This article was downloaded by: [York University Libraries]On: 20 November 2014, At: 12:34Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Nondestructive Testing and EvaluationPublication details, including instructions for authors and subscription information:http://www.tandfonline.com/loi/gnte20

EVALUATION OF MICROFRACTURE PROCESS IN ALUMINABY ACOUSTIC EMISSIONSHUICHI WAKAYAMA a & HISASHI NISHIMURA aa Department of Mechanical Engineering, Faculty of Technology , Tokyo MetropolitanUniversity , 1-1 Minami-Ohsawa, Hachioji-shi, Tokyo, 192-03Published online: 28 Nov 2010.

To cite this article: SHUICHI WAKAYAMA & HISASHI NISHIMURA (1992) EVALUATION OF MICROFRACTURE PROCESS IN ALUMINA BYACOUSTIC EMISSION, Nondestructive Testing and Evaluation, 8-9:1-6, 717-728, DOI: 10.1080/10589759208952745

To link to this article: http://dx.doi.org/10.1080/10589759208952745

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in thepublications on our platform. However, Taylor & Francis, our agents, and our licensors make no representationsor warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Anyopinions and views expressed in this publication are the opinions and views of the authors, and are not theviews of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should beindependently verified with primary sources of information. Taylor and Francis shall not be liable for any losses,actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoevercaused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematicreproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in anyform to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Nondestr. Test. Eval., Vol. 8-9, pp. 717-729Reprints available directly from the publisherPhotocopying permitted by license only

© 1992 Gordon and Breach S.A.Printed in the United Kingdom

EVALUATION OF MICROFRACTURE PROCESS INALUMINA BY ACOUSTIC EMISSION

SHUICHI WAKAYAMA and HISASHI NISHIMURADepartment of Mechanical Engineering, Faculty of Technology, Tokyo

Metropolitan University, 1-1 Minami-Ohsawa, Hachioji-shi, Tokyo 192-03

Microfracture process during the bending tests of alumina ceramics were evaluated by acoustic emissiontechnique. Specimens with different dimensions were used for the bending tests in order to investigatethe dependence of microfracture process on the specimen size. A remarkable point in AE generationpattern of each specimen, at which both AE events and energy increased rapidly, was observed beforethe final unstable fracture. It is important that the apparent stress, Gc, at those points were independentof the AE threshold level and specimen size.

The locations of AE sources, i.e. microcracks, were determined using the differences of arrival timesbetween 2 ch transducers. Those were distributed widely before the stress was lower than G" and theywere concentrated on the point which would become the origin of final unstable fracture. Using thefluorescent dye penetrant method, fracture process on the surface of bending specimens were observed.Those results demonstrated that the stress, Gc, corresponds to the critical stress for the maincrackformation due to the coalescence of microcracks and/or pores. Consequently, it was concluded that thecritical stress can be the new evaluation parameter, which is equivalent to yield strength in metals, forceramic materials.

KEY WORDS: Ceramics, acoustic emission, microfracture process, criticalstress, maincrack formation

INTRODUCTION

For the reliability of ceramics, the non-destructive detection and characterizationof microcracks, i.e. their nucleation time, location, size, velocity and fracturemode, are the most important problems. Especially, the establishment of thetechnique evaluating the critical stress for microcracking, which is the essentialnon-destructive evaluation (NDE) parameter in itself, is also necessary to under­stand the toughening mechanism [1,2] and to develop the statistical treatment ofceramics [3,4].

Therefore the applications of non-destructive evaluation (NDE) on the charac­terization of ceramics have been attempted. Especially, many acoustic emissionstudies on ceramics [5,6], including authors' [7,8], has been carried out because ofits easiness in the detection of the nucleation time and location of microcracks inthe material.

In this study, the bending of alumina ceramics using the specimen with severaldifferent dimensions were carried out. Two types of alumina ceramics withdifferent microstructures were used. Microfracture process during bending testswere evaluated by acoustic emission and then the apparent stress, 0c, at which bothAE events and energy increases rapidly, was determined. The fluorescent dye

717

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

718 S. WAKAYAMA ETAL.

penetrant method was also applied for the observation of fracture process on thesurface. Then, it was understood that ac corresponds to the critical stress formaincrack formation due to the coalescence of microcracks and/or pores. Thepurposes of this paper are to establish the evaluation technique of the critical stressfor microcracking and to demonstrate the possibility of the critical stress as the newevaluation parameter for ceramic materials.

EXPERIMENTAL PROCEDURE

Materials and Bending Tests

Materials used were two kinds of alumina ceramics with the different mean grainsizes. Coarse grain alumina (ADS-lO) has the mean grain size of 20 .urn, Young'smodulus of 385 GPa, porosity of 5% and purity of 99.5%. Those of fine grainalumina are 5 .urn, 227 GPa, 9% and 92%, respectively. Specimens were cut fromalumina blocks by diamond saw to the dimensions of 3 x 3 x 40 mrn, 4 x 4 x 40mm and 5 x 5 x 40 mm. Each of the specimens were chamfered to 0.1 mm andpolished with diamond powder of 1 .urn. In order to investigate the influence ofcorrosion by water on the microfracture process, "dry" specimens were alsoprepared by the dehydration in vacuum of 10-4 Torr at 150°C, 60 min.

Specimen sizes used for bending tests are tabulated in Table 1. All of the testswere carried out using the Instron-type tensile testing machine in air (temperature:20°C, relative humidity: 60%) with a bending test equipment, where crossheadspeed was controlled so carefully that the strain rate was constant as 4 x 10-6 Is.

Table t Specimen Sizes(a) Coarse Grain Alumina (ADS-IO)

Specimen Width Height Lower UpperName (mm) (mm) Span (mm) Span (mm)

5 x 5 3·Point 5 5 243 x 3 4·Point 3 3 24 84 x 4 4-Point 4 4 24 85 x 5 4-Poinl 5 5 24 8

(b) Fine Grain Alumina (ADS-SO)

Specimen Width Height Lower UpperName (mm) (mm) Span(mm) Span(mm)

3 x 3 3-Point 3 3 244x43-Point 4 4 24

3 x 3 S4-Point 3 3 9 33 x 3 4-Point 3 3 24 84 x 4 4-Point 4 4 24 8

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

AE Measurement

MICROFRACfURE PROCESS IN ALUMINA 719

The AE measuring system used in this study is shown in Figure 1, schematically.Two piezo-electric elements were used as AE sensors and attached directly on theboth ends of the specimen. In this study, AE source locations were calculated fromthe difference of arrival times between two sensors. Therefore, sensitivity of twosensors, especially those equilibrium, was calibrated carefully enough using thepencil lead breaking as a simulated source, before each testing.

SensorC;:::~~~~:J

Figure I AE measuring system.

Since it is well known that the AE activity of ceramic materials is low, theminimization of noise level of the system is indispensable for AE measurement ofsuch materials. In this study, noise-filter-transformers were used and the connec­tions between sensors and pre-amplifiers were modified as the same as differentialtype transducers. Consequently, the noise level at the input terminal of pre­amplifier was decreased to 14flY, then threshold level was selected as 18flY. AEsignals were measured by the AE system (PAC: LOCAN) with load signals, sent toa personal computer (NEC:PC9801 LX4) through the RS-232C interface andanalyzed with the computer.

Observation of Fracture Process

The fracture process on the surface under the tensile stress was observed using afluorescent dye penetrant method according to the following sequences.

1. Specimens, covered with penetrant, were loaded to some predestinated valueand unloaded immediately.

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

720 S. WAKAYAMA ETAL.

2. Penetrant was removed and the surface was observed using an opticalmicroscope under an ultraviolet light.

3. Specimen was loaded to higher value than the former, then these sequenceswere repeated until the specimen fractured.

RESULTS

Bending Strength

The bending strength obtained in this study is tabulated in Table 2. Sixteen totwenty specimens were used for each kind of specimen size. The strength of coarsegrain alumina was higher than that of fine grain alumina. In coarse grain alumina,Weibull's coefficients are almost constant for 4-point bending, while it was twice for3-point bending. In fine grain alumina, the dehydration enhanced Weibull'scoefficient but did not vary the average strength obviously.

Table 2 Results of Bending Tests(a) Coarse Grain Alumina (ADS-IO)

Standard EffectiveSpecimen Average Deviation Area Weibull's

Name (MPa) (MPa) (mm') Coefficient

5 x 53-Point 304 14.2 8.62 213 x 3 4-Point 270 20.3 27.4 134x44-Point 256 21.8 36.6 115 X 54-Point 267 27.8 45.7 9.7

(b) Fine Grain Alumina (ADS-80)

Standard EffectiveSpecimen Average Deviation Area Weibull's

Name (MPa) (MPa) (mm/] Coefficient

3 x 3 3·Point 228 14.8 4.50 154 x 4 3·Point 215 12.3 5.33 173 x 3 4-Point 186 12.2 26.7 154 X 4 4·Point 183 10.4 35.6 20

(c) Fine Grain Alumina (ADS-80): Dry Specimens

Standard EffectiveSpecimen Average Deviation Area Weibull's

Name (MPa) (MPa) (mm') Coefficient

3 X 3 3-Point 216 9.1 2.88 243 X 3 S4-Point 197 9.8 9.72 213 x3 4-Point 200 7.6 25.9 284 X 4 4·Point 194 9.1 34.6 21

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

MICROFRACfURE PROCESS IN ALUMINA 721

(1)

In the case of bending tests, it is considered that the strength depends on thefracture process on the surface. Therefore, effective area SE can be used as themeasure of the specimen size, i.e.

s£= Is (oloR)"'dS

= b[(L t - Lz)/(m + 1) + L z]

where 0 is the tensile stress on the surface area element dS, OR is the maximumstress, b is the width and L" Lzare the lower, upper span, respectively. Accordingto Weibull statistics [9], the mean strength.zz, and,uz, obtained from specimens withtwo different sizes (SEt, S£2) are related as

(2)

This equation means that the larger a specimen is, the lower an obtained strengthis, as actually shown in Table 2.

AE Behavior

In this study, electromagnetic noise was minimized as mentioned above. On theother hand, mechanical noise was removed by teflon sheets between loading rodsand a specimen, and then the negligibility was ascertained by the pre-loading usingthe specimen with twice height and same width. Furthermore, the AE sourcelocation was also used for the decrease of mechanical noise, especially for 4-pointbending, i.e. the events located out of the upper span were neglected. Therefore itcan be concluded that the sources of those AE events are microcrackings.

Figure 2 shows the result of 4-point bending test and typical AE generationpattern. In the figure, cumulative AE events counts are shown for the thresholdlevel of 26 dB, 30 dB and 35 dB, respectively. It can be understood that theapparent bending stress, at-which AE events increases raidly, is independent of thethreshold level. Therefore, those stresses were determined from each specimen, as

°c'In conventional AE studies on the bending tests of ceramics [5,6], the beginning

points of AE generation were focused. For ail example, Gogotsi et at. [5]demonstrated that the stress at those points had the same distribution as bendingstrength. But it is clear in Figure 2 that the stress at those points are influenced bythreshold level in comparison with 0c.

The microcrack locations, which were nucleated during the 4-point bending test,are shown in Figure 3. In this figure, the AE source locations were determined bythe product of the longitudinal wave velocity of the material (= 10,000 mls) and thedifferences of the arrival times between two transducers. It can be seen from thefigure that although microcracks distribute widely on the specimen surface before

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

722 S. WAKAYAMA ET AL.

300r------------....150

o a~Q.20u~-enen~ 100-tf)

-- Stress------- Events(Threshok:l =26d3)_. _.- Events(Threshok:l=:n:JB)_ ..- Events(Threshold=35c13)

500 1000Time I s

/

100 enC~w

50

150eP

Figure 2 AE generation pattern during 4-point bending test.

20 , 2CH• 1I000

E 10 • 0

• 00

E .0 c• 0

• 0 Q.I,- 0 Ec • • •0 •0 • •• • .-... • u

B •• '. •• ...-.... • •••• • • Q.Ia.

u • • If)0-10 •-l •

•-20 1CH0 10 20 30 40Oc Events

Figure 3 Locations of Microcracks detected during 4-point bending test.

stress reached ac, microcrack locations concentrate within the narrow band afterac. It is important that the narrow band has excellent agreement with the origin offinal fracture shown in the right part of the figure. It can be emphasized that thefigure demonstrates the excellent ability of AE technique to predict the time andlocation of the final failure.

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

MICROFRACfURE PROCESS IN ALUMINA 723

Size Dependence of 0c

Figure 4 shows the size dependence of 0c for (a) coarse and (b) fine grain alumina.Although it is well known that the bending strength, 0B, of brittle materials dependon the specimen size strongly as shown in Table 2, it can be seen from the figuresthat 0c distribute around the constant values independent of the specimen size forboth materials. It is interesting that the values of Oc are same for both coarse andfine grain materials. It can be understood from Figure 4 (b) that the 0c value isenhanced and the distributing range becomes small due to the dehydration ofmaterials.

1005 10 50Effective Area I mmZ

(a) Coarse Grain Alumina (ADS-IO)

ADS-10

5x53-Point 4x44-Point

------- --- ,- ----f - ---

3x34-Point5x5

14-Point1001

300

cf200~

~167

ADs-eo• :Wet0: Dry

300....-----......,.--...,.-------,----.

4x44-Point

3x3 3-Poirit

1005 10 50EffKtlve Area I mm2

(b) Fine Grain Alumina (ADS-SO)

1001'1---------..,--.::-----~;:';;---:;(

Figure 4 Size dependence of 0c.

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

DISCUSSION

MICROFRACfURE PROCESS IN ALUMINA 72S

Microfracture Process During Bending Test

In this study, the nucleation time and locations of microcracks during bending testsof alumina were evaluated by acoustic emission technique. Then the microfractureprocess during bending test is discussed.

It has been made clear from Figure 5 that the maincrack formation occurs whenthe stress is beyond 0c. In this study, the locations of AE sources, i.e. microcrackswere determined using the longitudinal wave velocity and the difference of thearrival times between 2 ch transducers, as shown in Figure 3. It can be seen fromFigure 3 that several AE events were detected before applied stress reached 0c.

Considering the shear and surface wave propagations, although some microcrackswere detected out of the upper span apparently, it is strongly suggested thatmicrocracks have been nucleated before the stress reaches 0c.

From Figures 3 and 5, the microfracture process, during the bending test ofalumina in this study, can be concluded as described schematically in Figure 6.

aTimill

c d e f

Tensile Direction( )

DOa b c

I I II II I II

I I II I \ I I II I I

II I I I

II I I I

eFigure 6 Microfracture process during bending test.

f

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

726 S. WAKA YAMA ET AL.

1. The microcrack initiation has been occurred until the beginning point of AEdetection at the latest (b).2. Microcracks increae (c) and once the stress reaches ac, maincrack formation

occurs due to the coalescence of microcracks and/or pores (d).3. The maincrack grows (e) and finally causes the unstable fracture (f).

Consequently, it can be concluded that ac corresponds to the critical stess formaincrack formation.

Properties of Critical Stress for Maincrack Formation

From the results of the present study, it was made clear that ac is independent ofthe AE threshold level and specimen sizes. Furthermore, ac has been understoodas the critical stress for maincrack formation due to the coalescence of microcracksand/or pores before the final failure. Therefore, the ac can be expected to becomethe new materials evaluation parameter for ceramics equivalently to the yieldstrength of metals, i.e. it has the role as the measure of the structural design, pre­loading stress in the proof testings and so on. Then the properties of ac have beeninvestigated.

Figure 7 shows the statistical distribution of ac, which are obtained from 4 sizespecimens of coarse grain alumina respectively. The points are fitted by a solid linedefined by the following equation.

(3)

1.co=a;>. .75E~~:=

&-goX.o .5uo2ltuc .25'8~

O.

ADS-10

-o5x5 3-Pointo3x3 4-Point-4x4 4-Point-5x54-Point

200crc I MPa

300

Figure 7 Statistical Distribution of 0c.

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

MICROFRACfURE PROCESS IN ALUMINA 727

where B (= 0.84) and m (= 4.1) are constants. Since Equation (3) is similar toWeibull's function except for a volumetric integral, Band m correspond to the scaleand shape parameter, respectively. From the figure, it can be concluded that the ac,obtained from 4 different size specimens, distribute according to one distributionfunction.

From the view point of the structural design, it is important to understand thestrain rate dependency of the parameter. Figure 8 shows the dependence of thebending strength, aB, and 0c on the strain rate. The materials used were coarsegrain alumina with dehydration (10-4 Torr, 150°C, 60 min). The temperature (20"C) and relative humidity (60%) during tests were controlled approximatelyconstant. As the strain rate becomes larger, both 0B and 0c become higher as shownin the figure.

The strain rate dependency of 0B and 0c can be expected as the influence of thestress corrosion cracking due to the water. The Weibull plots of 0c obtained fromthe wet and dry specimens of fine grain alumina are shown in Figure 9. It has beenmade clear from Figure 8 that the distribution function of 0c is equivalent toWeibull function, therefore the slope of the fitting straight lines in Figure 9 areconsidered as the shape parameters of Weibull function. The slope of 0c of wetspecimen is 5.8 and it becomes 19 due to the dehydration. On the other hand, theshape parameters of the bending strength, 0B of wet specimen were 15-20 andthose of dry specimen were 21--28, as shown in Table 2. Therefore it can beconcluded that the influence of the stress corrosion cracking by water on 0c are

350

300

&"200tf .

---- -_ .. -

~f--- .

--rf--_.-

..- ..---_ ..------

--------l-- -

~- -- . ----. --_.--1-----------

Strain rate I -15

164

FigUre 8 Dependence of Os and 0c on the strain rate.

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

728 S, WAKAYAMA ET AL.

larger than that on UB' Furthermore, it has significant implications that only thelowest values of Uc were influenced by the dehydration.

In this study, two kinds of alumina ceramics with different microstructures wereused. It is interesting that the Uc of both materials are the same value, as shown inFigure 4. Although further study is needed on the influences of the microstructures,e.g. grain sizes, porosities and intergranular glass phase contents, on the uc, it canbe concluded that the new characterization technique evaluating the advancedmaterials parameter, uc, has been established in this study.

§ .99:sE~1-::: 50:0 .~.8~o..,1-

bn. .1c'0~

ADS-BOo:Wet6.: Dry

150O'c I MPa

200 240

Figure 9 Influence of the Dehydration on uc.

CONCLUDING REMARKS

In this study, the nucleation times and location of microcracks during bending testsof two kinds of alumina were evaluated by acoustic emission technique. Specimenswith various dimensions were used for bending tests, in order to investigate thedependence of microfracture process on the specimen size. Furthermore, thefracture process was observed using a fluorescent dye penetrant method.Consequently, the following conclusions were obtained.

1. The remarkable point, at which both AE events and energy increase rapidlybefore the final unstable fracture, was observed. The apparent bending stress at thepoint, uc, was independent of AE threshold level and specimen size.

2. From the result of the fluorescent dye penetrant observation, it was under­stood that the formation of a maincrack due to the coalescence of microcrakcs and!or pores occurs at uc. Then the microfracture process was made clear, i.e. thenucleation of microcracks, the formation and stable growth of a maincrack and thefinal unstable fracture occur respectively.

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14

MICROFRACfURE PROCESS IN ALUMINA 729

3. The physical meaning of 0c was understood as the critical stress for maincrackformation due to the coalescence of microcracks and/or pores. Additionally, theinfluences of strain rate and dehydration on 0c were investigated.

Finally, it is concluded that 0c can be the new materials evaluation parameter inceramics, which is significant equivalently to the yield stress in metals.

ReferencesI. R.G. Hoagland and J.D. Embury, J. Am, Ceram. Soc.• 63, 7-8(1980).2. A.G, Evans and K.T, Faber. J. Am, Ceram. Soc" 67.4(1984).3. S.B. Batdorf and H.L. Heinisch, Jr., J. Am. Ceram. Soc., 61,7-8(1978),4. J. Lamon, J, Am. Ceram. Soc., 71,2(1988).5, G.A. Gogotsi, A,V. Drozdov and A,N, Negovskii, Proc. Ultrason. lnt., 83(1983),6, A. Katagiri, T, Nishiyama, T. Fukuhara and Y. Nozue, Proc. National Con]. on AE, JSNDI, Tokyo

(1983).7, T. Kishi, S. Wakayama and S. Kohara, Fracture Mechanics of Ceramics, Vol. 8, R. C. Bradt, et al.

(ed.), Plenum Press, New York (1985).8. S. Wakayama, T. Kishi and S, Kohara, Progress in AE 1/1, K. Yamaguchi, et ai, (ed.), JSNDI,

Tokyo (1986).9. W,A, Weibull, J, Appl. Mech., 18,3(1951).10. G. Sines and T. Okada, J. Am, Ceram. Soc" 66,3(1983).

Dow

nloa

ded

by [

Yor

k U

nive

rsity

Lib

rari

es]

at 1

2:34

20

Nov

embe

r 20

14