Reprinted from

TECTONOPHYSICSINTERNATIONAL JOURNAL OF GEOTECTONICS AND THE

GEOLOGY AND PHYSICS OF THE INTERIOR OF THE EARTH

'Iectonophysics 280 (1997 ) 83- 106

Temperature- and strain-rate-dependent microfabric evolution inmonomineralic mylonite:

evidence from in situ deformation of norcamphor

Marco Herwegh v", Mark R. Handy ", Renee Heilbronner "

~ GeO/OKischt's lnsünn. ßallcerstraßt' I. U" i."ersi/ii/ Bem, CH·30 J2 Bem. Swi/:erl anJb lnstiuu für Geowisst"lIchaft;m. Univen iliil GieJu ll. Gin sen. Gumany

~ Geologisches lnsunu urnf Abu' i/ unK f ür Win erl.lchafl liche Ph% graphie, Uni,'ersiliil Ba.lel. Bau l. Swnzertand

Received 27 January 19%: a~'Cepled 31 Ocrober 1996

~ELSEVIER

TECTONOPHYSICSEdit or s-in -Chief

J,·P. BRUN

1. ENGELDER

KP, FURLONG

F.WENZEL

Ho no rary Ed ll orS:

Editoria l Board

Universite da Reones, Institut da Geologie, Campus da Beauheu. Ave . du General Leclere, Rennes 35042 Cedex France.Phone: +33.99.28 61 23 : FAX : +33.99.28 67 80: e-mail: dirgeosc@univ ·rennesl,Ir

Pennsylvania sieie University. ~tege of Eanh & Mineral scerces. 336 Oeike Buildil'l9. University Park, PA 16802. USA.Pnoee. + 1.814 .865.362(1,1466.7208; FAX : +1 .814.663.7823; e-mail : e~der@geosc,psu,edu

Pennsylvania stete University. Department 01Geosciences. 439 Deike Build ing. University Park, PA 16802, USA. Phone+1.814.863 .0567; FAX: +1.814.865.3191 ; e-mail : kev in@geodyn.psu,e<lu

Universität Frideridana Kar1sruhe, Gaophysikalisches Institut, Hertzstraße 16.Bau 42. 0·76187 Kartsruhe, Germany_ Phone:+49 .721 .608 4431 : FAX: +49.721 .711173: e-mail : fwenzel@g~wapl .physj k. l.mi-kar1sruhe.de

M. FnedmanS. Uyeda

D.l. socesco. Pasadena, CAH.G, Ave Lallemant, Hccston, T XE. garoa, BarcelonaZ, Ben·Avraham, Tel AvivH. Ele<'ckhemer, Koenigste inC. BloI , SOllies·PonlG.C . Bond. Palisades, NYG.J . ecoaoane. Thunder Bay, Ont.BC. 8urchfiel , Cambridge, MAK C . Burke. HouSlon, T XS. Cloelingtl, AmsterdamP,R Cobbold, RennesD. Denharn. Canberra, ACTJ,F. Dewey, OxlordG,H . Eisbacher. KarlsrutleE.R Engclatll, Denver, COE.R FIOtl, KielK. Fujrta. EastLansing. MIY. Fukao, TokyoR Geiler, TokyoJ.·P, Gratier, GrenobleA,G. Green, ZOnctlRH Groshong. Jr.,Tuscaloosa, ALHK Gupta. Hyde rabad

T.W,C. aaoe.College Slabon, TXA. Him, PansF.H0fV8l tl, BudapeslE.S. Husebye, BergenH. Kanarrort,Pasadena, CAS, Karate, Minneapolis, M NRJ. Knipe, LeadsM. Kono . TokyoX , Le Pichon, ParisG.S, uster,Claylon, vc.Ri. Mada nage , ParisY. Man, HailaM. McNutt, Cambridge, MAWO. Means, Albany. NYK. Mengel , Clausthal-Zell erfe ldA. Nicol as, Montpell ierG. DeneI, Los Angeles, CAA. Perez-Estaun, OviedoH.N. Pollack, Ann Arbor, MIC.McA. Powell, Ned lands. W.AL, Ratsctlbactler, WOrzoorgE.H. Rutter, Mand1eslerM.P. Ryan, Reslon, VAD.J. seroerseo. $outtlampton

S.M. Sctlmid, BaselW.M. Schwerdtner, Tororl1O, öo r.C. ~engör, IstanbulT. sero. Tok yoShi Yang-Stlen , NanjingN, Sloop, Stanlord, GAS. SoboIev, PotsdamCA Stein, Chicago, ILP. Suhadolc, Tna steK. Tarr.aki, TokyoM. Terne, Barcelonaer.Trilu, Kings ton, 001­J. Tullis, Providence, RID.l. TufColle, nrece. NYBA van der Plu ijm. Ann Arbor, MIR. van der Voo, Ann Arbor, M IB.G, VendevHle, Auslin, TXR.l.M, veseo. UtrechlJ,S. Watkins, College Sta lion, TXH-R Wenk, Berkeley, CAG. Weslbrook , BinninghamB,F. Windley,leicesl erM.J.R Wonel. UtrechlPA Ziegl er, Binningen

Sc ope 01 Ih e journalTectonophysics rs an ir nematoner med ium lor me publ ication 01 original stud ies and comprehensive reviews in Ihe field 01 geolectonics and tbegeology and pbyscs cf me eartn's crusl and uueocr. The eosos will end eavour to maintain a high scien lific rever and it rs hoped that wi th üsinternational coverage me jo urnal will con tributa to tne sound development of this lield.

(Text continued on inside back rover)

(l 1997, ELSEVIER SGIENCE B.V. AU AIGHTS AESERVED. 0040-1951/971$17.00

Th is ioumal and the individual contributions ccotaoec in it are prolected by ne copyright 01 Elsevier sceoce av..and the follo wi09 tenns andconditions apply to their uee :

Pholocopying: Sing le photocop;es cf single emctes may be made lor personal use as allowed by nafional copyrighl taws. PenniSSion of tnepubHsher and payment of a lee is requ ired lor a ll other photocopying , inchlding multiple or sysiernanc copyi09. copying for advertising or promolionalpurposes. resere . and all forms of documenl delivery, scecer retes are available for educational institutions fhat wistl lo ma ke pholocopies fornon-prolit educabonal ctassroom useIn ftle USA, users may clear perm issions and make paymentlhrougtlltle Copyrigtlt Clea rance Center, Inc. , 222 Aosewood Drive, Danvers. MA 01923,USA. tn ltle UK, usan; may elear perm issions and make paymenlS ttlrougtl the Copyrigtlt Licensing Agency Rapid Clearance Service (CLARCS), 90Tollentlam Coun Road, Landon W1 P OLP,UK In other COllnlries where a local copyrighf clearance cenfer ex ists, please eanfact if for inlorma tion onrequired perrnisssions arld payments.Deri vallve Works : Subscnben; may reprod uce lables of contents or prepare lists 01anicles includ ing abstracts for intem al circulalion wiltli n Iheirinstitutions. Perm iss ion 01l he Publ isher is required lor lesale or d istribution outside ftle instifution.Permission of lhe Pub!isher is required lor all ottler denvafiva werks, incl uding compilations and translations .Eleclronic Siorage: Penniss ion of ttle Publ istler is requ ired to store elect ronica lly any matenal contained in this joo mal, includ i09 any anicle or panot an anicie. Confactlhe Publistler at Ihe address indicated.Excep t as OOllined above, no pan 01 lt1is publication may be reproduced, stored in a retrieval system or transmitted in any form or by any me"lnselectronic, mechanical . ptlotocopying, recording or otheN/isa , wrthouf wntten pennission cf Ihe Publ istler.

Noflee ; No responsibility is assumed by Itle Publisher for any injury and/or damage to persons or property as a maller 01produCls liability. negligenceor oltlerwise, or from any use or operalion 01any meltlods. prodllClS, instruc lions or ideas contained in the malenal herein.

€I l ll e paper used in lhis publical ion meets ttle requirements 01ANSVNISO Z39,48-1992 (Permanence cf Paper).

PR INTED IN THE NETHERLANDS

ELSEVIER Tcetonoph)'Mcs 280 (1997) 83-106

TECTONOPHYSICS

Temperature- and strain-rate-dependent microfabric evolution inmonomineralic mylonite :

evidence from in situ deformation of norcamphor

Marco Herwegh '", Mark R. Handy !>, Renee Heilbronncr "

• (N"I"g iM:hu hlS/itul. Ball:..n troß<' I. Unil't'rsität &rn. CH·JOI2 8ern. S" 'it:rrfand• Institut /Ur G..o ..it!<'nsdwjt..n. U"i.'t'rsitiit Gi..s....n. Gi..!.,«, " . (NfJII(lI/)"

~ Grologitdlu /" st.tul und A.bt..ilung/Ur lA'iSJ.." M:hajtlich.. Plwtogrophie. U"i .'t'rsitiit Bin ..l. Bos.-f. S..-it:;.-rlond

Abstract

Norc amphor (C7H,oO) was subjected 10 plane seain simple sbear in a see- tbroegh de formation rig at four differe nlstra ln rate and tem perature co ndiuons . Two tran sjent stages in the rmcrofabnc evohruon 10 steady stare are distinguisbed .Tbe gra m scale mechanisms assoc iated whh the microsuuctural and texrural evolunon vary with lhe applied temperarure.strainrate and strain, In high-temperarure-low-srrain-rate expenmems. cornputer imegrated polarizarion microscopy revealsthat the rexnr re evclution is closely retared 10 the crystatlographic rotanon path s and rota tion rares of individua l gra ms.High c-axis rotano n rates at low 10 intermed iate shear slmins are related to the developmeruof a symmetrical c-axis crossgirdle by the end of the lirst transienl stage (y = 1.5 10 2). During the sece nd trans ient stage (y = 1.5 to 6). the crossgirdle y ields tc an oblique c-axis single gird le as c-axi s roration mies decrease and the relative uctivity o f grain bo undarymig ration recrystallizanon increuses. Steady stete (y > 8) ts characterized by a stable end orientation nf the semple textuteand the cyclic growth, nuauon and consumprion of ind ividu al grains within rhe aggrega te.

Keywords: experimental suuctural geology: rock analogue: rnicrofabric: texture:dynamic recrystalllzanon

I. Introduction

Over the past thirty years. rnicrcstructural anal­ysis of experimentally and naturally defonned my­lonite bave shown that dislocation glide and creepare intimately associated with the development of acrystallographic preferred orientat ion (CPO), usuallytenned ' texture' in the materials science literature .Mylonitic texture yields Information about both the

" Correspond ing aulhor. Tet.: + 4 1 (3 1) 631·8764; fall: + 4 1 (3 1)631-4843; e-mail: herwtgh @g:oo.uniht.ch

physical conditions and kinemancs of deformation,001the difficulty in distinguisbing the effects of tem­persture. strain-rate and strain on texrure remainsone of the basic limitations in applying textures tointerpret the deformational history of mylonite . Itappears that the relative comribctions of these effects10 textute ca nnot be discerred without understanding

the way in which texture evolves with strain.Although most workers now agree that texture

fonnation in polycrystalline aggregares involves therotanon of intracrystalline slip systems. there is con­siderable controversy as 10 the prec lse way in which

0040- 1951197f$17.00 iO 1997 Elsevier Science B.V. All rig: hls reserva l.Pfl S00 4 0 -1951(9 7 )OOI 3 9 -X

84 M. Herl'>'rg h ef at. / Tt>cumophysics 280 ( / 997) 83-106

thcse systems rotate. One concept. the so-ca lled's table end orientarion' concept. envisages the ro­tation of slip systems into irrotational orientat ions.Intracrystalli ne slip occurs para llel 10 the overa llshear ing direction within glide planes oriented paral­lel to the shear zone boundary (e.g., Schmidt , 1927).This idea is suppcrted by complete texture analysesof neturally and expe rimenlally defonned mylonitesand assumes that the stable end orientaticn maxi­mizes thc resolved shcar stress on the active slipsystems (Schmid and Casey, 1986; Schmid Cl al.,1987; Etchccopar and vasseur, 1987; Law Cl al.,1990; Schmid, 1994). Dynamic recrystallization isbelieved to play a fundamental role in preservingthis stable end orientation by elirni nating grains thatare unfavourably oriented for intracrystalline glide(Schmid and Casey, 1986). A more recent concepr.the so-called ' ideal orie ntat ion' concept (Wenk andChristie , 1991) , rela tes texture to orientation-depe n­dent rotation rates of the active intracrystalline slipsystems. To paraphrase Wenk and Chri süe (199 1).idea l orientat ions are pole figure maxima corre­sponding to orientations for which the rotationsof slip systems are stowest (p. 1096 in Wenk andChristic, 1991). This rate-oependent rotation of shpsystems has been simulated in cornputer models us­ing both the Taylcr-Bishop-Hil l (TBH; Lister eral., 1978; Lister and Hobbs, 1980) and viscoplas­tic self-consistcr u(VPSC) theories of polycrystallineplasticity (Wenk er al., 1989). In the absence of dy­namic recrystallizaüo n, howeve r, no stable end ori­entations can ever be attained in such dynamic sys­terns. The modelli ng work of Jessell (I988a,b) andJessell and Lister (1990) shows that modifying TBHtheory to account for dynamic recrystalli zation canproduce start lingly realistic model quartz microfab­ries.

So far, neither of these conceprs has been ade­quately tested bccause rnicrofabrics of samples fromnatural shear zoncs and from high-pressure experi­ments record mostly finite strain, precludi ng a recon­structlon of the crystallographic roration trajectoriesof individua l grains during shearing. An alternativeapproach adopted below is to conduct experimentson organic analogue materia ls with a relatively lowmelting tempe rature and to monitor their microfabricas it evolve s under the optical microscope (Means.1977, 1989). Although the crystallograp hy of ma nyorganic rnaterial s is not weil known, previous workhas shown that the textures and microstructures pro­duced in such materials bear a striking resemblanceto those found in naturally deformcd rock (e.g..Means, 1989; Herwegh and Handy, 1996) .

In the following, we report on a series of expcr­iments conducted on norcamphor which we com­bined with computer-integrated polarization rnicros­copy (CIP; Panozzo-Heilbronner and Pauli, 1993)to document continuously the evolution of a sreadystete texture durin g simple shearing. The expe rt­ments were carri ed out ar four different homologoustemperature- strain rate conditions:

(I ) high homologous temperature-Iow strain rateexperiments (HT- LS);

(2) high homologous temperature-high strain rateexperiments (HT- HS; see also Herwegh and Handy,1996);

(3) intermediate homologous temperature- highstrain rate experiment s (lT-HS); and

(4) low homologous temperature- high stra in rateexperiments (LT- HS; see Table I for expe rimentalconditions).

The c-axis rotation paths and velocities derer­mined in norcamphor with the CIP method suggestthat texture evolution in polycrystalline aggregates

Table IE~perimen lal conditions: high tempe ranrre-low strain rare (HT-LS), high tcmpera ture-hlgh strain rate (HT- HS), intermediate tcmpera­lure-high strain rare (IT~HS ) and low tcmpcrature-higb strain rare (LT- HS)

HT- LS HT-HS IT- HS LT- HS

T ("C) zs zs 10 4T, 0.8 1 0.8 1 0.77 0.76Shear strain rate 4.0 x lQ_3 5.5 X lQ_4 5.5 X 10- 4 5.5 X 10- 4

Maximum allained shear struin (y) 7.9 10.5 9 sExperiment duranon (h) " 5.25 4.5 2.5

ss

involves elements of both the stable end orienta­tion and ideal orientation co ncepts ouüined above.We conclude the paper with a list of criteria thatshould enable geologists tc disunguish the effects ofsrrain and tempcrature in quartz mylonite that hasundcrgonc simple sbcaring.

20Sampie preparation and analyt lcal procedu res

We used a Means-Urui rig to deform norca m­phor (C7H IOO) in plane strain. simple shear (secJessell . 1986; Means. 1989; fig. la in Herweghand Handy. 1996 for a dctailed desc ription of theexperimental co nfigurat ion). Herwegh and Handy( 1996) have found many str iking similarities be­tween the microfabric of experimenta lly deformednorcamphor and that of naturall y deformed. my­Ionnie quartz . Norcamphor is uniaxial negati ve andprobabl y has hexagona l crysrallographic symmetryas inferred from hexagonal dendrites formed duringsublimation on glass (see fig. 2.5 in Bons. 1993). Un­Iortunarely, the high sublimation rate of norcamphorat room temperature precluoes the use of standerdx -rey gon iometry to identify its crystallography andpotential glide systems.

We prepared the sampie in the same way asalready described in Hcrwegh and Handy (1996):Norca mphor was mixed with corundum grindingpowder prior to cold and hot pressing. In cen trast toour previous sampie preparation procedure, however,hot pressing of the eurrent sampIe batch lasted longer(96- 172 h) at temperatures ranging from 35" 10 45"C.Fig. I dcpicts pole figurcs for the Initial. hor-pressedtextures of the sampies prior to the simple shear­ing experiments. Note that coaxi al flatten ing duri nghot pressing induces c-axis point maxima parallel tothe Y direction in Fig. Ib .c (see HT- LS. IT-HS andLT- HS columns in Fig. Ib .c ). lbe initial texture ofthe sa mpies in the HT- HS experiments compri!\Csconcentric c-a xis small circ le patterns about the prin­c iple stress (UI) direction for hol pressing (HT- HScolumn in Fig. Ib .c ). In contrast. the initialtexturesof the other experiments eo ntain c-axes maxi maalong the periphery of the pole figures in Fig . 1b (seealsoelongate maxima in Fig. l e ). These textures mayretlect sampie f1 0wage para llel to the fros ted grips ofthe glass slides during f1auening in the Y direet ion inFig. l a. Note that the incomplctc small eircle alo ng

thc periphery of the pole figure in the initial HT-LStexrure (Fig . I band Fig . 2) reffccrs an artefaet of themeasuring method. Becausc the HT- LS run was thefirst experiment in which we applied the CIP method(see below), we did not rcalize rbar the ti lting anglewe applied during digita l imag ing was too low. Con­sequently, norcamphor c-axes wi th az imuths in thc220-29<f' range at interrne diatc to high incli nauonswere not imaged co rrecrly. Based on later experi­rncm s. we surmisc that co rrec t measurement of thetcxture would have yiclded a small cirele c-axis pat­tern symmetr ically disposcd ubout the hor pressing"I direction.

We applied two metbods of rextute anal ysis in theana logue expe rime nts:

( I) a mod ified oprica l 'Achsenverteil ungsanalyse'or AVA (Sander, 1950) involving V-stage measure­ment of oorcamphor c-axes whose gene ral orienta­tion within thc microstrucrure was estahlished fromthe interference co lour of grains (for a detailed de­scr ipnon of the procedure. sce Herwegh and Handy.1996); and

(2) computer-integrated polarizanon microscopy(CIP; Panozzo- Hcilbronner and Pauli. 1993. 1994).

Using the V-stage has a basic problem: rapidannealing of the sa mpIes prcctudes defo nni ng thesampie to higher strains after r- axis measurement.Thu s. thc experime nt must bc repcated several times.euch time to a different shear srrain. such that thec-axcs are measured in a different sampie at thc endof eac h ron. A quantitati ve cvaluation of the rotationpaths of c-axes for individual grains can thereforenot be ob tained with the V-stage. The main advan­tage of the CIP method (method 2) is that it allow ssuch an eva luation and we have app lied it to trackc-axis trajectories in the new experiments reponedbelow. Tbc CIP so ftware p<!Ckage generales a c-axisorientation image from 22 digital infrared imagesby calculating grey value images reprcsenti ng theazimuths and inclinations of the c-axes at eac h ofthe pixels of the emire grain agg regate (Panozzo­Heilbronner and Pauli. 1993. 1994). Eighteen of thetwenty-two images are taken at 1(1' rotat ion intervalsand four images in tilted orientations (see configura­tion in Herwegh. 1996). The c-axis orie ntations areassigned characteristic colours (see colours and the irrefe rence code in Figs. 2-4 ). The c-axis orientat ionsare then ploued in a standard c-axis pole figure using

M. Hl'nu Xh l'1 a/. / T«''''''>php in l HO ( / 99 7} 83- 106

HT-LS HT-HS IT-HS LT-HS

b) ~ ~ ~ ~y y y

cr,x x x z x

preparation •

-a- -a- -a- -a-c) z

~z z z

~cr, x x ~ x

experiment

~ ~ ~T = 35°C T = 35°C T = 450Ct = 1nhrs t = 168hrs t = 168hrs

a) weight Z

1 ±; IUIU"~' tqZ ~ shear ' "y~zone ~ J"X-.. t,..,__

OB y

preparalion experimentFig . I. Equal-area projccuons uf initial textures aftcr hot pressing during sampie preparat ion. (a l Plaucmng plane is oriented east-wcstin the pule figures. perpendicular tu the (1"1 direction of hot pressing (hl al'k arrow sj, (h ) Simple shear plane is orie ntcd east-west in thepo le figurcs. corre sponding to the SZB at 450 to the (1"1 direetion dur ing the cxpcrimc nts twhitc armws), Note Ihat the initial textures inthe fiNt and second mws arc idcntical: a 90" c joc kwise rotarinn aho:ll.I t thc X sampie preparation axis trun-Jcr ms the pole figures in (a ) Inthose ,mN," in (b ). (c l Spccimcn axes d uring sample prepareeion and simple shearing expcrimcnts. T = hör pres.sing tempe raturc. 1 =hOl prcs.sing duralioo.

the program 'Stereoplor XL' {Mancktelow. 1993).Table 2 shows the methods employed and the fre­quency of texrural unalysis pcrformed during eachof the experirrems. In the HT- LS experiments. thestrain rate was sufficicntly low (4 x 10- 5 S-I ) toobtain seventeen orientation images without inter­rupting the experiments. In ce ntrast. all high strainrate experiments had to be interrupted for 10 ro 15min in order to record rhe 22 digital Images requiredfor each orientation Image. To minimize sample an­ncaling during thesc intcrrnissions. the number oforientation images per increment of shear strain wasrcduced (see IT- HS and LT-HS columns in Tablc 2).

The resuhing dccrease in temporal resolution wascompensared for by taking additional digital imagesevery 5 min without intcrrupting rbe experiment. Thepublic domain software NIH image 1.57 tRasband.1995) allowcd us to anirnatc thcse images and tocalculate changes in the areas of individual grainsand grain aggregares with strain Ie.g.• Fig. 5).

3. Micro fahric evolutlon

Our previous work with norcamphor has shownthat microfabric changes involve the slmutrancousactivity of several interacnvc mechanisms from the

,\1. f"'I",<'.~h .'1 <11./ 'fi.'clO1I" I ,hn in 28(J (IW71<'lJ - /110

Fi~. 2. IlT-LS microfahric cvolunon . Finile vtrain cllip-c-, fliN coluuuu CII'·~cnl·ralcll oricnWlion illl,,~e' ("",,,,,,li eolullln ) .1I111 colourrctcr cncc pole ßgurc (hol1"m "I' secend column}, r omou rcd ('·a, i, pole liguln ühird l'"I Ul1l l1 l. Sh"ar /"11" hl'unll:lr~ Idushcd horivumalline' ) und long '1, i- "I' thc finite strxin cllipw Iline l..bcllcd S(/ ). Comour imcrvaI, of thc pote tigur",:ore 0.25. 0.5. 1.0 (ll..shcd ). 2. -I. K.]6. 32 time, uniform distributinn.

I;i~ . J. IT~US micn.fahri<: evohmon. l-irnrc crain cIliJ"e" I fiNl... lumn], ''f'lical phulo~raph~of mk:n>fahric 1...."Condl..>lumnJ and l""t.IOIlrretcrcnce pole figure tbonom of ...."Condlrolumm. coruourcd ,··a,i, polefigurL"' ubird columnj. Sole Ihal. lnc retcrcnce culuur pole figurcr= CUIll,...,;t>p ic image I CIllTc~I'l'lllh ,,, a ...,;1 up whcre the ana l)'ler. 1'l,1;lriJcr arKi cumpcn-alur arc Illlall'd -I5~ cuumerd.>cl.."'i....: (rum

rhcir ~t;mrJarrJ povinon. Crach lhlach arr""'~l. shcar zone htounrJ;lry "Li~ht.'d huri/llfllal linc'l and In"1! ;ni, ur IOC finite -train etlip-crlille lnbellcd SIl) are indicncd in lhe I'l,le figurcv, Conmur inter';II~ ur lhe pole figure-, are 0.25. 0.5. 1.0 1.L1,hl:rJJ. 2. -I. X. 16. .12 tin...·'unifurm di-aribution.

M. HenlY:gh et alr Tectnnoptiysics 280 (1997) 83-106 89

sub- to the supragranular scale (Herwegh and Handy,1996). We applied the conce pt of a mechanism as­semblage to describe the simuttaneous activity ofsuch mechanisms, with the term ' mechanism' usedin a rathe r broad sense to refer 10any grain scale pro­cess that involves changes in microfabric . Dur newwork below indicates that texrural and microstruc­tural changes leuding to a steady state microfabricare strongly dependent on ternperature and strainrate. The recognition of different relat ive activi tiesof rnechanisms with strain allowed us to discerntwo rransier ustages in the microfabric evolution. Al­though the nature end duration of these stages variedwith temperat ure and strain rate, both stages werediscerned in all of the experiments described below.

3. 1. High temperature- low strain rate (HT-LS)experiments

3.1.1. First transient stageAt y = 0 to 1.5, glide-induced vorticity ('shear­

induced component of vorticity' of Liste r, 1982;Lister and Will iams, 1983) is manifest by theglide-i nduced rotanon and elongation of individ ualgrains. This results in an oblique shape preferredorientation (SPO) at 60" to the SZB. Strain-depen ­der u changes in the CIP-generated colours of grainsreffect changes in grains' c-axis orientations (Fig. 2).These oriemauo ns can be read from the colour-codedpole figure at the bottom of Fig. 2. The glide-inducedrotanon of the c-axes as weil as the simultane­ous growth of yellow and magenta grains via grainboundary migration leads to the rapid consu mptionof grains that are unfavourably orien ted for intracrys­talline glidc parallel ro the SZB (e.g., appearance offavourably orienred yellow grains and disappearanceof unfavourably oriented blue and purple grains inFigs. 2 and 5a) . The avetage grain slze increases(Figs. 6 and 7) and a weak domainal microfabricdevelops at the end of the first transient stage. Atsomewhat higher shear strains (y = 1.5), the textureconststs of a c-axis cross girdle that is symmetrica llydisposed with respect to the long axis of the finitestrain ellipse (sec third column in Fig. 2). The con­centratton of c-axes in the centre of the pole figure ispartly inherited from the initial texture (Fig. Ib.c) butalso reflects incipient roration of some c-axes into anorientation consistent with prism glide parallel to

the SZB (see discussion section below). The c-axespoint maxima of the cross gird le at the peripheryof the pole figure in Fig. 2 were generated with themechanism assemblage described above.

3.1.2. Secend transient stageBetween y = 1.5 and y = 6, intense grain bound­

ary migration recrystalli zation combined with the co­alescence of yellow grai ns (Mean s and Dong, 1982;see fig. 9 in Herwegh and Handy, 1996) strength­ens the domainal charac ter of the microfabric untilthe microfabric only consists of yellow and magentagrains (Fig. 2). Thc average grain area increaseslinearly with strain, reflecting the high activity ofgra in bounda ry migration (Figs. 6 and 7). The samephenomenon was observed in one of the octachloro­propane expcrtmenrs of Jessell (1986 , fig. 7, ron TD­63). Progressive subgram rotation recrystallization('rotational recrystatllzanon' of Poirier and Gui1­lope, 1979) is especiall y active in yellow grains andcontributes to the complere dynamic recrystalliza­tion of the enrire rnicrofabric at shear strains above2.8. Interestingly, the average grai n SPO already be­comes strain invariant at shear strains of 1.5. This isconsistcnt with observa tions of SPO in octachloro­propane (Ree. 1991) and norcamphor deformed athigher srraln rates (Herwegh and Handy, 1996). Theprogressive strengthening of the domainal microfab­ric is close ly related 10 thc consumption of magentagra ins by yellow grains (see Fig. 5a) and involvesa texrural transition from a symmetrical c-axis crossgirdle to a c-axls single girdle. The lauer is orientedsubperpendicular 10 the SZB but obliquely orientedwith respect 10 the main foliat ion (see second andthird columns of Fig. 2).

3.1.3. Steudv state

At shear strains above 6, borh the texture andthe relative area of yellow and magenta grains be­come strain invariant, indicating the artainment oftexrural and microstruct ural steady state (Fig. 2).Although microfabric appears 10 be strain invari­ant on the sampie scale. the continuous activityof high- and low-angle grain boundary migrationand subgrain rotanon recrysraülzatlon replenisbesthe rnicrofabdc with a steady supply of new, pre­sumably less-strained grains (see below). The detailsof this renewal process are quite complex: when

co

Fig.4

Fig.5

M. Herwegh 1'1 ul.lTeclOllOphysics 2XO (/997/ X3-106 91

Tahle 2Chart showing rhe methods uscd rc doc vmen t the microstruceural and rcxiural evclurion during thc expcrirncnt s: digita l infrared pictures,normal 35 mm photographs. ClP (computer-integraIL"d polarizadon microscopy). V-SI. (V-stage mcasurement of c-axes)

yo

2

3

4

5

6

7

8

9

10

11

HT-LS HT-HS IT-HS LT-HS

CIP .,U-StJ rCIP ICIPCIP ~U-St J iCIP • rCIP ICIP

~U·St J0

o-@] • ICIP •CIP •CIP --jU-StJ ~ C[P ICIP • !CIP !CIP •

•~ --jU-StJ c CIP

CIP •CIP •CIP •CIP

.,U·StJ ~•• o--[QjpJ•-1U-StJ•

o digital pietute • photograph

The evolution of the microstructure in all cxpcrimcnts was also documcntcd with videotapc (HT- HS) or computer animation of thedigital images (HT~LS. IT- HS. LT-HS).

obscrving the grain boundaries of yellow and ma­genra grains, we were surprised 10 find that oldmagenra grains are never completely consumed bythc ycllow gralns. Instead. they undergo cycles ofshrinkagc and growth. Burg et al. (1986) observedsimilar oscillatory growth and consumption of grai nsin polyerystalline iee undergoing simple shear. Theoscillatory bchaviour of magenra and yellow grainspreserves a stable proportlon of diffcrcntl y eo louredand criented grains within thc samplc (Fig. 5a). A

statistically constant proportton of di fferently ori­cnrcd grains on the samplc scale is also manifest inthe texturc which compnscs two stable c-axis pointmaxima , one paralle l to thc Y spcctmcn axis and theother slightly oblique to the SZB normal (see bonomtwo rows of Hg. 2). Note thut Figs. 6 and 7 scem 10

indicate that grain slze lncreases continuousty withstrain and therefore has not rcachcd stcady state.Uufortunately, thc weak colou r contrasts bctwcensimilarly orientcd grains at high shear strains pre-

Fig.4. LT-HS micro fabnc evolution. Finite srrain ellipses (fir't column ) C1P-generatcd r- axis orien talion images (sceond column) andcolo ur refcrcnce pole ligure (top left-hand corner Hf each orientatinn imagc). comourcd r- axis pole figures (rbird column). Shear zoneboundary (dashed horizon tal hnes) and long axis of the finite strain ellipse (linc Iabellcd Su). Contour intcrvals of thc pole figures are0.25,0.5. 1.0 (dashcd) , 2. 4. 8. 16, 32 nmes uniform distnbution.

Fig . 5. Relative urcal proponions of grains with a specific crystallngraphic one ntation as a function of increasing shcar strain for HT- LS(a), HT- HS (b). IT-HS (c) and LT- IIS (d) expcnrnents. Corresponding Cü'-gencreted colour refercnce pole figurcs are shown.

92 M. H~"'...~gh ~f al.lT«/OIIoph.nin 180 (J997J81-/06

HT-LS I HT-HS I IT-HS I LT-HSI I II I ~.,

.,~ 'T~ "u1 r =O I 0.4 Y= 0 I

~:u0.4 't> 0

cz c .a . 3

f

I0.1 ; cz

I ".. ,., .., , I0 0 1.0 0 0 1.0, , z a • s • r o 1.3

ua

"u 'I~ "uIr - 2.8 0-4 "(= 2 o.s Y= 1.7 0.4 = 1.

'" na .. .,"

nz ,. cz. d .. cz ..

0 0 0 1.3 o 0 1.0 o 0 1.00 , , 3 • s • ,(/)

.: 'T~c

0.3

Ir =4.0 "u 00 .4 "(= 4 0.8 "( = 3.4 0 .4 y= 4

'2 oa O• "t::,., ,.. ,., 0

c.t

1c., ... . , .t ,., O. 0

0 0 1.3 0 0 1.0 0 0 1.0 ...0 , e , • , s r c.

o.a

"LJ 'T~Cll

Ir - 5.5 0.4 y= 6 (18 1= 5.1 IQ)...

0 '

",. Cll

,., cz " Q)

d., I,., " .>

c 0 0 - ' 1.3 o 0 1.0 -c , a , • s s , Cll

'I~Q)

c.a

"u...

Ir =7.0 0.'1 Y= 8 Ic.e r =6.85

'2 0.'0.' I

I0' 0.', .t

Io ..I 0.' 0.'

0 0 1.3 I o 0 1.00 , z , • s s r I03 I

'T~"uIr- 7.9 0.4 Y= 10 I 0.8 1=7.7ca c.a I c.e..

Io.z I c.•.. I cz

0 0 0 1.3 I 0 0 1.00 • z a • s • ,grain area (mrn-)

Fig. 6. E\I'Olulion of , rain area distribulron .. i m shear strain in the experiments. 1he pain area evolution depenoß on temperemre andslnIin rate (~ text for uplanillion). NOle mal ee scaling on ee horizont.ll grain areaaxis and the verucal area proportion iUis vary inthediffaenl columns .

M. H""""t'f;:h l'(al.l T.,('/onoph.n;o 21:10 ( / 997) 83-J06 93

Fig. 7. Change in grain area with shcar strai n for thc differentexperimental conditions (sec tcxr for explanat ion ).

c1ude the accurate mea sureme m of grain shape andsize, so that the estimates for the HT- LS runs atY 2::. 6 in Figs. 6 and 7 may not acc urately reflect rhctrue grain size.

8

The HT- LS microstructures describcd hcre arcvery similar to the HT-HS microsrructures rcportedin Herwegh and Handy (1996). ind icat ing thut strainrate has a relat ively modest effect on microfabricevolution at high homologous tcmpcra ture (compareFig. 2 of this paper with fig . 2 of Herwegh andHandy, 1996). Nevert hclcss. there are some differ­ences between the HT-HS and HT-LS microfabricsthat ca n be related to varied strain rate. During thefirst transie nt srage in the evolution of the HT-HSmicrofabrics. subgram rotation recrysta llization andrigid body rotanon are the predominant mechanisms(Fig. Ba). wherca s grain boundary migration is lessactive than in the HT- LS experimcnrs (Fig . 8a) . Thisis associatcd with a significam decrease in grainsize (Figs. 6 and 7) as weil as with the develo p­me nt of a c-ax is cross gird le at relati vely lugher

a LT-HS

10 12, 6

shear stram (y)

o HT·HS t. IT-HS

a

° HT-LS

@

10.3 10-4

log stram rate (s")

subgrain,otation<:\om;nal&d regime~

~/&J.:~IT'Hcryslal plaslicity ' ,~ ~I.i

, • SblittH! lailuft!__ LT·H

1-a

10

100

t:•~ 10•0E•l'

102 4. 6 8snear snafn <ll

o

IHT-LS I _ : -"'- . -v~o;:.:~_~~,!­~ ,

~ '0-'-,J.,. ---'__-€~_ __I~ I~ .--------- _~o ! : ~ .

o 2 4 6 8 10

'Iiiiiltii.......... : MCOI'ICl b1l_ Olea""

sbear strain hl

• subgrain rotanon dominated regime b c• crystal plasticity & brittle fallure

! LT- HS I 1.1"'...."", : ,~ghde ncoceo vorticity

~

~, 11 rigidbody rotaneo,-0- _ :

~, m subgrain rotation":" - - ---

•> >V scouereccs nocieeuce•~OV grain boundary migration0 3 s

shear strain (r) VI fraeturil1 g

10

­,I 'T- HS I

ri '~::!.:::.:;:.: ::=~:::.::.:::: ~.::.::::=.eO _,_ _.. _- -

o 2 4. 6 8

Fig . 8. Sehemaue diagrams showing the variat iun in mechanism as-.cmhlage as a functi on of shear strain fur the different experiments: (a )HT- LS and HT- HS. tb} IT-HS. Ic ] LT- HS. «n Temperalure vs . strain rate plot o f the dominant grain sealc mechani'ms in expe rimentaldeformation of norcamphor (see text fOT explananon). Nute the LT-LSR (10'" tempcraturc- Jow strain rate) experimcnts arc nut reportedin this work hutihey yield thc same microfabrics as the HT- HS experiments.

M. H~",,'~h ~I ol. I Ti'CtflflOph.niC'S 2J«J f 1997) 83-106

shear stralns (y = 2) in the HT-HS expenmems.In the second translern stage . the width of the do­mains in the HT-HS microfabric is narro....'er rbanthat of the HT- LS microfabncs (370 ILm vs. 570ILm). At steady srate. both cxperimem typcs havea similar mcch anism asscmblagc (Fig. 8a). In caseof the HT-HS expcrimcm s. however, grain boundarymobilit y is lower and spontanco us nucleation of newgrains is an add itional meehan ism associated withthe co nservarion of a strain invariant mic rofabric(Fig . 8a) . Spo nraneo us nucleation is discussed belowin greater de tail.

As already demonslrated in Herwegh and Handy( 1996). the high-temperature norcamphor microfab­ries are very similar ro thosc of natura lly quanz my­loni te deformed under uppe r greensc hisr to amphi­bo lite facie s eonditions (compare fig. 12 of Herweghand Handy. 1996 with Fig. 2 of this work). In boIhnorcampho r and quart z samples . thc microsuucrurecomprises an ob lique SPQ and retauvely Jarge grainswith bulged grain boundari es. These Ieatures reßectthe streng eo mponent of grain boundary migration insimple shear. In norcamphor. we ooserved that smaUequi-axed gra ins originate by progressive subgramrota rion. We infer a similar origi n for sueh grainsin quartz. Neven heless. d ifferenees in the positionsof c-axis point maxima bcrween quartz and nor­eamphor text ures probably rcflect d iffere nces in therelat ive act ivity of slip syste ms (eo mpare fig. 12 ofHe rwegh and Handy, 1996 wilh Fig. 20f this work).

3.2. Intemlediute remperarure- high slrain rale(rr- HS) experiments

3.2. / . First transient stageThe mechanism assem blage fo r the first transient

stage ( y < 2) at IT-HS conditions is a co mbinat ionof subgrain rotation recrystaJlil..ation. glide-inducedvorticity. grain boundary migra lion . with subord inateacti vity of spontaneous nuclea lion a nd rigid body r0­

tat ion (Fig. 8b). The rota tion of inlracrystalline glideplanes towards an orie nlation favou ring slip parallelto the SZB (Fig. 3) is associated primarily with pro­gressive subgrain rotali on recrystallil...ation. leading10 a strong decrease in the ave rage grain size (Figs . 6and 7). Rigid body rotation of grains is unimportantcompared 10 the HT- HS ex periments. The gra in SPOfarms a similar angle wilh the SZB (60" ) as Ihose

observed in the high-temperature expcrimenrs andthis angle becomes srrai n invaria nt already duri ngthc first transient stage. Herw egh and Ha ndy (1996)related this a ngle to the average 60" ang le bet weenthc SZB and do minant microshear zones within thesa mpie (see also Hcrwegh and Handy, 1997). A weakdomainal microfab ric conslsung of yellow a nd ma­ge nra grains develops at the end of this stage (Figs. 3and 5c) . As in the high-tempcrature ex perimerus. asymmctrical c-ax ls cross girdlc forms pcrpendicularto the long axis of the finite strain e llipse calc ulatedfor thc e mire sample (Fig. 3).

3.2.2. Second transient stageWith progressive deformation. a co rnbination of

subgram rotaden. glidc-induced vonicny and grainbou nda ry migration srrengrben rbe domainal mi­crofahric (Pig. 3). In panicular. we noti ced thatnoncoa xial sheari ng of grain boundary bulges ro­tate s these bc lges into paralle lism with the grainboundari es. Renewed bulging along these shearedboundari es contrtbnes to the growth of domain s (seefig. 10 in Herwegh and Handy. 1996). The averegewidth of these dorneins (170 ILm) is Iess than thedomain width in the bigh-rernperarure experiments.The average grain size still dccreases during thesecond tran sien l stage duc to the continued activ­ity of subgrain rotanon recrystallization and spor ua­neous nucleation (Figs. 6 and 7). We apply the term'spontaneous nuclea tion' to the sudden appearanceof minute grains in animaled d igilal images of thcevolving microstruclure . These grains grow rapidlyat the ex pense of other grains until auaining dirnen·sions comparable to those of 0plk al subgrains. It isunclear if the ir nucleat ion invol ves the progressiverolation of TEM-size subgrains. the rotalion of '>mallgrain fragments. or even c1assical nucleat ion (Druryand Ura i. 1990). We favour the first possibiJity. how­ever. because spontaneous nucleat ion tends to occu rtogether with subgra in rotati on recry stallization nearthc bou ndaries of larger host grains (e.g .. Fig. 8a.b).The microfabri c is completely recrystallized at shearstrains greater than ~ . During the second tran sicntstage . the c -ax is cross girdle rapidly reduces to anoblique single girdle (last Iwo columns of Fig. 3).Interestingly, several small crads opened perpe ndic­ular to the incremental stretching axis. These crackslengthened as they rotatcd synthetically inlo paral -

9'lelism with the main axis of rbc finite strain ellipsetcompare with LT-US experirrents below).

3.2.3. Steodv stateAI Y > 6, grain size, domain width. gram SPO,

and textute a ll become invariant with srrai n on thesampie scale (Figs. 3 and 5c). On the grain scale .howcver, subgrain rotano n and the cyc lical growrhand co nsumptio n of individual grains or subgrainsconstantly ' refres hes' thc microfubr ic (sec Fig. Sb).The elongate fractures that formed end rotared dur­ing the second translern stage remain open parallel 10the ma in axis of finite strain elli pse. Com pared to thehigh-rernperature microsrucrures. the IT-HS steadystete microstructu re shows a much smaller avcra gegrain size (Figs. 6 and 7), a smaller domain width(Figs. 2 and 3), a sreeper SPO. a differen t rrecha­nism assembtage and other relative mechanism acuv­ities (Fig. 8). These pronounced differences berweenhigh- and intermediate-temperature microsrructuresare also man ifest in the sready-stare rextures: pointc-axis maxima typical of high-temperature deforma­tion are replaced by oblique r-axis sing fe girdleswith a stable obliquiry with respcct to the SZB atshear stra ins grcater than 6 (right-hand column inFig.3).

Natural cquivalents to thc IT- HS norcam phor mi­crostructures are rather common in greenschisr fadesquartz myloni tes (c.g.. Knipc and Law, 1987; Lawct al., 1990) and havc also been gcncrated in bothcoaxial shear (regime 3 rnicrostructures of Hirthand Tullis, 1992) and spli! cylinder shear e:\pcri­ments on quartz ite (Dell'Angelo and TuJlis, 1989 ).In the e:\periments by Hin h and Tullis (1992), quanzalso begins to recrysla llize along the boundariesof hosl grai ns by a combination of subgrain rOla­tion and grain boundary migration recry stallization .AI strains corrcsponding 10 57'k axial shonening.complelc recrysta llizalion of Ihe sampie eradicaleda11 vestigcs of thc core-manlle structure in quanz(see fig. 6d in Uin h aod Tullis. 1992). Spontaneousnudeation and veining such as obse nted in the nOT­camphor e:\perimenls were not reponed from any ofIhc quanz e:\periments in the literalure. ldentical te:\­lures 10 the IT-HS norcamphor te:\lures documenledin Fig. 3 can be found in natural quartz aggregalesIhat were subjec ted to plane slrain simple shearingundcr gree nschisI lo lower amphibolile facies condi-

rions t Schmid and Casey, 1986 ; Law er al., 1990). Inquanz, this type of rexrure pauem is interpreted roreßect slip on the prism. basal and rhomb planes inthc (a) direct ion.

3.3. Low temperature-high strain rate (LT-HS)experiments

3.3./. First transient stageAn oblique ribbon grain SPO develops already

during the first incrcmcnts of shear strain. Th isrc flccts a significam componenl of glide-inducedvorticity (Fig. 4). Com pared 10 the previous exper­imcms. dynam ic recrysrallization is strongly sup­pressed and involves Iimited grain boundary mi­grau en (Figs. 4 and 8). The texture comprises asymmetrieal c-axis cross girdle {Fig. 4).

Ouring this low-temperarure crystal plasuciry,Iractures develop along gra in boundaries with anInitial orientation of 1350 to thc SZB. i.e . parallel10 the inferred GI direction of simple sbear (Fig. 9 ).With conunued strain. rbe following (wo rypes ofcrack evolution can be discemed :

( I) Cracks open parallel to the incre menlalstretching direct ion and propaga te para llel 10 thc0 1 direction. They then imerconnect 10 form a bigvein thaI truncares the entire shear zone [white cracklabclled 2 in Fig. 9). Pinning of such veins at thcSZB prevents their rotation during the simple shear­ing cxperiment.

(2) In cases where cracks do not interconnect,they remain relatively small and rotale syntheticallyinlo concordance wilh the SZB. The cracks open aslong as their long axes lie wilhin Ihe shortening field.001dose obliquely upan rolat ion inlo Ihe e:\tensionalneid. OllCe the cracks are d osed. only a few OObbletrails subparalle l lo Ihe long axis of Ihe finile strainellipse mark Ihe fonncr local ion of the cracks (seebubbles labelIed I in Fig. 9).

3.3 .1. Seeond lranS;enl Slagt'Dynamic recry stalli zation involving a combina­

lion of subgrain rotalion recrystallizalion and grainbountJary migrat ion Iypifies Ihis slage. This is asso­ciated with a reduct ion in Ihe average gra in size andIhc dcvelop mcnt of a core- mantle slructure (Figs. 4,6 and 7). The large veins continue to dilate. 001 noncw fraclures nucleate in Ihc norcamphor belween

,',I. H l' n n:x!l 1'1 ,,/. /Tl'u"""p!l,n;n 2lio (1997) II3- f ()fJ

Fig. 'J. Crack evolution in the samplc dcfonucd at LT- HS condi­nons. Type I crac ks nucleate at grain boundmes onc ntcd 135"ro the SZB, Tbcy then opcn and rotate symhcticully towards theSZB. These cracks closc upon rotanon into the short cning Iicldof the inc remental strain ellipse (0-90" rc SZB j, Note bubblcsoutlining thc tracc of the closcd crac k. Type 2 crucks nuclcatcas abovc. propagate pcrpendlculur 10 the incrementa l streichingdirec tion and imcrconnect to form a big vein.

....... extension

"/ compression

• crack propagation

__shear movements- along fractures

thc Iract ures. A c-axis texture dcvelopv in uncrackcdregions, suggcsring thut intracrystalline glide and dy­namic recrystallization are able to mai ntain strai nco mpatibi lity in the parts of the aggregate betwee nthe cracks. Thus. the loca lized cracking does not up­pear to bave a strong influence on texture evolution,at least at low strai ns.

It is important to note that close spatia l and tem­poral rela tionship of mtcrosuucrures related to frac­turing and intracrystalli nc ptasucity is a diagnosticcritcrion for thc brinlc to crystat plastic transition innatural fault rocks (i.e. ' frictionul to viscous transi­tion' of Schrnid and Handy, 1991). Altho ugh syn my­lonitic cracks havc bee n recognizcd in greenschistfac ies quartz mylonites (e.g., plates 4.20 und 4.22 inHandy, 1986), the big cracks formcd in the LT- IISnorcamphur expc rime nt are probably not vcry real­istic . The unrealistically low co nfining press ure inou r experiment favours the opening of such cracks.Mc reover, because these cracks are pinned at theSZB, they cannut all rotate into an orie nta tion whichwould lead to their clos ing. Fina lly, the lack of afluid phase in the experiments inhibits sea ling of thecracks via solution-precipitation mcc banism s. Un­fort unatcly, the inability of thc hig cracks 10 rotateand scal lead to disintcgration orthe sample. prevent­ing us from reuehing shcar strains of grea ter than 5in the LT- f1 S cxpc rimenrs. Therefore. sready statcwas ncverauatncd.

To summarize this secnon. our experimcnts in­dicatc that microfabric cvolunon in norcamphoris high ly tcmpcrature- and straln-ratc- dcpc ndcru.Fig. 3d shows the relationship bciwecn the tcmpcr­atu re-strain rate conditions and the dominant grainscatc mechanism inferred from the microfabrics inour expcrimcms. Shcaring of norca mphor at high bo­mologous temperatures favours grai n bou ndary mi­gration recrystallization as the do minan t grain scalemechanism accommodating glidc-induccd vorticitywithin the aggregate. Increa sing strain rate at rhishigh tempcnnure tendv to tncreasc the relative activ­ity of subg ram rotati on rccrys lallization and, at lowstrai ns. also to fuvour rigid body rotation of 'hardgrains' {i.e . gra ins poorly orie nted for sl ip parallel tothe SZB). By co mpariso n. microfabrics ge nerated insimple shear at an intermediale homologous tcrnper­ature are much fi ncr grained. ind icating the prcdo m­inance of subg ram rota tion recryslall izat ion. In ex-

M. Henvegh 1'1 ul. / Tectonopbvsics 280 ( / 997) 83-106 97

peri ments conducted at the lowest homologous tem­peratures, the inabi lity of crystal plastic mechanismssuch as intracrystalline glide and dynamic rec rystal­lization to accommodate the bulk strain compatiblyleads to the opening of tensional cracks along grainboundaries. The aggregate defonns at the brit tle toductile transition. In alI expe riments, however, themicrostructural changes are strongly linked to a tex­tura l evolution characterized by the transltion from asymmetrical c-axis cross gird le to an oblique c-axissingle gird le with respect to the long axis of thefinite stra in e llipse. In the next section. we examinethe way in which the rota tiona l history of individualgrains rela tes to the formation and preservanon of asteady stete mierofabric.

4. Crystallographtc rota non patbs and rates

The texture of a polycrystalli ne aggregate com ­prises grains with different crystallographic orien­tati ons and rotanon histories. Up to now, such ro­tattonal histories could only be approximated bymeasuring the cryst allographic axes on the U-stageonce the expe riments were sropped (Jessell, 1986;Herwegh and Handy. 1996) or by inserting a Berekeo mpensator during experi mental runs to interme­diate shear strains (y = 104, in Ree, 1991). Wetherefore app lied rhe CIP method (deseribed above)during the HT-LS runs to reco rd a sufficiently denseorientation image distribution per shear strain inere­ment to calculare rhe rotation paths of individualgrains.

Basically three different types of rotational pathea n be observed in the pole figures in Fig . lOa,b:type I, shallowly inclined c-axes at the periphery thatrotate synthetically with respect to the bulk senseof vortici ty: type 11, steeply indined c-axes that re­main in the centre of the pole figures; and type111, intermediate to steep c-axes in the centre of thepole figures that rotat e towards smaller inclinations.These general c-axis paths are shown schematicallyin Fig. l 1a. The relative rares of the orientationa lchanges eorresponding to these three types of rota­tiona l paths are best seen in those parts of Fig . 10that dep ict changes in the azimuth (Fig. IOc,d) andinclination (Fig. l Oe.f) of norcamphor c-axes as afunetion of shear strain. The steeper the curve. thefaster thc rate of erystallographic rotation. Therefore ,

grains with paths types 1 and 11 eorre spond to the flatcurves in Fig. lOe,f, indicating a low rate of inc lina­tional change, whereas gralns with path type 111 havesteep curves diagnostic of high rares of inclinationalchange. Intere stingly, there is no general correlationbetwe en grains ' ortenrational path type and changesin azimuth in Fig. IOe,d. Some grains maintain theirazimuth (flat curves ) whereas others show initia llyhigh rares of azimuth change that dec rease withstrain (co ncave curves in Fig. lOe,d).

Thi s rather complex relationship betwee n crysta l­lographic orien tatio n and rota tion rate is summarizedin Fig. I lb, which outlines fie lds of relative c-axisorientational stability (= low c-axis rotat ion rates).The boundaries of the stability fields in this dia­gram derive directly from the angles corre spondingto horizontal dashed lines that separate differentl ysloped curves in Fig. lOe-f. c-axes with orie ntationsin the white fie ld of Fig. Ilb rota te quickly towardorientetions with slower rotation rates (hatched field sin Fig. IIb) before being eliminated by dynarmerecry stallization (short curves between dashed Iinesin Fig. lOe- f, e.g., grains 2, 5, 14, 18). Intersectionsof the left- and right-hatched fields in the pole figurecorrespond to the relative ly narrow range ofaxialorien tatio ns tha t are characterized by low azimuthaland inclinational rotation rates. lt is probably notcoincidental that these fields of relatively high axialstabüity also eo ntain the c-axis point maxima makingup the bulk texture in norcamphor at intermediate tohigh shear strains (shaded elliptical areas in Fig. 11.reeall Fig. 2). Also, newly nucleated gra ins usuallyshow c-axis orientations within, or at least very closeto. the field of maximum orientational stability inFig. I lb (e.g., grains 19,25,27, 30 in Fig. 10).

The distribution of the orientational stability fieldsin Fig. 11b bears impli cations for the evolution oftexture on the scale of the entire norcamphor sampie.The symmetrical c-axis cross girdle that fonns at lowto intermediate simple shear strains in the first tran­sient stage of the HT-LS experiments (Fig . 2; y = 2)consists of grain orientations in the low rotationalrate fields of Fig. 1Ib. As mentioned above, theserotations primari ly involve glide-induced vorticity(Fig. 8a). In analogy with quartz, slip is inferre dto occur both on the prism [m] slip system and oneonjugate basa l g lide planes in the (a) direction (dis­cussion in Herwegh and Handy. 1996). Thi s allows

911 .\1. n ..,..-r1:h rt al. ITn ',m'Jf'/r.n in JXO ( I W7J 1U-106

-48 27

ps

37J

3235~Y',

q ;37

""35 '

b

6

ps

­•24/

j

o data point 0---:; c-axis rotanon direction * new grain • grain consumed

8r6s•"'3 ---------- --

2 ", .

2 , s 6 , 8a •

" "" e" __ .,: ;23__ __

•

•1 ~ ~- -- - - - - - - - - - - -

19 13ra

00.,

90

'"

sheer strain (1) shear strain (1)

Fig. 10. c·.u i~ n....lion p;>lh~ for indi\'idual grai ns, Eqaal-<Uea pole liga=-~' ing c·a" i~ rotation path~ of indi" idual grams ( iL bl.ALimalh V~, ..tIcar "'r..in di..gr.aJ1b (c. dl . InclillOilion ,,~. ..tIcM ",,,, in diag rams te . O. Tu aid "iwaliL<ilion o f Ihc c-n is rolaliooal hi"'ories.orien t11ions fmm <lflPO'ile quaoJ",nts of the pole figures are combined so thai oo ly a 1lW'" u imulh range is d.;:pickd on tbe "mini axesof (cl and Id ,. Thi~ is jusntied by Ihc ~) mlTl<"try of mo:N: c-iL\ i~ rotaliooal p;>th~ ..ith ~I to the C('fllre of the pole Iigures . Da.Jledh..ri t l....taJ Iines in (c ) 10 10 d.:hneale orirnlalional stabilily dom.ain~ Jepictcd in FiJ;. 12 ( <,« In l for npl;m;ttion).

s1

M. H..,,,,~h et uf.IT« fO'l'>ph,".•;CJ :!xt) 11(97) 83-106

s1

99

-szs-

pi

Type I c-axis rotation pathType 11 c-axts rotanon pathType 111 c-axis rotaticn path

stable c-axis point maxlmametastable c-axis point maxima

pi

low azimuth rotanon rates

low inclination & azimuthrotanon rates (high axial stabi lity)

low incli nation rotation rates

Fig. 11. (a l Tbree general types of c· ;ni, TOI.;lIion paths derived from pole Iigun::s in Fig. 1000b. (b) Pole tigures "oo..,.'ing Iiekh 01 variedr-axis roIalioo rares tsee tUl for explanalion).

pure shcar ex tension ur the norcamphor agg regateparalle l 10 thc long axis ur thc finite strain clli pse .Why doc s suc h a sccmingly stable cross girdlc tcx­turc yicld 10 the c-axis single girdle in the secondtransicnt stage? A glancc al thc c-axis rraicctor icsat the periphery of the pole ügures in Fig. l üa.bprovides a possible explanation: grains that have at ­taincd an or ientation in the periphera l high stabilitytie ld IFig. I lb) cominue to rota ic at a low rare syn­thet ically with respect to the hulk sense of vorti cityIe.g.. grains 4. 6. 20 in Fig. 10). As shown in Fig. 5a.the volume proportion of such grains (vio let and ma­gema grain s) decreases dramatically at shear stra insbctwee n 2 and 4 due ro the increased relative activ­ity of grain boundary migration recrysta llization (scealso Fig. 8a ). Thi s slow rotation of individua l grains'c-axes in the presence of syntcctonic gra in bou ndarymigra tion co ntinucs. with the r- axcs of most surviv­ing grains au aining an orie ntanon coincident withthe steady state poinl maxima (c.g.. grains 35. 36in Fig. 10). Up to stcady slale . thc e rystallographic

rotanon of individual grai ns is acco mpanied by aslight symbetic rotanon of the singfc girdlc skeletaloutlinc with rcspcct ro thc SZ B (tcxt ures for y = 4to 6 in Fig. 2).

At steady statc. roration cf the pcriphcral pointmaxirna with respect to thc SZ B stops. but thc c-axesof individual grains within the aggrcgate continuc tcrotate. Th is is shown in Fig. 12. whic h contains anorientation image seq uence through an orien ranona lslice (azimuth ranges 170-190" and 350-010" at in­e1inations of 0" to 10") correspondi ng to the periph­eral c-axis poin r maxima for the steady stete HT- LSrextute. In this figure . the grey Iones correspondro Z' intcrvals within this azimuth mnge (see up­per left-hand come r of Fig. 12). Deformation bandsand psismaue subgrei ns in Fig. 12 have lew-angleboundaries (2- 100) that are oriented subperpendicu­lar to the SZB. Both the boundaries and the c-axesof the subgrains rota re synthetically (sec deformationband labelled 1 in Fig. 12 and progressively darken­ing subgrain to the [ert of this band). This rotation

100

SZB

M. Herwegh et al. /Tectonophysics 280 ( /997) 83-/06

4 1-.._ ­1

Fig. 12. Changes in crysta llographic orientation and microstructure in a steady state HT-LS experime nt. Pole figure in the upper Ieft-handcorner shows the grey tones corresponding to 2° azimuth intervals for flat-lying c-axes within the periphera l steady state point maxima(see text for further explanation).

M. Hr"" ...glt rl al. I T«llHloph.mcs 280 ( 1997} 81- 106 101

co ntinues until the subgrains are co nsumed by oneof two mcchanisms: (I ) lew-angle grain boundarymigrat ion of adjacent subgrains whosc basa l planesare oriented subparallel to the SZB (subgrain la­bellcd 2 in Fig. 12); o r (2) high-angle grain boundarymigrat ion of nc ighbouring grains (grain labe lIed 4in Fig. 12). A few grains even manage to rotateout of the favourable slip orientaüon before beingeonsumed (see grai n labelled 3 in Fig. 12). The ro­tation path of such grains is similar to that of grain24 in F ig. 10a. Most blae k grains in Fig. 12 arefavourably oriented for prism gfide. as inferred fromtheir c-axis orientations para llel to the Y fabrie di­rection (steady state poinr maxiraum ar the ce ntre ofthe pole figures in Fig . 2). The black grain clusterlabelIed 5 in Fig. 12 undergocs cyclica l growrh . co­alescence. grain size reduction. and dismembennentduring grain boundary migration recrystal lizaricn.Thus, the continuous rotanon of cryst allographicaxes observed on the gra in scale is not perceived onthe sarnp le scale beca use grain boundary migrationusually consumes grains whose predominam slipplanes rorate ou t of an easy glide orientarion. It is in­teresung to nore the similarity of this behaviour withthat predicted by the twinned fibre domain modelby Cobbold and Gapais (1986). In their model, slipdomain boundaries rotate and migrate unti l slip isacco mmodated primari ly (hut not exclusively) alongplanes oriented para llel to the SZB (see the ir com­bined müdes l and 11). This demonstratc s thc im­portance of dynamic recry stallizat ion in preservingsteady state textures.

S. EfTect of lemper.llure and slrAin rale ontexture

In a11 eÄperimcnt s. simple shearing is associatedwith the fonnation of a symmetrie c-u is eross girdlethat is progre ssively replaced by a rotat ing. obliquec-axi s single girdle . which at stcady state e ither stabi­lizes (IT- HS eÄperimems) or yields to a stable c-aÄispoint max ima (HT- LS and HT- HS experiments: seeFig. 13). Despite the similarity of these bulk textures.thc detail s of crystallographic rotation histories areinferred to vary between thc experiments due to dif­ferences in the grain sca le mcchani sm assemblage .Tbe maximum density of orientation images perstra in increment in the IT-HS and LT-HS experi-

ments was too low to enabl e us to track the r-axes ofnorcamphor grains. OOt the generat similarity both ofthe textures and of their strain-de pcndenr transitionsin Fig. 13 leads us to believe that the generat concep tof rotational stability fields (recall Fig. Il b) apphcsro norcamphor in all of thc expcriments. Slight dif­ferences in thc geornetry of the texrure s suggest thattbe relative size of these stability fields varies withtbe extri nsic conditions of simple shearing. For ex­ample. Fig. 14 shows that the ope ning angle betweenthe two legs of the c-axis cross gird le (Fig, 13) issmaller at lower temperatures and hig her strain ratesthan in the HT-LS experiments. Similar tempera­rure-dependem variatio ns in this angular relationshiphave also bce n observed in c-axis cross girdle pat­tems from high-grade quartz mylonites (Behr, 1968).ostenslbly deformed under coaxial co nditions (Lis­rer and Dcmslepen. 1982). This indicates that rhefields of the relati vely fast crystallographic rotationin Fig. ll b (white areas) increa ses at the expense ofthe low rotational rate fields with increasing temper­ature and/ar decreasing strain rate . We attribute this10 the observed increase in thc relative activiry ofsubgrain rotat ion recrystallization at lower tem pera­tures (compare microsuuctures in Figs. 2 and 3; seealso Fig. 8).

The asymmetry of the single and cross girdleswith respec t both to the main folia non. Sa. and thcSZB chan ges as a function of shcar strain in all theexperiments (Fig. 13). Thi s is seen more clearly inFig. 15. where the angle ß betwec n the SZB andthe central segment of the c-ax is ske letal outl inesdec reases with progressive shear strain. In order tovisualize the relative rates of textura! rotation andpassive material rota tion within the sampie. Fig. 15also contains a solid curve represeming the anglebelween the shon axis of the finite strain ell ipse andthe SZB. The fOIation rale of the c-axis cross gird lesduring the first trans ient stage coineides approÄi­mate ly with that of the finite strain e llipse . Duringthe second transient stage ( y = 2 to 6). however. thetextura l rotation rate increases abruptly as the texturechanges from a cross girdle to a single girdle pauern .Thi s has also bcen observed during simple shear dc­fonnation of olivine aggregates (fig. 4 in Zhang andKarato. 1995). Note that the cross gird le to singlegirdle transition in norcamphor generally occ urs atlower stra ins for higher temperatures. presumably

102 M. Herwegn era/J Teclonophysics 280 ( / 99 7) 83-/06

y HT-LS HT-HS IT-HS LT-HS

8 @8-0 - . ' "~ - Q ' .., -'

, '

. ''fjj) '~. - . - ~-

s~- ,<:;,~ -- ' _.'2 ~~ 0

. __ , ' __-----f ~ - - - -J--s - " /saJ-- ' - "'-

4 - S@ S$ s€fJ 0-- , - - -- , - :

6 ~:@sa s(1) ~€&, se ,ffiaJÖ a 1La "Sa

s~, 0 - se1 .1 first transient

,"" stageo seco nd transient

stageo steady state

Fig. 13. Summary ur pole tigurcs showing sample c-axis texw res for all the experiments. Equal-area u mlour imervals are 0.25 , 0.5, I,4,8. 16, 32 umes uniform distribution fOTthe HT- LS. IT-H5. and LT- HS experiments . contour intervals are 1,2.4.8. 16 times uniformdisirib urion for the HT- HS experimerus. Dashcd lincs represer urne shcar zone Iloundary (SZB); Sa. refers [0 rhe long 311is ur the finite

strain ellipse.

due to the higher relative activity of grain boundarymigration recrystallizalion in the high-ternpcrarureexperiments (Fig. 8). AI steady stare. the tcxtures

SlOP rotating and the ß angle stabilizes at values of86--88", The sampie textures have artained a stabteend oricntation with respect to the SZB.

M. Hern'eKh er al. /Tecwrwphysics 280 ( / 997) 83-/06 103

6. Resolu tion of a long-standing debate?

rotational orienta tion with respect to the SZB duringfurther strain (e.g., p. 37 in Law et al., 1990). There­fore, the stable tcxture of an aggregate at steady steteis believed to reflect a stationary orientation of thcindividual grains making up the aggregate. In con­trast. the ideal orientatio n concept tnvolves cxplici tassumprions about the way in which strain compati­bility is maintained within thc aggregate in ordcr 10

generate a model texture. As stared in thc Introduc­non, rnost models that invoke the ideal orientationconcepr neglect dynamic recrystallization (excepr forJessell. 1988a,b; Jessel l and Lister, 1990) and as­sumc either uniform strain {Lister et al., 1978) oraverege (i.c . self-consistent) strain within the aggre­gate (Wenk ct al., 1989). These assumptions leadto discontinuously rota ting slip sysrems. In quartz.this is associated with point maxima whose obliquitywith respect to thc SZB appears to be opposite tothat observed in nature and experiment (e.g.• see fig.9 in Wenk and Chr isüe, 1991).

Our norcamphor experiments suggest that al­though both approachcs explain certain aspects oftextute forma tion , ncither approach is entirely valid.While the experimental textures certa inly attain astable end orientation on the seale of the norcamphorsamples. the slip systerns of individual grain are nOIstationary at steady state and can even undergo alimited amount of rotation out of easy glide orienta­tions before being consumed by neighbouring grainsduring dynamic recrystallization. In fact, dynamicrecrystatlization plays a central role in maintainingstram compatiblll ty within a polycrystallinc aggre­gate defonning at high strain s, as Schmid (1994)and co-workers (Schmid and Casey, 1986; Schmidet al., 1987) have already noted. In addition. pre­vious experiments in norcamphor (Herwegh andHandy. 1996) have shown that microshearing on thcsupragranular scale wirhin the aggregate cor ur ibutesto maintaining strain compatibility on the granularscale. Both of thesc effects should be incorporatedinto future modelling of polycryslalline aggregates.

We end this section on a semantic note by rec­ommending that the tenn 's table end orientation' beapplied only to bulk textures at steady state. Only onthe bulk (i.e. aggregate) scale is the crystallographicoricntation of grains statisticaJly invariant with slrainand time. This tenn should be avoided, however,when dcscribing the cryslallographic behaviour of

SZB~•

60.. 50 • HT-LSsn0 - - - - - - ~ - -• • • HT-HS~ 40 - ~ --0

(j)- - - -·c • IT-HS• 30 • &c,

0 LT·HS20

0 2 3 4 5 6shear straln (y)

130

C 120~

• 110C>

1000es

90

800

. HT-LS . HT-HS .A IT·HS .LT-HS

Fig. 14. Tempereture and strain rate dcpendence of the openingangle between legs of the c-axis cross girdles in Fig. 13.

2 4 6 8 10 12shear strain (""t)

• HT·LS • HT·HS .A IT-HS • LT-HS

Fig. 15. Texrural obfiquity with respect to rne SZB (angle ß )

v,. shcar strain at different experimental condilions. Solid curveis the angle betwee n the ,hon axis of thc finite strain ellipseand thc SZB. Filled and open symbols represcnt the main limbs.respectively, of thc cross gtrdles and single girdles. Dashed lineat 90" depicts the normalto the SZB.

Thc resuns of thc texrural analysis presentedabove may shed some ncw light on the contro­versy regarding texture evoluticn outlined in theintroduction . Before discussmg thc norcamphor ex­periments in this context . it is important to considcrthe different approaches under lying the srable endorientation and ideal orientation conce pts. Implicitin the concept of a stable texrural end orientation isthe notion that slip systems in the grain aggregaterotate until thcy attain an orientation that maximize sthe resolved shear stress in the slip direction (i.e. thecasy glide orientalion; Schmid and Casey. 1986). Inaccordance with Ihe modelling work of Elchecopar(1977) and Elchccopar and Vasseur (1987), theseplanes are thcn assumcd 10 maintain this stable. ir-

'01

indiv idual grains ur subgrains wirhin an aggregate.bccause the microfabric studies above show thatsteady state on the granular scale is both herero­geneous and dynamic. involving the cyclic growtb.rotat ion and consumprion of individual grains.

7. Geological applications

Thc mic rofabri cs gencru tcd in the norcamphor ex ­pcrimcms dcscribcd ubove bcar directly on the wayin which qua rtz microstructu res ca n bc used in fie ldstudies. Fig. 16 shows that no sense of shear canbc derived at low sbcar strains unless the SZB iscxposed. because thc gra in SPO and the c-axis crossgirdle pattem are symmetrically disposed with re­specr both 10 each other and to tbe main foliation,Sa. Al intermediate 10 high shear suai ns. howeve r,the obl iquity of the skeleta l outline of cross or singlegird les with respect to thc main foliation . as weilas the prono unced SPO in dynamica lly recrysrallizedaggregares. are both good kinemati c indicators (see

also Burg and Laurem . 1978; Simpson. 1980; Simp­son and Schmid. 1983). Note that symmetrical crossgirdle panerns form during the first tran siem stageof simple shearing and do not necessarily indicat ean ove rall regime of coax iai llow (Hudleston. 1978;Bouchez and Duval. 1982; Hcrwegb and Handy.1996).

Tbc microfabrics and mcchanism asse mblagescan also Oe used to makc qualit ative inferences abouttempcrature and strain rate. In genoral. a predo m­inance of grain boundary migration rec rystalliza­tion is diagnostic of high homologous temperature.whcreas subgrain rotation recrystallization indicatesrelativel y low homologous tempcrarure and/ar highstrain rate . A similar depcndence of dynamic re­crysta llizat ion mechanism on temperature and strainrate has been interred for deformed minera ls (e .g.•quartz. Schmid and Casey, 1986; ce tcire. SchmidCl al., 1987) and inorgan ic rock analogues (magne­sium. Drury er al .• 1985; sodium nitra te, Tungau andHumphreys. 1981).

shear I textural element SPO fjeld criteriastrain ßcross giedle c-eoe cross giedle

I~- subperpendicular il

SO@~B10 SPO anc Sa .-

~cross girdle "'>no shear sens e! Ei ä

109' < ~ > 120· "'0cross girdle is ~ .

asymmelrie os- ~.

:.- posed 10 meSZB=> shear sense

Iransiljon cross · single girdle- •

$~B - c-eos cross giedle •eees Q1rd1e and si llg le girdle n0

~90"<11 >106 stay perpend lC1J lar .0

10 Ihe SZB g~

- =~ both are asym - • •0......- metrically orienled ~.

11 -90 " 10 Sa andSPO •"'> shear sense :.-stable single girdle sfable oonr rnaxima- - ~

~~Beso......- c-axe single giedle

~86<11 >90 " and point maJlima •0.

stay oblique 10 lhe -cpoint mau"a SZB. saand SPO •

1 86 < 11 > 90 " => shea r sense iü- - 0;,

Fig. 16. Field cnrenaIor shear sense determination in mylomtic quanz-bearing ruck al differenl shear slmins.

M. Hn wf:}lh et u/. / Tecum ophysics 280 ( / 997) 83-106 105

Finally, we point out that the experimental de­formation of rock analogucs is no substitute forcareful experimentation on real mineral aggregates.In particular. further studies are needed to see if ourobservations and inferences also hold for hexagonalminerals such as quartz or ice. Only the calibrationof the basic texrural and microstruetural relation­ships observed in in situ experiments can improvethe application of microstructures to solving geo­logie problems.

Acknowledgements

We wish to thank Tectonophysics reviewers C.Passchier and Anonymous for thcir helpful com­ments. The financial support of the Swiss NationalScie nce Foundation in the fonn of project grants 21­30598.9 1 and 21-338 14.92 to Mark Handy and grant21-36008 .92 to Renee Heilbronner is acknowledgedwith gratitude.

References

Behr, H.J.. 1968. Zur tektonischen Analyse magmatischer Körperun ter beso nderer Berücksichtigung des Quarzkomgefüges. I.Freibe rget Forsch. C215 , 9-59.

Bons, P.D., 1993. Experimen tal dcformation of polyphase ruckanalog ues. oeor .Ultrajeelina 110. 207 pp.

Bc uchez, J-L., Duval, P.. 1982. The fabrie of poly crystaltineice deformed in simple shear : experiments in rorsion. naturaldefonnal ion and geome trieal interpretatio n. Text. Mic rostruct.S. 17 1-190.

Burg, J .P., Laurent. Ph., 1978. Strain analysis of a shear zcne ina granodiorile. Tec tonophysic s 47, 15-42.

Burg, J .p.. w üson. CJ-L. , Mirehel. J.C.. 1986. Dynarme re­crystalliza rion and fahrie developrnent during simple sheardefonnation of ice . J . Struct . Geo l. 8. 857-870.

Cohbold. P.R., Gapais . D., 1986. Slip syste m domein s. I. Planestrain kinemauc s of arrays of coherenr bands with twinncdfibre orientations. Tectonophysics 131.11 3-132.

DeIl'Angelo, L.N .. Tul1 is. J .. 1989. Fabric developrnent in exper­imema lly sheared quam.ites. Tectonophysics 169. 1-21.

Drury. M.R., Urai, J .L.. 1990, Defonnation·relaled recrystalliza­tio n pruces..;es. Tectonophysics 172, 235- 253.

Drury, M.R.. Humphreys. E J., While, S., 1985. Large straindefonnation sludies using polyeryslall ine magnesium 3,' a ruckanalogue, 11. Dynamie recryslal1ization meehan ism al highte mperatures. Phys. Eanh Planet. Inter. 40 , 208--222.

Elehecopar, A" 1977. A plane kinematic model of progressivedeformation in a polyerystall ine aggregate. Tectonophysies 39,121- 139.

Etehecopar, A.. Vasseur. G., 1987. A 3-D kinematie model offabrie developmenl in polyerystalline aggrega les: eomparison

with experimental and natural examples. J. Struct. Geol. 9.705- 7 17.

Handy. M.R., 1986. Th e Structure and Rheological Evolutionof rhc Pogallo Fault Zon e, a Deep Crustal Disloc ation in theSouthern Alps of Nc rtbwestern ltaly (Prov. Novara) , Unpub­lishcd Ph.D, Thesis, umv. of Basel. 327 pp.

Herwegh, M., 1996. Mic rofab ric Evolutio n in MonomineralieMylonites; An Experimental Approach Using See-TbroughAnalogue Materials . Unpublished Ph.D. Thesi s, Univ. ofSeme,

Herwegh, M., Handy, M.R.. 1996 . The evolution of high ternper­ature mylonitic micro fabries: evldence from simple shearing ofa quart l analogue (norcamphor). J. Struct. Gecl. 18. 689-7 10.

Herwegh. M.. Handy. M.R. , 1997. The origin of shape preferrredorientations in mylonite: interences from in-shu exper nnenteon polyc rystalline norcamphor. J. Struct. Gco l., submiUed.

Hirth. G.. Tullis. J.T., 1992. Dislccation creep regimes in quartzaggregate s. J . Struet. Geol. 14. 145-159.

Hudleston, P.J.. 1978. Progressive deformation and developmentof fabric acros s zones of shcar in glacial ice. In; Saxena. S..Bbauac harji. S . (Eds.). Energetics of Goological Processes .Springe r Verlag, Berlin, pp. 121- 150 .

Jessell. M.W.. 1986. Grain boundary rmgranon and fahrie de­velopment in experimentally dcformcd octaeh loropropane. J .Struct. Geol. 8. 527-542.

Jessel1 , M.W.. 1988a . Simulation of fabrie developm ent in re­crystallizing aggre gates. I. Descnption of thc model. J. Sttuct.ceot 10, 771-778.

Jessetl. M.W., 1988b. Simulation of fabrie development in re­c rystalliz ing aggregates. 11. Example model runs. J . Struct.Geol. 10,779-793.

Jessell. M.W., Liste r, G.S., 1990. A simulation of the lemperaturedcpendence of quart z fabric s. In: Knipe. RJ.. Rutte r, E.H.(Eds.), Deformatio n Meehanisms. Rheclo gy and Teetonics.Geo logieal Soc iely, London. pp. 353-362.

Knipe, RJ ., Law, R.D.. 1987. The inftuence of crystallographicorientation and grain houndary mig ration on microstructura land texrural evoluucn in an s--c mylonite. Tectonophysics135, 155- 169 .

Law. R.D., Schmid . S.M., wheeler. J ., 1990. Simple shear defor­mation and quanz fabric s: a possible natura l exam ple Irom theTorridon area of NW Scotland. J . Struct. Geol. 12. 29-45.

Lister, G.S., 1982. A vorticity equalion for lattice reorien tationduring plastic deformatio n. Tcctonophysics 82. 351-366.

Liste t, G.S. , Dom siepen, V,F.. 1982. Fahrie transitions in theSaxony granulite terrain. J . Struct . Geol. 4, 8 1-92.

Lisle r. G.S .. Hobbs, B.E., 19SO. The simulal ion of fabrie de­velopment during plastie defo nnation and its app lieation toquartzite: lhe influenee of defonnation hislory. J . Siruet. Gco1.2. 355- 370.

Lisler, G.S., Wi1l iams. P.F.. 1983. Th e partilioning of defonna­lion in ftowing rock masses. Tcelonophysies 92 . 1- 33.

Li,ter. G.S., Paterson. M.S., Hohhs, RE., 1978. The simulal ionof fabrie development during plastic defonn ation and its ap­plieation 10 quartzi le: lhe model. Tectonophysi es 45 , 107- 158.

Mancktelow. N., 1993. Compuler program SlereoPlotXL. ET H.Züric h.

106 M. H('nI "{;'Xh r:1 111.1 1i.·clmw phy.•ic:" 21M ( ( 997) 1i3-/O{)

Mcans. W.D.. 1977 . A deformation expenmcnt in transmitredlight. Earth Planet. Sei. Leu. 35. 169- 179 .

Means. W.D.. 1989. Synkinematic rmcro-ccpy of tran,p;.ren\polycrysials. J. Struct . Geo l. 11. 16J-174.

Meao !>. W,D.. Dong. H,G.• 19~2 . Some unexpectcd effects ofrecr)'!'lall izalioo 011 tbe microcrectcres o f matenals dr:formedat high tenpwllure. Min . Grol. In.... Eidg. Tech. Hoc h<.ch.Univ. :lunch 239a. 205-207.

Panouo-Heilbronncr. R., PallIi. C . 1993. lntegrated~Iial andorienlalKln anal ysis of quaru c-exe, by oomperer-aiced mi­CfO'oOOPY. J. Struet. GeoI . 15. 369-383.

Panouo-Hc:i lbronncr. R.• Pauli, C~ 1994. Oneneucn and mi,..oricnLltion imagi ng: integration (Ir microslruc tur.ll and 1("1IuI';I1

;malys.is. In: Bunge. HJ., Sicge>;mulld. S .• Skrotrl:.i.W~ Weber.K. (&Is. ). Textures of Geo logieal Materia ls, DGM Informa­lion 'ge~lIschafl Ver lag. pp. I·H -I M .

Poiric:r. J.P.. Gui l1ope, M .• 1979. [),:formali Oll ioduced recrys tal­

hzation of rninerals. Bult. Mineral . 102. 67- 7.f.Ra'ihaod. w.. 1995. Computer Pmgram Image 1.57. National

lnstiune of Health. Research Services Branch. developed atthe U.S. National In'liMe' of Hcalth and available on Internetat hllp:flrsb-info.nih.gov/nih· imagc.

Ree. I. H.• 1991. An expe rimental , teadY" lale folialion. J. Stru ct .

GL"OI. 13. 1001-1011.

Sander. B.. 1950. Einführung in die Gefügekund e der geologis­chen Körper. 2. Die Komgefüge . Springer Verlag. 409 pp .

Schmid. S.M.. Im . Textures of geolog ical materials: computermodel predidions versus emprocal tnterpreranons based onrock deformauon Ul'C'rimenls aod neid studies. In: Bunge.H.1.• Siegesmund. S~ SkrotAl. W.• Weber. K. (Eds.). Tu ­tu~ of Geological Materia ls. DG~t. Informa lion' l!est' llsetlaflVerlag. pp. 179-30 1.

Schmid. S.M.. Casey. \I.. 191\6. Cornplete fabrie analysis of

",me commonly observcd quart ' c-axis panems . In: Hobbs.B E.. Heard . H.C. (Eds.], Mineral and Rock Deformauon :LaboralOl)' Sludie, - Thc Pallcrson Volume. Am. Gcophys.Union. Gcopbys. ~1onogr. 36. 161-1 99.

Scbmid. S."-I_. Handy. M.R.• 199 1. Toward' a genetic da." ifi.caticon of fault roch: gcological usage and IrCtonophy, icalimplications . In: Müller. D.W.. McKrnzie. tA.. Wei, <;CT1. H.tEds.). Cont fO\'ersies in Modem Geology. Acade mic Press..Loedoe, pp. 9S--11O.

Schmid. S.M.. Paoouo. R.. Bauer. S.. 1987. Simple!hear upc1'"­imen ts on cakue rocks: rheoolot:y and microfabric. J. Struct .f".oeol. 9. 7.f7_ns.

SchmidL w.. 1927. UnteNIChungrn über d ie Regelunj! derQuarz j!eftige kristalliner Schiefer. FOf1seltr. MineraL KOstal­logr. Petrogr. 11. 33+-3·H.

Simpson. c.. 1980. Ob lique girdle oriema tioe patteras of quanzC·axes from a shear zone in ee ha--.emenl coee of tbe ~1aj!gi a

Nappe Tk mo. Switzerland. J. SITUCI. Geot, 2. 243-2.f7.Simpson. c.. Sc hmid . S.M.• 198 3. An evaluanon of criteria to

deduce the sense of movement in sheared rocks. Bull. Geol.SOl.'. Am . 94. 1281-1 288 .

Tungatt. P.D.. Humphreys. F.J.. 1981. An in situ op ncat inves­ligatiun of the dcformation bchavior of sodiurn nitrate - ananalngue of calctte. TCCI" nnphysic, 711, 661---676.

w cnk. H,-R.• Christie , I .M.• 1991. Commenls on the intcrpre­tarion of dcformanon texfures in roch . J. Struct. Geol. 13.J()9I - I I IO.

w enk. H.-R.. Cancva. G.. Molm ari. A.. Koch. U.F.. 1989.Visroplastic modelling of tevture develop ment in quartzue.J. Geophys. Res. 94. 1789S--1 7906 .

Zhang, 5.. Karate. 5.• 1995. Lausee preferred on entaüon ofoli,·ine aj!gn-gate<; deformed in simple shear. r\ atun- 375.774...n7.

Publ lctl tlon InformetlonTectonoptrysics (ISSN 0040-1951). FOf 1997 volumes 265-:-278 ere scheduled 10<pubhcahon . Subsctiplion prces ere av ailab!e upon request Iromlhe publiSh8f. Subscript'ons are accep ted on a prepaidbaSIS onIy and ere entered on a cerercer year baSIS. rssues are sent by surface mai t exceot10 Itle followiog countries where air 6el ivery via SAL IS ensured : Argentlna, Austra l18, Br8z,I, Canada, Hong Kong, India, Israel, Japan , Malay sia ,MexiCO, New Ze ala nd, Pak istan, PR Chi na, Singapore. Sout h Alrica . South Korea. Taiwan , Thai land , USA. Fo< all o ther coun tnes airma il retes ereavaila.ble upon request, CJaims lOf mis siog rssces must be made wilh in six mootns 01 our publi calion (mailing) oeie Fo. croers. claims, productenqu iries (no manuscnpt enquiries) please contacl the Customer Support Departm&nl at the Reg ;onal seresOffice nearestto you:N_ Yotk, Elsev i8f Science, P.O. Box945, New YorlI , NY 10159..0945, USA . Tel: (+ 1 ) 2 1 2~·3730, (Toll Free numoer 10. North American ccs­tome rs: 1-888-4ES-tNFO (437 -4636)1, Fax : (+1) 212-633-3680, E-mail; usinfo-IOelsevier,comAmsterdam, Elsevier Science, P.O. Box 211, 1000 AE Ams lMdam, The Nelhe rfands. Tel: (+31)20-485-3757, Fax: (+3 1) 20-485-3432, E-mail:nlinlo-lOelsevi er.nlTokyo . Elsevie r sceoce. 9-15, Higashi·Azabu t-eere. Minato-k u, ToI<yo 106 , Japa n. Tel: (+81) 3-5561· 5033. Fax (+61)3-5561-5047, E·mail:kyf040350nifty sa rve.o. ,jpSlngapote. Elsevier scerce.No. 1 Temase k Avenue , 117-01 Millen ia Tower . Singapore 039192. Tel : (+65) 434-3727, Fax : (~) 337·2230, E-mail:asiai nto Oelsevier.oom.sg

US mai liog nctce - Tectonophysics (ISSN 0040-1951) is published bi-weeldy by Elsevier seerce BV (Molenwerf " Postbus 211, 1000 AEAmsterdam). Annual subscrtptce price in the USA US$ 2505 (US$ price valid in North , Ce ntra l and South America on ly) , including ai r speeddelivery, PEHiodicais postage paid at Jamaica, NY 1143 1.USA POSTMA STERS: Send address changes 10 Tectonophysics, Publications Expediling , lne., 200 Meacham AVllfl lle, Elmon l. NY 11003.Airfreighl end mailing in lhe USA by Publications Expediting.

AdVertising In' ormetlonAdvertising oroers end erqoutes may be sant 10: Elsevier seeree. Adv ertising Departmenl. The Boulevard, Langford Lane, Kidlington , OlCIord, OXS1GB, UK, tel.: (+44) (0) 1865 843565, lax : (+44 ) (0) 1865 843952, In fhe USA anc1 Canada: Weston Media Associales, atln. Dan üpner, P.O. Bo x1110, Greens Fanns, Cl 06436-1110, USA, tel ,: (203)261 2500, lax: (203) 261 0101. In Japan : Elsevier sceoce Japan, Marf<eting sevces. 1-9-15Higashi-Azabu, Minato-ku, ToI<yo 106. Japan, lel .: (+61) 3 5561 5033, fax : (+61 )35561 5047 .

NOTETO CONTRIBUTORS

A detailed Guide for Authors is avai lable on requesl. Please pay all8f1tion 10 the foIlowing noles:

Uu",m..The otficiallanguage 01 tf1e jou rna l is English ,

Preparation of tfl9 ttm(a) The manuscri pl should pre ferably be prepared on e ward peccessorand plinted with double spadng end wide margins and include an abstract04 nol more tben 500 words,(b) Authors should use lUGS tem1inology. The use of S.I . unrts is also recommended.(c) The lilie page Should include lhe name(s) 04 lhe author( s), their alliliations, lax a l1<l e-mai l numbers In case of mors \ha n one author, pleasairldica le 10whom the correspondenee shoukl be addressed.

RefereflC6s(a) Aelerences in the text con sist cl the surnarne ollhe author(s), lollowed by lhe year of publicalion in parentheses, All relllfenc8S eiled in the lelltshould be given in the reference list and vice versa .(b) The relerence list Shou ld bEI in alphabetical order.

TabiesTables shQukl bEI complled onseparate sl'leets and shoul d be numbered accordlng 10 their SBqUllflCll in Itle lellt. Tables can also bEI sant as glossyprints 10 avoid errors in typesetting.

IIlustra/ions(a) mustrations should be submille(! in lripl icate. Please nol e that upon submission 01 a manuscripl rtIree sets01 all pho lographic material prin tedsharply on gIossy paper 0. as fugn-i1e~'nition las8I prinls must be prov ided 10enable meanin gful review. Photocoples and ofher low-quality prinls willnot be accepted fo<rev iew.(b) CoIour l igure s can be accepted providing lhe reproduction costs are met by lhe aulhor. P~e consult the pvbIisher 10. lurther informa tion

Page proofsOne seI of page proofs will be sant 10 the corre spondi ng aUlhor, 10be checked lOf typesellinglelllling. The aulhor is nol expecl ed to make changesOf corrections tha I con stitute departures Iram the Micle in ils accep led form , To avoid pos lai delay, autho rs are requested 10 retum corrections 10 Itledesk-edi to., Mr. He rrnan E. Engelen, by FAX (+3120,4852459\) or e-ma il (h.angelen @elsavier,nl), pt eferably within 3 days .

ReprintsFifty replints ot ooch article published are supplied Iree of charge. Addilionaf reprinls can be ordered on a repli nt order form , whM:h will be sent to thecorresponding aulhor upon acc epl ance 01lhe artide,

Submission 01manuscnp/sTh ree copi es shoukl be sub mined to: Ed itolial Off ice Tectonophysics, P.O. Box 1930, 1000 BX Amsterdam, The Nelherfands.Submiss ion of an article is understood 10imply thatthe article is orig inal and unpubl ished and is not being conSldered /or publicat ion elsawhere.Upon acceplance of an arficle by the journal, lhe auth or(s) will bEI aske(! 10 transf er lhe copynghl of the art ic le to lhe pub lisher, This transfer willensure the widest poss ible disseminat ion 01information under lhe USoCopyright LawThe indicalion o f a fax and e-mai l number on submission of lhe rnanuscripl cou ld assis t in speedIng communicat ions, The fa x number lo r theAmsterda m office is +3 1-20-4852696 .Aurtlo~ ,n Japan. please /lOte: Upon req uest, Elsev ier Science Japan will prov ide autho~ with a list cl peopl e wI10 can check and improv e theEngli Sh 01thelf paper (oofore submission ), Please conlacl our Tokyo olhcB: Elsev ier Science Japa n. 1·9· 15 Higashi·Azabu, Minato·ku, Tokyo t06:Tel. (+8 t) 3 556 1 5032 ; Fax (+8 t) 3 556 t 5045.THERE ARE NO PAGE CHARGES

Submission ot el8Crronic texlIn order to publ ish tM paper as quickfy as possible aller accep tanee aUlhors are e<lCOuraged 10 subm.t lhe final te xt also on a 3.5 ' or 5.25 ' d,sketteEsse ntial is tflal the nam e and vers ion 01the wordp rocess ing program. type of com puter on wh,ch lhe lexl was prepa red . and lormat 01 the text filesare clearly Ind icated . Authon> are f8q Uested 10 ensure that apart trom any such sma illast·mlnule co rrectlOf'S. the d,sk vers ion corresponds exaclly10 lhe hafdcopy.11availab le, elecl ron ic lil es of !he figures shoukl also be Included on a separate 1I0ppy d'Sk.