Tozawa, “Experimental measurement of electro-magnetic emissions possibly related to earthquakes in Japan”, Journal of Geophysical Research, 87, pp.7824-7828 (1982).

10) J.W Warwick, C. Stoker and T.R. Meyer, “Radio emission associated with rock failure: Possible ap-plication to the Great Chilean Earthquake of May 22, 1960”, Journal of Geophysical Research, 87, pp.2851-2859 (1982).

11) T. Rikitake, “Nature of electromagnetic emis-sion precursory to an earthquake”, Journal of Geomagnetism and Geoelectricity, Vol. 49,

pp.1153-1163 (1997).12) N. Gershenzon, M. Gokhberg, V. Morgunov,

“Sources of electromagnetic emissions preceding seismic events”, Izv Earth Physics, Vol.23, pp.127-135 (1987).

13) U. Nitsan, “Electromagnetic emission accompany-ing fracture of quartz-bearing rocks” Geophysical Research Letter, 4, pp. 333-336 (1977).

14) T. Ogawa, K. Oike and T . Miura, “Electromagnetic radiations from rocks”, Journal of Geophysical Research, 90, pp.6245-6249 (1985).

15) G. O. Cress, B. T. Brady and G. A. Rowell, “Sources of electromagnetic radiation from fracture of rock samples in the laboratory”, Geophysical Research Letter, 14, pp. 331-334 (1987).

16) I. Ymada, K. Masuda and H. Mizutani, “Electromagnetic and acoustic emission associated with rock fracture”, Physics of Earth and Planetary Interiors, 57, pp. 157-168 (1989).

17) V. Hadjicontis and C. Mavromatou, “Transient electric signals prior to rock failure under uniaxial compression”, Geophysical Research Letter, 21, pp. 1687-1690 (1994).

18) T. Lolajicek and J. Sikula, “Acoustic emission and electromagnetic effects in rocks”, Progress in Acoustic Emission VIII, pp. 311-314 (1996).

19) Y. Mori, K. Sato, Y. Obata and K. Mogi, “Acoustic emission and electric potential changes of rock samples under cyclic loading”, Progress in Acoustic Emission IX, pp. II: 1-8 (1998).

20) I. Iida, Y. Mori, Y. Obata and K. Mogi, “Measurement of AE and electric potential changes in fracture of brittle materials”, Progress in Acoustic Emission X, pp. 325-330 (2000).

21) Y. Imaizumi, Y. Mori et al., “Measurement of acoustic and electromagnetic emission and radon emanation from rocks”, Proc. Int. Workshop on Experimental Methods in Acoustic and Electromagnetic Emission, Brno, Czech, pp. 54-62 (2002).

22) J. Sikula, T. Lokajicek and Y. Mori, “Cracks characterisation by electromagnetic and acoustic emission”, Proc. of 8th ECNDT, AEND, Barcelona, CDR No. 62 (2002).

23) S. Yoshikawa and K. Mogi, “Experimental study on the effect of stress history on acoustic emis-sion activity – A possibility for estimation of rock


– 14 –

stress”, Journal of AE, 8, 4, pp. 113-123 (1989).24) Y. Mori, “Kaiser Effect of acoustic emission in

metals and alloys – Kaiser Effect Ratio (K. E. R.) and Microplasticity”, Proc. 4th Acoustic Emission Symposium, JSNDI, 7: 32-42 (1978).

25) K. Yamamoto, H. Yamamoto, N. Kato and T. Hirasawa, “Deformation rate analysis for in-situ stress estimation”, Proc. 5th Conf. On Acoustic

Emission/Microseismic Activity in Geologic Structures and Materials. Trans Tech Publications, pp. 243-255 (1995).

26) S. P. Hunt, A. G. Meyers and V. Louchnikov, “Modelling the Kaiser Effect and deformation rate analysis in sandstone using the discrete element method”, Computers and Geotechnics, 30, pp. 611-621 (2003).

– 15 –

岩石のその場応力の評価における電磁放射とAEカイザー効果

森　　康彦，小幡　義彦

概　　要

アコースティック･エミッション（AE）のカイザー効果は，岩石のその場応力（in-situ stress）の大きさの推定評価にしばしば用いられている。しかしながら，岩石にすでに導入されていたき裂面どうしの接触あるいは摩擦による雑音が再負荷の過程で計測されて，カイザー効果が明瞭でない場合がある。このような場合，荷重負荷の過程で観察されるAEの活動だけに基づいた簡便な方法による予応力の評価は難しい。カイザー効果の試験では，したがって，２回目の負荷で新たに生成されるき裂から発生するAEから，そのような雑音を除去する必要がある。そこで，繰返し負荷を行って１回目と２回目の負荷で観察されるAE活動の差を解析する，あるいは，AE振幅分布の勾配の変化を解析するなどの方法が提案されてきている。本論文では，岩石の最大予応力を，事後にいかなる信号処理も必要とすることなく，直接的に評価できる新しい方法を提案する。この新しい方法は，AEと電磁放射の同時計測を岩石の負荷試験に採用するものである。岩石試料の変形における電磁放射は，試料の中に割れに伴って新生表面が生成されるときだけに発生するものであり，電磁放射の源は新生き裂面間に生じる電化である。本論文は，AEと電磁放射の同時計測が岩石のその場応力を評価することに有効であることを述べる。

– 16 –

Biographical Sketches of the Authors

Yasuhiko Mori was born in Nagano, Japan, on September 4th 1940. He received the B.E. and M.S. degree in electrical engineering from Nihon University, in 1963 and 1967, respectively, and the Ph.D. degree in mechanical engineering from Nihon University, in 1980. Dr. Mori is a professor of the Department of Mechanical Engineering, College of Industrial Technology, Nihon University. His research interest is the materials evaluation using acoustic and electromagnetic emission technique.

Yoshihiko Obata was born in Niigata, Japan on December 26th 1952. He finished Doctor Course, Graduate School of Nihon University, and received his Doctor degree of engineering. Dr. Obata is an associate professor of the Mechanical Engineering, College of Industrial Technology, Nihon University. His working topics are acoustic emission measurement and interpretation of materials behavior on loading and droplet burning of an emulsion. He is a member of The Japan Society of Mechanical Engineers (JSME) and The Japanese Society for Non-Destructive Inspection (JSNDI).

electromagnetic emission and ae kaiser effect for ......electromagnetic emission and ae kaiser...

Documents