test in Lac du Bonnet granite - Civil Engineering - University of Toronto

CCNBD of Lac du Bonnet granite. The rock sample(diameter=20 cm, thickness=5.1cm) is fromCanada's Underground Research Laboratory wherethe concept of deep radioactive waste disposal isbeing studied. This CCNBD test was designed tomeasure the fracture toughness at the tip of a “V”-shaped notch as well as the AE response to theinitiation and propagation of tensile fracturing, (Fig.1).Fig. 1. The CCNBD specimen of Lac du Bonnet granite in aMTS loading frame with 20 transducers attached to it for 3DAE coverage.Ultrasonic data was recorded on 16 channels,sampled at 10 MHz. Each triggered event consistsof 4096 sample points, equating to a 0.2 ms timeperiod. The hardware has the potential to capture 20AE events per second and experiencesapproximately 50 ms of recording downtime afterevery event, when the AE data is written to the PC.The Giga RAM Recorder was designed in responseto this downtime limitation. Ultrasonic waveformsare recorded continuously onto the 40 GB RAM,which when full, begins to overwrite the oldest data.Therefore the RAM always contains the previous268.4 seconds of ultrasonic waveforms, which canbe saved for later download at any time. Therecorded ultrasonic signals passed through external20 dB pre-amplifiers (PAC model 1220A). Threechannels i.e. 11, 13 and 15 were amplified at 40 dBin order to be able to record higher amplitude eventsduring the test. (see conceptual diagram, figure 2).Four transmitters were dedicated as active sourcesfor velocity surveys and excited using aPanametrics 5072PR transmitter box. An ESGinterface box switched between the fourtransmitters. The signal to noise ratio was improvedby stacking. In this study we are not reporting onthe velocity survey results obtained during loadingof specimen.The displacement, force and time base of theexperiment were recorded with MTS TesStarsoftware using another PC. Both the PCs weresynchronized through a GPS system.2.2. Sample preparation for fracture toughnesstestA special large fixture was designed and fabricatedat the workshop of department of civil engineering,University of Toronto which helped to ensure thatthe chevron notches are exactly in the centre of thedisc and the geometrical dimensions conform to thegiven tolerance. A circular diamond saw was usedto cut the required notch. Lac du Bonnet graniteshowed strong anisotropy once wave velocitieswere measured along various directions beforemachining the notches in the specimen. Figure 3shows the direction of measured wave velocitiesprior to notch and the results are tabulated in thetable of the same figure. The wave velocitymeasured along AA’ line being around 5290 m/swas found to be the fastest direction and EE’ wasfound to be 4550 m/s. Wave velocity measurementsusing pulsing method and lead break method wasused for sensor location calibration prior to testing.Transversally isotropic model with P- wave velocityof 4800 m/s in the notch direction and anisotropyratio of 0.75 was set for slow direction, in responseto velocity measured perpendicular to the notchorientation. Taking into account the sensor thatprovide the best 3D coverage around notch A, thelocation error was found to be around 3 mm. Thegeometry of CCNBD is illustrated in Figure 4 andin table 1. ISRM suggested method [19], was usedto prepare a large size (diameter=20 cm,thickness=5.1 cm) cracked chevron notchedBrazilian disc. The advantage of this method is thecore axis can be oriented to any direction withrespect to any anisotropy features such as planes ofweakness or preferred micro-fractures. The fracturetoughness (K IC ) of the specimen shall be calculatedby the following formula:

Ultrasonic WaveformsPre-AmpsDiscreteTriggeredDataGiga Recorder20 AE eventsPer secondPCSampleContinuousDataRecord1 x 260SecondFig. 2. Conceptual diagram of experimental set up used to record ultrasonic data, (active source for velocity surveys not shown).Scan LineA-AB-BC-CD-DE-EF-FH-HG-GA -AVelocitym/s529051705190470045504640499050705300BAGCHDFE E ’F ’D ’H ’C ’G ’B ’A ’Y = µ e(2)*. α1min.υwhere µ and ν are constants determined by α 0 , α Bonly using a separate table.Y*min= Critical dimension less intensity valueP max =maximum load at failureAll the dimensions of the geometry should beconverted into dimensionless parameter withrespect to the specimen radius R and diameter D asshown in table 1.Fig.3 Anisotropic wave velocity measured in Lac du Bonnetgranite.Pmax *K IC= Ymin(1)whereBRTable 1. CCNBD geometrical dimensions in this testDescriptionsDiameter D (mm)Thickness B (mm)Initial cracklength a o (mm)Final crack lengtha 1 (mm)Saw diameter D s(mm),Y*min(dimensionless)Valuesmm194.851.1016.6756.08147.40.76Dimensionlessexpressionα B =B/R = 0.52α 0 =a o /R= 0.17α 1 =a 1 /R=5.75α S =D s /D= 0.75Fig. 4. The CCNBD specimen geometry with recommendedtest fixture.

2.3 Microstructural InvestigationMicrocrack observational techniques include 1) dyepenetration prior to thin section preparation, 2)radiography and X-ray, 3) SEM and 4) TEM, [20].In recent years development of computer aidedimage analysis programs has greatly facilitatedmicrostructural characterization through analysis ofdigital images obtained from the thin sections, [1,21, & 22]. These techniques are based on directmeasurements of crack length, orientation, grainsize and shape measurements from thin sections.These methods provide easier handling of largeramount of data collected and therefore provide amore representative assessment of microstructuralproperties. Six large thin sections (7cmx5cm)perpendicular to the fault trace were prepared fromeach notch. About 70 digital images taken fromeach thin section were tiled together. In this studyonly the result of one thin section from each zone(notch A, notch B and far field) is shown.Microfracture density and mean orientation weremeasured from notch A and B and compared withfar-field microfracture densities obtained from thinsections prepared from the edge and far from theloading area. Microstructural properties such asmicrofracture densities, orientation, microfracturelength and width were measured to evaluate thedimension of the process zone.3. EXPERIMENTAL RESULTIn this paper the AE data for only one of the notchin CCNBD, i.e. notch A (the one in contact withmoving platen) is analyzed to present hypocenterlocation and focal mechanism based on first pulsemotions polarities. In all about 2500 events werestored in the Giga recorder which was first autoprocessed for their temporal and special analysis. Inorder to minimize the hypocenter errors only six toeight transducers which were found to have the best3D coverage around notch “A” were used tomanually locate about 350 events during pre to postfailure time period. However microfracture analysisinvolves both the notches. With the sudden drop ofthe load at failure the test was terminated and as aresult fracture propagation was seized in somesection close to the edge of notch B, as confermentlater by thin section studies. This provided a rareopportunity to study the fracture process zonelength in front of the tip of the propagating macrofracture.3.1. Hypocenter distribution and fracture growthFigure 5(a) reveals the variation of all the AE hitsper second and force versus time of loading for theCCNBD of Lac du Bonnet granite recorded fromboth notches. Stress intensity factor at the criticalload leading to the failure was measured usingISRM suggested method [19], is shown in thefigure. Cumulative detected number of AE eventsand load versus time is shown in 5 (b). Loaddisplacement graph is shown in figure 5(c) wheremaximum displacement at peak strength wasrecorded to be 0.28 mm.A close up window of various stages of failurebased on maximum AE heat/s recorded at differentstations around the notches is demonstrated infigure 5(d). The evolution of AE events and theirlocations with time during tensile fracture initiationand propagation at various phases in notch A isdemonstrated in Fig.7.3.2. Fault growth stages3.2.1. Stage A:Except few isolated events were located in thebeginning of the test, but the first noticeable eventsat this stage is related to a linear and progressivemicrofracture initiating at a distance of about 5mmahead of the tip of the notch extending up to alength of 14 mm. This corresponds to an area ofabout 50 mm 2 (10.5 mm x 4.7 mm) of rectangularshape. This period starts from time 553 th secondsand continues till the end of 562th seconds(15:41:30 to 15: 41:39), as shown in figures 5(d)and 6 (Stage A). The load rises at the displacementrate of 0.0005 mm/s to 32.02 KN and stressintensity factor (K I ) measured at the tip of the “Vshaped”notch reaches up to 1.48 MPa.m 1/2 which is95% of the sample’s fracture toughness. At this load0.267 mm of displacement (inbuilt MTS loadingmachine’s LVDT) is attained which is also 95% ofultimate displacement before failure. Total numberof AE events located manually around the tip andthe plane containing the notch is about 13 events.3.2.2. Stage B:This period extended from 562 th second to 563 thsecond where load increase to 32.32 KN anddisplacement measured is of the order of 0.278 mm.Calculated KI at this load was found to be 1.52

(a)(b)(C)(d)Fig. 5. (a) Applied load and AE hits/second versus time, fracture toughness value for critical stress intensity factor is marked in thefigure. (b) Applied load and cumulative AE versus time, (c) Variation of load with displacement for CCNBD at failure and (d) Aclose up window of Load and AE hits (sensed at various locations ) versus time for fracture initiation and dynamic failureMPa.m 1/2 (97.7% of critical stress intensity factor).Total of 40 localized AE events were located in thisperiod marking almost same rectangular shapedarea advancing about another 5 mm outwardstowards the loading edge of the specimen (Fig. 6,stage B). From AE density map it can be inferredthat AE concentration increases at a distance of 10mm from the fracture tip extending for 12 mmlinearly outwards towards the region where thenotch ends at the base of the wedge. Still fracture isgrowing quasi-statically at this stage and remainsstable.3.2.3. Stage C:This period extended for four seconds starting from563 th to 567 th second when load is increased to amaximum level of 32.72 KN and stress intensityfactor, K I attains highest value of 1.55 MPa.m 1/2becoming equal to the fracture toughness, K IC ofLac du Bonnet granite. Highest displacement of0.28 mm is experienced at this failure point. As it isclear from the spatial occurrence of 93 AE eventsand its calculated density in figure 6 stage C, whenfracture propagates laterally in all directionsunstably. Figure 6 illustrates various stages oftemporal and spatial fracture evolution along twoplanes of XZ and XY i.e. parallel and perpendicularto fracture propagation directions respectively. Thefracture propagates unstably and dynamically themoment the stress intensity factor, K I reaches K ICand the length of the growing fracture and the areainvolved exceeds its stable conditions. It is evidentfrom the added widely distributed numerous AEevents at this stage that microfracture interactionand coalescence was facilitating unstable growth ofthe fracture mainly under tensile mode.3.2.4. Stage D:This stage is characterized by a sudden drop of loadas quasi-statically stable fracture propagation leadsto unstable dynamic fracture propagation. The loaddrops from 32.72 to 21.6 KN in a period of timebetween 568th second to 570 th second i.e. from time15:41:43 to 15:41:45. At this stage71 AE events arerecorded from similar areas to previous stage asrevealed by AE density couture map. This stage canbe considered as the beginning of the post failurezone as well. Generally it is sited that AE

hypocenter distribution reflects directly thedistribution of cracking in rock under stress. In thisstudy AE density maps (Fig. 6) are used to study thefracture initiation and propagation from the tip of anotch in its own plane with a process zone ofintense cracking. The time interval between earlyappearance of AE close to the notch tip and furtherprogression of fracture front and later on finalfracturing was approximately 14 seconds. Migrationof the front of rapidly increasing cracking along thefailure plane prior to the dynamic rupture can beconsidered as quasi-static rupture velocity [5 & 6].using similar approach it was found that fracturenucleation in front of the notch grew to 2.8 cm/sand reached approximately a volume of 6.5 cm 2 x1.25 cm = 8.12 cm 3 quasi-statically before itpropagate dynamically in stage B. At this levelstress intensity factor became 97.7% of its criticalvalue. In stage C fracture front extends dynamicallyat much faster rate laterally and towards the edge ofthe specimen. It is also noticed from figure 5(d)that the maximum AE activities are recorded atcouple of channels in middle of stage C, indicatedas well by AE density map from entire failure plane(XZ) where the fault reached the edge of thesample, yet the dynamic rupture happens in themiddle of stage D, about 2 second latter. It isevident from the AE density couture pattern thatduring this time interval, intense cracking tookplace from the area where triangular wedge shapednotch ends. Rupture of larger volume of rock i.e.21.3 cm 2 x 1.2 cm = 26 cm 3 , (process zonecalculated from the AE distribution andmicrofracture density of the larger rectangular zone)takes place at stages C and D.3.2.4. Stage E:Stage E starting after the main rupture of the notchA and marks the beginning of the post failureactivity. This stage starts minute shear failure as thefailed surfaces slide past each other. Figure 6 (stageD), shows the spatial distribution of the AE eventsand its density map from this stage. Clustering ofthe AE events happens to be from the same areawhere maximum fracturing was experienced inprevious stage. Figure 6 (stages A), shows thecumulative AE events and its density maprepresenting all the events from the beginning of thetest till the end of the stage E. The highest densityloci were found to cover an area of 19.3mm X 13.18mm at a distance of 18 mm from the fracture tip.5. FOCAL MECHANISMFor manually located 220 AE events occurring up tothe end of stage D, i.e. before receiving post failureevents, focal mechanism based on first motionpolarity was calculated [4]. Using similar methodAE events manually located (135 events) from stageE i.e. post failure events have been calculate to findout if these two sets of events could bediscriminated based on their focal mechanism. Thepolarity value of a single event is calculated by:K1iPol = ∑ sign(A1)(3)Ki=1where K is number of channels used for hypocenterdetermination. In this analysis based on pencil leadbreak and artificial pulses used for sensor positioncalibration, compressional pulses are characterizedwith positive polarity. Analysis of pre failure events(220 events) revealed that 75% of the total events(159 events) are of tensile type as their polarityvaried in the range of 0.25< Pol ≤1. 15% of the totalevents were found to be of shear type with theirpolarity varying in the range of 0.25≤ Pol≤-0.25.The rest (10%) had their polarity in the range of -0.25 ≤ Pol ≤ -1, believed to be of implosion type.On the contrary first motion polarity analysis of thepost failure events (134 events) reveals the fact that40% of the events are of shear type versus only 39%of tensile type. 15% prove to be of implosion type.6. PROCESS ZONE AND FAR FIELDMICROFRACTURE ANALYSISThe fracture process zone (FPZ) in rocks is definedas the region affected by microcracking andfrictional slip surrounding the visible crack tippropagating under stress, [13 & 16]. The width ofFPZ is defined as the longest distance betweenvisible cracks on either side of the macro-fractureand/or fault and it’s length is defined as the lengthbetween the fault tip and the crack with the greatestdistance in front of the fault tip.

YZXXStage A: 15:41:30 to 15:41 39, 13 eventsStage B: 15:41:39 to 15:41: 41, 40 eventsStage C: 15:41:40 to 15:41: 43, 93 eventsStage D: 15:41:43 to 15:41 45, 71 events(a)Stage E: 15:41:45 to 15:42 08, 135 eventsAE densityStage A to D, 15:40:35 to 15:41:45, 220 events10 mmLocation magnitudeFig 6. Hypocenter locations of 350 AE (dots) and their density maps shown during various stages of brittle fracturing as stressintensity factor KI reaches its critical value of fracture toughness at the tip of notch A in Lac du Bonnet granite.

6.1. Micro-structural data form Notch A:Figure 7 shows the mapped macro-fracturedeveloped in front of the notch with smallerfractures close to it and away from it. Grey Lac duBonnet granite is relatively large a grain granitewith grain size of 1-10 mm. The width of thefracture process zone optically evaluated form thethin section was found to be around 20-25 mm andthe original length of the macro-fracture is 85 mm.Cumulative AE events (Stages A_E in Fig. 6) aresuperimposed on this thin section wheremicrofracture and macro fractures have beenmapped. It was not possible to observe the length ofthe fracture process zone as the macro fracture inthis notch has reached the edge of the specimenalong its plane of propagation. But the length ofFPZ based on AE event location density map wasfound to be around 8 mm from the tip of the notch(state A in Fig.6). Moving towards the macrofracture,microfracture density increases from abackground value of 4.7 cm/cm 2 up to 10.7 cm/cm 2in the red box on either side of macro-fracture. Themicrofracture density comparison and theirorientations (rose diagram) measured from differentarea is shown in figure 8. The orientation anddensity analysis of microcracks were carried outsystematically parallel to the direction of macrofractureat increments of 2.5 mm width movingoutward from the micro-fracture trace towards theouter edge of the thin section. In figure 8 the spatialanalysis of microfracture density and orientation isillustrated. Far field micro-fracture orientation ismarked by several maxima in various orientationswhereas microfractures within the process zonedemonstrate a single maximum parallel to themacro fracture direction.6.2. Micro- structural data from notch BThe background fracture density being around 4cm/cm 2 increases to 11 cm/cm 2 in the process zone.Fracture propagation was seized in some sectionclose to the edge of notch B as the test wasterminated after sudden drop of load from 34 KN to20 KN. This provided a rare opportunity to studythe fracture process zone length in front of the tip ofthe arrested macro-fracture (Fig. 9). In the fractureprocess zone length, microfracture density wasfound to be equal to 44 cm/cm 2 which is inagreement with former process zone length fracturedensity studies [1]. Substantial increase of themicrofracture density observed at the tip of anarrested macro fracture is explained [23]. Theymentioned that dissipation of fracture energy nearthe tip of a crack coming to end makes local stressesnear inhomogeneities (in this case such as preexisting micro fractures in quartz) more important.As a result the crack begins to ramify, thus creatinga fractal structures that consumes larger energy andeventually gets arrested. Microfracture orientationanalysis of the FPZ and FPZL in notch B indicatesdouble maxima with one maximum parallel to mainfracture and the other one orientated at an angle of25 degrees form it. Far field micro-fractureorientation analysis is marked by several maximaoriented at various orientations.7. MICROFRACTUE DENSITIES COMPAREDTO AE DATAFigure 10 shows the microfracture density and AEnumber plotted versus the distance to the macrofracture trace. This plot is prepared based on similarincremental distribution of microfracture densityand AE number as a function of distance from themacro fracture trace line. Logarithmic trend ofincrease in densities of micro-fracture and AEevents confined to the immediate surroundings ofthe macro-facture is evident form the figure. Thenumber of AE drops off with distance to zero levelin the background micro-fracture.8. DISCUSSIONAcoustic emission localization and microfracturedensity analysis has been commonly used as adiagnostic tool for characterization of fractureprocess zone in geomaterials. AE hypocenters areused to investigate spontaneous fracture nucleationand its quasi-static growth from a notch tip underMode- I in CCNBD specimen of granite underconstant stress loading. AE density contour mapsillustrate the temporal and spatial correlation ofevent clusters during the quasi-static growth undertension even up to dynamic fracture propagation,(Fig.6 stage A-D). AE technique was found viablemethod for estimating the length of FPZ in testsperformed under Mode-1 [16, 17 & 18] and Mode-II [1, 3 & 4] in various geomaterials. Earlier testperformed under Mode I indicated that the intrinsicprocess zone developed near failure and had adistinct size for a particular rock type which’s widthappeared to be a martial property related to grainsize [15 & 16].Based on AE density contour map, we observed thatFPZ length ahead of a notch in Lac du Bonnetgranite is about 5-8 mm and its width varies along

Fig.8. AE (red, green and orange, 355 dots) events superimposed over micro and macro fractures traced from thin sectionprepared from notch A.Fig.9 Microfracture density and orientation (rose diagram) measured in process zone and farfield area in a in notch A. Theextended red box shown the width of the FPZ and AE events are superimposed to show the width of the process zone..the macro fracture up to a maximum of 20 mmacross. We as well noticed that fracture nucleationin front of the notch grew with the speed of 2.8 cm/sand reached approximately a volume of 6.5 cm 2 x1.25 cm = 8.12 cm 3 quasi-statically before itpropagate dynamically in stage B. This study showsthat the FPZ width based on the microfracturedensity analysis and distribution of AE isapproximately the same and increasessymmetrically and logarithmically as a function ofdistance towards the macro-fracture (Fig 10).Microfracture density analysis from the testsperformed under Mode II in rocks [1, 4 & 24] andin nature from major faults, [13 & 14] indicate anasymmetrical logarithmic increase towards the faulttrace.

Fig.10 Microfracture density and orientation (rose diagram) measured in process zone and farfield area in a seized fracture innotch B. The extended red box shown the width of the FPZ and smaller red box in front of the fracture contains the muchmicro-fracture density in FPZL.9. CONCLUSIONNumber of AE140120100806040200AE EventsMicrofracture dinsity-17.5 -15 -12.5 -10 -7.5 -5 -2.5 0 2.5 5 7.5 10 12.5 15 17.5Distance from fault trace in notch A, mmFigure 10 - Process zone- related micro-fracture density (leftordinate) and AE number distribution (right ordinate)perpendicular to the trace of main tensile fracture innotch A.Micro-fracture orientation analyses of shearfractures also reveal that the compressionalmicrocracks exhibit a smaller angle with respect tofault propagation direction and larger angle ofdilational microcracks with fault planes. Suchphenomenon is not observed in this study mainlydue to the fact that the main mechanism of fracturepropagation is tensile stresses under Mode I.121110987654Microfracture density, cm/cm 2Detailed microstructural analysis using AE andpetrographical techniques were performed onexperimentally induced fractures that propagateddominantly under tensile Mode. AE techniques canbe used to delineate the dimensions of the fractureprocess zone in front of a “V-shaped” notch inCCNBD specimen. Unlike what is usuallyobserved in shear fracture tests, AE distribution andmicrofracture densities are shown to be in closeagreement to one another and increaselogarithmically and symmetrically towards themacro-fracture.Focal mechanism analysis (based on first motionpolarity method), of pre failure events reveals thatabout 75% of events are of tensile type, 15% are ofshear and 10% belong to implosion type. On thecontrary 40% of the events in post failure region areof shear type, 39% of tensile type and 16% werefound to be of implosion type.Based on AE density contour and focal mechanismanalysis a tensile fracture initiated at a distance of5-8 mm ahead of the notch A and propagated withthe velocity of 2.8 cm/s quasi-statically. Thedynamic propagation and rupturing happened due to

unstable growth of expanding tensile fracture whenit’s size grow to a volume of elongated shape (8.12cm 3 ). At this time K I = K IC =1.55 MPa. (m) 1/2 at tipof the notch. Within ∼ 14 second initiation anddynamic rupturing of the tensile macro fracture tookplace.Acknowledgment:The authors wish to acknowledge the contributionof Dr. David Collins in using INSITE softwareduring the analysis of the data and many otherhelpful discussions.References:1. Moor, D.E. and Lockner, D.A. 1995. The role of microfracturingin shear-fracture propagation in granite. J.Structural Geology. 17: 96-114.2. Lockner, D.E., Byerlee, J.D., Kuksenko, V.,Ponomarev, A. and Sidorin, A. 1991. Quasi-static faultgrowth and shear fracture energy in granite, Nature.350:39-42.3. Zang, A. Wagner, C., Stanchits, S., Dresen, G.,Andresen R., and Haidekker, M.A. 1998. Sourceanalysis of acoustic emissions in Aue granite coredunder symmetric and asymmetric compressive loads. J.Geophys. Res. 135: 1113-1130.4. Zang, A. and Wagner, F.C., 2000. Fracture processzone in granite. J. Geophys. Res. 105:23651-23661.5. Lei, X, Kusunose, K., Rao, M.V.M.S., Nishizawa, O.,and Satoh, T. 2000. Quasi-static fault growth andcracking in homogeneous brittle rock under triaxialcompressive using acoustic emission monitoring. J.Geophys. Res. 105: 6127-6139.6. Lei, X, Nishizawa, O., Kusunose, K., and Satoh, T.1992. Fractal structure of the hypocenter distributionand focal mechanism solutions of AE in two granites ofdifferent grain size. J.Phys. Earth, 40: 617-634.7. Dieterich, J.h. 1992. Earthquake nucleation on faultswith rate and state-dependent friction. Techtonophysics.211:115-134.8. Ohnaka, M. and Kuwahara, Y. 1990. Characteristicfeatures of local breakdown near crack tip in thetransition zone from nucleation to dynamic ruptureduring stick-slip shear failure. Techtonophysics. 175:197-220.9. Ohnaka, M. 1996. Nonuniformity of the constitutivelaw parameters for shear rupture and quasi-staticnucleation to dynamic rupture: A physical model ofearthquake generation process. Proc. Natl. Acad. Sci.U.S.A. 93: 3795-3802.10. Kato, N., Yammamoto, H. and Hirasava, T. 1992.Strain-rate effect on frictional strength and the slipnucleation process. Techtonophysics. 211:269-282.11. Lockner, D. E. 1996. Brittle fracture as an analog toearthquakes: can acoustic emission be used to develop aviable prediction strategy, J. Acous. Emission, 14:88-101.12. Scholz, C.H., Dawers, N.H., Yu, J.-Z. and Anders,M.H. 1993. Fault growth and fault scaling laws:Preliminary results, J. Geophys. Res. 98:21952-21961.13. Vermilye, J.M. and Scholz, C.H. 1999. Faultpropagation and segmentation: insight from themicrostructural examination of a small fault. J.Structural Geology. 21: 1623-1636.14. Wilson, J.E. Chester, J.S. and Chester, F.M. 2003.Microfracture analysis of fault growth and wearprocesses, Punchbowl fault, Sand Andreas system,California. J. Structural Geology. 25: 1855-1873.15. Zietlow, W.K., and Labuz, J.F. 1998. Measurement ofthe intrinsic process zone in rock using acousticemission. Int. J. R. Mech. Min. Sci. 3: 291-299.16. Labuz, J.F. Shaw, S.P. Dowding, C.H. 1987.Thefracture process zone in granite: Evidence and effect,Int. J. R. Mech. Min. Sci&Geomech. Abstr.24: 235-246.17. Mihashi, H. and Nomura, N. 1996. Correlation betweencharacteristics of fracture process zone and tensionsofteningproperties of concrete. Nuclear Engineeringand Design, 165:359-376.18. Koji, O and Hidehumi, D. 2000. Fracture process zonein concrete tension specimen, Engineering FractureMechanics, 65: 111-13119. ISRM Testing Commission, Suggested method fordetermining Mode I fracture toughness using crackedchevron notched Brazilian disc (CCNBD) specimens.1995. Int. J. Mech. Min. Sci. 7 Geomech. Abstr. 1: 57-64.20. Kranz, R.L. 1983. The role of micro-cracking in shearfracture propagation in granite. Techtonophysics, 267:91:119.21. Nasseri, M.H.B. Mohanty, B. Prasad, U. 20002.Investigation of micro-structural properties of selectedrocks and their effect on fracture toughness. NARMS,Toronto, Canada2: 961-968.22. Nasseri, M.H.B. Mohanty, Robin, P.-Y.F. 2005.Characterization of microstructures and fracturetoughness in five granitic rocks. Int. J. Mech. Min. Sci.7 Geomech. Abstr. In Press.23. Chelidze, T. Reuschle, T and Gueguen, Y. 1994. Atheoretical investigation of the fracture energy ofheterogeneous brittle materials. J. Phys.: Condens.Matter, 6:1857-1868.24. Janssen, C. Wagner, F.C. Zang, A and Dresen, G. 2001.Fracture process zone in granite: a microstructuralanalysis. Int. J. Earth Sciences, 90:46-59