dislocation pile-up (Schulson and others, 1984; Cole,. 1988) or crystal elastic ... the influence of micro cracking on ice ductile deformation, provided tests are not ...
Annals rifGlaciology 19 1994
©
International Glaciologi cal Society
Microcracking and shear fracture in ice M. A. RI ST , S .]. ]ONES A:\,D T. D. SLADE Ins filllleJor Ma rine Dy namics, .Valional R esearch Council . P.O . B ox 12093, SI J oll1l's, S ewJoundland AIB 3 T 5 . Canada
ABSTRACT.
Th e rela ti o nship betwee n mi ero crae kin g a nd ice streng th has bee n exa mined usin g tri axial a ppa ra tu s in whi ch c rac k d a m age can be inhibited by th e impositi o n of co nfining press ure. Sh ea r fr ac ture in ice is o bse ryed to be a rapid , unstable process with no a ppare nt indi ca ti o n of tensile c rac k locali sati o n or interac tion prior to failure a nd no a cco mpa n ying la rge-sc al e volum e tri c cha nges, at leas t to within lms of th e occ urrence of m ac rosco pi c fa ilure. Sh ea r fr ac ture stre ng th displays littl e o r no d ep end ence o n confin em e nt a t m od era te press ures (P = j 20 :"IPa ), a nd th ere is no evidence of signili ca nt c rack sliding befo re m ac rosco pi c fr ac ture und er th ese co nditio ns. Wh ere Oow with di stributed mi c roc rac kin g occ urs, yield stre ng th ca n a lso remain rem a rk a bl y unaITec ted b y confinin g press ure, d es pite redu ced c rac k dam age . Pa rtic ularl y und er conditi o ns wh ere mi crocrac ks a re indu ced b y pred o min a ntl y el as ti c stra ins, th ey m ay re m a in sta bl e a nd no n-interac ting even a t high \'olum e tri c d ensiti es.
INTRODUCTION
tha t occ ur durin g hig h d efo rm a ti o n ra te inte rac tion s ice noes a nd o ITsh o re stru c tures . Tri a xi a l tes tin g a lso prO\- id es a m ea ns of im-estiga ting th e influ ence of mi cro crac kin g on ice du c til e d eformati o n , provid ed tes ts a re no t co ndu cted too nea r the press ure m elting po int , sin ce hydrosta ti c press ure inhibits c rac k fo rm a ti o n but has a minim a l e ITec t o n pl as ti c !low _ H e re we prese l1l res ults fro m tri ax ial tes ts th a t prov id e insig ht into th e na ture of mi croc rac kin g as it a ITects ice streng th a nd th e sh ea r fra c ture process . b e t\~-ee n
It is frequ entl y o bserved th a t ice re tains un co mm o nl y brittl e behaviour a t hi g h ho m o logo us te mpera tures wh en compa red to oth e r polycr ys ta lline materi a ls. Ind eed , o n th e grain -bound a ry level mu c h consid e ra ti o n has bee n g iven to th e m echanism und erl ying mi croc rack fo rma ti o n du e to loca li sed te nsil e stresses, th e process bein g a ttributed to g ra in-bound a r y sliding (Sinh a, 1984) , di slocation pile-up (Schulson a nd oth ers, 1984; Col e, 1988 ) or cr ys ta l el as ti c ani sotrop y (Col e, 1988; Sund er a nd Wu , 1990) . On th e m ac rosco pi c level, te nsil e fr ac ture is rea dil y ex pla in ed since it is co nsid ered to occ ur as th e res ult of th e unsta bl e pro paga ti on o f a sin gle crac k previously form ed by one of th e a bove m echa nism s. On th e other hand , th e na ture of m ac roscopic brittl e fr ac ture a nd fa ilure und er a n ove ra ll co mpressive fi e ld is ra th er poorl y und erstood. Althoug h crac ks in ice a re known to nuclea te in compression du e to localised te nsio n , it is no t clea r ho w th ey pro paga te o r inte rac t to ca use ultima te fa ilure. Part of th e probl e m has bee n th a t th e comm o nl y used uniaxial compressio n tes t has provid ed a ra nge of fa ilure mod es including ax ia l splitting, shea r fr ac ture, explosive fr ac ture o r a mi x ture of th ese mod es, d epe nding fr equ e ntl y on th e no n -unifo rmit y of th e imposed stress nea r th e sp ec im en-m ac hin e inte rface . A prac ti cal m ea ns of appl ying a more unifo rm sta te of stress is provid ed by th e triaxi a l tes t, wid ely employed in th e ex p erimenta l exa mination of briLlle rocks, in whi ch a h ydrosta ti c press ure a nd a uni axial stress a re superimposed (princ ipal stresses 0'1 > 0'2 = 0'3 ) . ""h ere brittl e ice Ij'ac ture has bee n inves tigated using this tec hniqu e, fa ilure occ urs b y th e fo rmation of a sin gle distin c t shea r fa ult (Durh a m a nd oth ers, 1983; Murrell a nd o th ers, 1991 ) . Fa ilurc of thi s typ e is fr eq uentl y o bserved , fo r exampl e, in associa tion with oth er mec ha ni sm s und er th e compl ex stress sta tes
GENERAL POLYCRYST ALLINE SHEAR FRACTUR E Th e m ec h a ni cs of sh ear fr ac ture in polyc r ys ta llin e m a te ri a ls uncl er compress io n has rece ived parti c ular a ttenti o n with res pec t to th e fa ilure of brittl e rocks, fo r w hi c h a wid e bod y of th eo re ti ca l and ex perim ental wo rk ex ists (P a te rso n, 1978; J aege r a nd Coo k, 19 79; Sch o lz, 1990) . Brittl e fa ilure has been fo und to be strong ly d epend ent o n confinin g press ure and fo r this reaso n is co mm o nl y inves ti gat ecl using tri a xi a l tes tin g . Sh ea r frac ture d eve lo pm ent is co nsid ered to be a ve r y compl ex process in vo h 'ing th e g ro wth a nd illlerac ti o n of te nsil e mi crocrac ks (vV a we rsik a nd Brace, 19 71 ; H a llbau er and o th ers, 19 73), the complexity a rising because it a ppears th a t a n in cl in ed crac k d oes no t propaga te by shear in its own pl a ne in a brittle iso tro pi c m edium. Eve n und er a llro und co mpressio n a sliding crack will tend to fo rm m od e I (tensil e) ex tensio ns th a t a lig n th emse lves in th e direc ti o n of th e m ax imum prin cipa l stress, p a ra ll e l to 5 MPa (Fig. 3) . Pressure independence of ice brittle shear stren gth h as also been observed at -40°C for confining pressures a bove 10 MPa (Murrell and others, 199 1; Rist a nd Murrell, 1994) and at temperatures between - 11 5 and - 196°C when the confin em ent is above 50 MPa (Durham a nd others , 1983) . The 45 ° fracture orientation is also unusual, occurring in the direction of maximum sh ear stress, and is accompa nied by a very narrow fault zone. Brittle failure does not appear to be associated with an y large-scale volumetric ch anges within the specimen and neither has it been possi ble to identify any evid en ce of crack coalescence or shear localisation preceding fai lure, eith er fr om thin 136
section examina tion or from high-sp eed video observations. The pressure ind epend ence points to a mechanism that does not depend on a critical crack density and other as pects of the failure process outlined above appear at odds with the notion of interacting tensile micro cracks forming a broad damaged process zone. I t is also interesting to note that the ductile yield strength for 1 mm grain-sized ice d eform ed a t 10- 2 S- I shown in Figure 9 is close to the fra cture strength of the 5 mm grain-sized sp ecimens illustrated in Figure 3. Some of these observations maybe resolved by specula ting that the fracture process is triggered by localised plastici ty, thus explaining the almost volum e-conserving deformation , the pressure ind epend en ce of strength a nd th e direction of fracture that m aximi ses the sh ear str ess, a lth ough su ch a mechanism wou ld be unprecedented . On the other ha nd , even though it h as been d emonstrated that the sh ear fracture surface in ice propagates at least as fast as 100 m s- I, this is still an order of magnitude slower than the m eas ured tensile crack propagation speed (1000 ms- I; Sato a nd Wakaha ma, 1980) and does not preclude extremely rapid localised tensile micro crack growth and interaction immediately prior to fracture . The reloading tests on damaged ice also provide insight into th e nat ure of microcracking and in this respect th e hysteresis effect observed in Fig ure 7 is particularly informative. Closed sliding cracks, as well as open cracks or fl aws, are known to lower the effective modulus of polycrystalline m ate ri a ls. In fact , simil ar behaviour to th a t illustrated by th e loading- unloadin g curves in Figure 7 is observed in som e brittle polycrystalline rocks where the elastic modulus is reduced during loading by the presence of closed microcrac ks that und ergo sliding. Such cracks only slid e in the opposite se nse after the load has dropped sufficiently to overcome friction in th e reverse direction, producing a high modulus on initi al unloading (W alsh , 1965 ) . However, such materials contain randoml y ori ented fl aws a nd display significant permanent strains once the load is removed because the frictional stresses that d elay reverse sliding a lso prevent all th e microcracks from ret urning to their origin al relative positions. Th e predomin ant a lignment of cracking in the direction of loading co mmonly observed during compression testing of ice, and the large stresses need ed to induce sliding (see Fig. 3), as well as the a bsence of significant permanent axial d eform ation, make this an unlikely sce na rio here. Instead, it is suggested that the larger region of purely elastic behaviour during unloading is a conseq uence of the di rectionali ty of ice a nelasti c behaviour (see Duval and others, 1983) . It can be concluded that no significant crack sliding has occurred a nd the non-linearity during loading under these conditions is due almost entirel y to the co n tribution from a delayed elastic strain component, not due to microcracking. This conclusion is reinforced by the near identical behaviour during repeat loading to ultimate fracture (Fig. 7) which occurs at the sam e high stress as for undam aged ice. Although other workers h ave reported a signifi cant influence of damage on ice deformation behaviour, it is important to point out th a t not a ll of this may be du e to presence of microcracks per se . Stone a nd others ( 1989 ),
Rist and others: Microcracking and shear fracture in ice for example, induced d a mage by deforming at low strain ra te past the upper yield stress to 2% strain. Und er these conditions it is likely there will have been ex tensive permanent plas ti c d eformation of th e ice crystals as well as crack formation , crack growth , crack interaction and m ay be even grain rotation leading to large volumetric d eformation s as illustrated in Figure 6. From this it is diffi cult to isolate th e influ ence of th e presence of mi crocracks alone. The results presented here indicate that where microcrack damage is indu ced without significant permanent creep strains, subseq uen t stressstrain behaviour may be indistinguishable from th a t obser ved during d eformation of und amaged ice.
SUMMARY AND CONCLUSIONS An experimental examination of microcracking and shear fracture at hig h strain rates in ice has led to th e following con cl usions: - th e shear fai lure process is very rapid and proceeds through polycrystalline ice in compression at least as fast as 100 ms- I. - shear frac ture strength displays little or no press ure dependence und er moderate confinement even though static fri ctional forces between sliding ice surfaces a re strongl y influenced by confining pressure. - no large-scale volumetric changes are observed to precede shear fractu re a nd no evidence h as been found to indicate crack coalescence or localisatio n pri or to failure. - micro cracks re m ain closed (or d o not open significantl y) but show no evidence of significant sliding before failure occurs . Although the a bove observations do not preclud e tensile crack interaction at a very late stage to form th e observed shear fracture, they can a lso be ex plained by spec ul a ting that th e unusua l shear fractures observed here ma y have been trigge red b y localised plastic instability. G enerally, it appears that when microcrack damage is induced by predominantly elastic strains th e cracks m ay rem ain remarkably stable and non-interacting even at hig h stresses. Where ductile d eformation occurs by yield with distributed cracking, the presence of mi crocracks does not necessaril y influ ence th e failure stress, even at high volum etric d ensiti es .
REFERENCES Ashby, M. F. a nd S. D . H a ll am. 1986. Th e failure of brittle solids containing small cracks und er compressive stress states. Acta M etall., 34,497- 5 10. Cole, D. M. 1988. Crack nucleation in polycrysta llin e ice. Cold Reg. Sci. Teelmol. , 15 (1), 79- 87.
Costin , L. S. 1983. A microcrack model for the deform ation of a nd failure of brittl e rock. ]. Ceophys. Res., 88 (B 11 ), 9485- 9492. Durh am, W . B., H. C. H eard and S. H. Kirb y. 1983. Experim enta l deformation of polycrystallin e H 2 0 ice at high pressure and low temperature: preliminary results. J. Ceoplzys. Res., 88, Suppl. , Pt. I , B377-B392 . D uval, P., M . F. Ashby a nd I. Andermann. 1983. Rate-contro llin g processes in the creep of polycrysta llin e ice. ]. Phys. Chem., 87 (2 1), 4066--4074. H a ll bauer , D . K ., H . W ag ner and N. G. W . Coo k. 197 3. Some observations co ncerning the mi croscopic and mechanical behaviour of quartzite specimens in stiff, triaxia l compre sion tests. fnt. J. Rock ,Hech. Min . ci. Geomech. Ahstr .. 10, 712- 726. J aeger, J. C. and N. G. \\'. Cook. 1979. Fundamentals oJ rock mechanics. Lond on, Chap man and H a ll . K a lifa, P., P. Du va l and M . Ri card. 1989. Crack nucleation in polycrys ta llin e ice under comp ressive stress states. In Sin ha, N. K ., D. S. Sodhi and j . S. Ch ung, eds. Proceedings oJ the Eighth b,ternational
COliferCllce on Offshore Nl echanics alld Arctic Engineering, Th e Hague, The Netherlands, March 19- 23, 1989. Vol. l rr. Arctic and polar technology. New York , Ameri can Society of Mechan ical Engineers, 13- 2 1. K a lifa, P ., G . Oui ll on a nd P . Du va l. 1992. Mi crocracking and the fa ilure of polyc rys ta llin e ice under tri ax ia l compression. ]. Claciol., 38 ( 128), 65- 76. K emeny, j . M. a nd N. G. W. Cook. 1987. Crack mode ls for the fai lu re of rock under compression. In D esa i, C. S. and others, eds. Constitutive laws Jor engineering materials: theory and applications. London , Elsevier, 879887. M urrell , S. A. F., P. R. Sa mmonds a nd M. A. Risl. 199 1. Strength a nd fai lure modes of pure ice and multi -year sea ice und er triaxial load ing. In j ones, S. , R. F. M cKenna, j . Tillotson a nd 1. j ordaan,
eds . Ice-structure interaction. l U T llil1-IAHR Symposium, SI. J ohn's, NewJoundland, Canada, Aug. 14--17, 1989. Berlin. e tc. , SpringerVeriag, 339- 36 1. Paterso n, M . S. 1978. Experimental rock de.formation - the briUleJield. Berlin , etc., Springer- Verl ag. Rist, M. A . and S. A. F . Murrell . 1991. R ela tionship between creep a nd frac ture of ice. In Cocks, A. C. F. a nd A. R . S. Ponter, eds. M echanics oJ creep-brittle materials. Vol. 2. London, E lse\'ier, 355- 369. Ri st, M. A. a nd S. A. F. Murrell. 1994. I ce tri axia l deformation a nd fracture. J. Claciol. , 40 ( 135), 305- 3 18. Samm is, C. G. a nd M. F. Ashby. 1986. The failure of brittle porous solids under compressive stress states. Acla M etalL. , 34, 5 11 - 526. SaLO, A. and G. vV a kah a m a. 1980. Measurements of th e veloci ty of crack propagation in ice. Low T emp. Sci .. Ser. A 39, 193- 195. [InJ a pa nese. l Scholz, C. 1-1 . 1990. The mechanics if earthquakes and Jaulting . Cambridge, Camb ridge Un iversity Press. Schu lso n, E. M ., P. N. Lim a nd R . W. Lee. 1984. A brittl e to du ctil e transition in ice und er tension. Philos. Mag. , Ser. A. 49 (3), 353- 363. Sin ha , N. K. 1984. Intercrys ta lli ne cracking, gra in-bo und a ry sliding, and d elayed elasticity at high temperatures. J. Maler. Sci., 19 (2), 359-376. Stone, B. M ., l.j . j ordan, S. J . J ones and R. F. M cKenn a . 1989. D a mage of iso tropic polycrystallin e ice und er moderate confining pressures. III Axelsson, K . B. E. a nd L . A. Fra nsso n, cds. POAC 89. The
10/h International Conference 011 Port and Ocean Engineering under Arc/ic Conditiolls , June 12- 16. Lulea, Sweden. Proceedings. Vol. 1. Lulea , Tekniska H 6gskolan i L ulea., 408--419. Sunder, S. S. and M. S. Wu . 1990. Crack nucleation due LO elastic a niso tropy in po lycrystallin e ice. Cold Reg. Sci. T echnol., 18 ( 1), 29- 47. Wa lsh , j . B. 1965. Th e effect of cracks on the uni axia l elastic co mpressio n of rocks. ]. Ceoph)'s. Res. , 70 (2), 399--4 11. W awersik, "V. R . and ' V. F . Brace. 197 1. Post-fai lure behaviour of gra nite and diabase. Rock M ech. , 3, 6 1- 85.
The accuracy of riferences in the text and in this list is the responsibility of the authors, to whom queries should be addressed.
137
