Journal ifClaeiology

........... ::~. .... C.. .. •• .." :: ••

.. .. ............ .. .. .... ..

B •• • . .. ......

.......... o o

00 0

o 00

o o

o

o A

. . .

t-

t-

1 0-7 +--r---r-..,....,...,.......T'"TT----T--.,.--r-TT"1'""1""Tr--r-~1_ 0_1

A

B

c

10

Octahedral shear strain (%)

W-"CD ••••• '0 •

. '.-':',: : . ; ... : .~ -: /J.i '0, t.: .. ' .. :. ;\: :.: l'

: ': ,.' ' 0 ' ,I . ' .

.-:.~ CD~ .;~

~~ \V~

Fig. 2. CrreJ) curvesfordiformation lests 011 rectangular oblong samples in simple shear ai - 2°C and 0.3 i1IPa. Samples were initialLy isotrolJie (test .r!), exhibiLing a two -maxima oystal­

Jabric paltem (test B), and exhibiting a single-/Jole Clystal­fabric pattern (lest C). Also showlI aTe the Sehmidl equal­area c-axisJabric diagrams at the start (left) and conclusion ( right) if the tests. For IheJabrie diagrams, the shear direc­tion is 10 the right, and jor the initial sam/Jle if test B, the samjJle length was aligned lift to right during the comlJression test.

woul d ha\'e been no minim.um stra in rate: the creep curve would have approached direc tly to the steady-state strain­rate value. The crys tal-fabr ic pa ttern at the conclusion of this tcs t was simila r to but slightl y stronger (th e m edi an c-ax is angle to thc \ 'ertical changes from 20 to 17°) tha n the ini tial pattern .

Tests B (horizontal shear o n the ice with vertical com­pression fabrics ) shO\I' st ra in rates decreasing to a minimum \'a lue, then increasing with stra in, again to a steady-s tate \'a lue simil ar to that in the o ther tests. Th e minimum strain ra te is (for bo th cases ) a fac to r of ",2.5 greater than the iso­tropic minimum strain ra te; i.e shear of ice exhibiting a small-circle g irdle or a two-m ax im a fabric pa ttern, in the direct ion perpendicul ar to the sym metry axis of the fa bric, generates easy glide with a stra in rate a fac to r of rv2.5 great­er th an the minimum iso tropic strain rate. \ Vith increas ing stra in, the st rain rate increases ( 0 a terti a ry ra te similar to

672

those for the o ther shea r test . Th e crystal fabric a t th e con­clusion of this test, aga in, is simil a r to the fabric a t the con­clusion of the o the r two tests.

CONCLUSION

Vertical compression fabri cs a ft er la rge strain tend to ha\'e a degree of concentration of c axes towards the ve rtical char­acterized by (I) in unconfined compression, at higher stres­ses and temper a tures leading to rec rystallization, a small­circle girdle pattern wi th medi a n a ngle to the ve rtica l of' rv25° to 400

; (2) in compression co nfined in the line of fl ow, aga in at higher temperatures and stresses leading to rec rys­ta llization, a two-m ax ima pattern with simila r m edian a ngle; or (3) a g irdle-like rotat ion fabric generated a t lower stresses and temperatures than requi red for forma tion o[ a true girdl e.

The hori zonta l shea r rates of samples with verti cal com­pression fabrics exhibit fl ow enha ncements of rv2.5 relati\'e to the minimum stra in rates [or isot ropic ice. \Vith increas­ing shear strain the enhancement increases to rv lO with the de\'elopmelll of a strong single m aximum fabr ic. These fab­ri c and strain-rate developments a re similar to those occu r­r ing in ice sheets from vertical compression near the surface to high shea r with increasing depth

Antarctic CRC and Australian Antarctic D ivision, Box 252-80, Hobart, Tasmania 7001, rlllstralia

17 September 1998

REFERENCES

LIJCN TH.JACKA

A ll ey, R.B. 1992. Fl ow- law hypot heses lell' ict-shect modeli ng. ] Gtnrioi., 38(129), 245- 256.

Budd. W. E a nd T H , J acka. 1989. A rcv icw of ice rheo logy 101' ice sheet modell ing. Cold Reg. Sri. TedllloI.. 16 (2), 107 144.

Gao. X . Q 1992. Laborato ry studies of the development of a nisotropic crys­ta l struc ture and the fl ow properties of ice. Ph. D. thesis. Un iH'rsit), of :\ lelbou rne.)

J acka. T H. and R . C. Lile. 19fH. Sample preparat ion lechn iques and com­pression apparatus for ice flow sludies. Cotd Reg. Sri. TeclIllOI., 8 (3), 235 2+0.

J acka. T H. and :\'1. i\l accagnan. 198+. Ice crysta ll ographic a nd st ra in rate changes with stra in in compression a nd ex tension. Cold Reg. Sri. Teelinoi .. 8(3), 269 286.

LiJ un, T H. J acka a nd \\'. 1'. Budd. 1996. D eformation ra tes in combined comprcssion a nd shear for ice wh ich is init ia ll y isotropic a nd afier the dCI'Clopment o f st rong anisotrop), . .-1 1111. Glaciot. , 23, 2+7 252.

SIR,

TeclllZiqllejoT impTOving COTe quality in intermediate-depth ice drilling

One of the factors that de termine the quality of pa laeoe nvi­ronmental records obtained by thc a na lysis of ice cores is the q uality of the ice core itself. Broken or cracked co res a re dif­fi cult to sample even for simple m easurements a nd can be unusable for contaminati on-sensiti ve studies such as trapped-air and trace metals. Core qua lity in inte rmediate­depth dry-hole drilling is di sc ll ssed by Gillet a nd o thers (1984), Schwander a nd Rufli (1994) a nd Shoji (1994).

In the course of dry-hole mechanical drilling a 270 m

Downloaded from https://www.cambridge.org/core. 05 Feb 2022 at 07:10:42, subject to the Cambridge Core terms of use.

deep core on Law D ome, Eas t Anta rcti ca, over the 1997- 98 season, it was found that a weight of ",4 kg resting on th e core during cutting g reatly reduced core breakage. An Eclipse IT drill (Bla ke a nd others, 1998) was being used to drill in the high-accumul ation (0.7 m a I ice-equ i\ 'a lent ) a rea near the summit of Law Dome. This drill, which takcs Im long, 80 mm diameter cores, is designed to be used in a dry hole, but the L aw D ome drill had a sealed mOLOr and gea rbox assembly which allowed drilling in a sma ll dept h of fluid. Immersing the head and pa rt of the co re barrel in fluid has been sugges ted as a method of redur ing core break­age (Na rita, 1994).

Drilling down to 90 m proceeded sm oothl y, with the re­trieved core being of ,"cry good qua lity. Below 90 m, co re qua lit y started to de teriorate, and by 102 m 111 0st of the re­covered co re was in the form of fl a t d iscs, a few cm thi ck, with multipl e internal fra cture lines. Efforts to improve core quality included drilling at different speeds (110- 75 rpm ), \'ar ying the depth of cut (5 to < 2 mm/re\'.), va rying cabl e tension and drilling shorter co res (650 mm as opposed to

1 m ). Other tests involved elimina ting the load of chips on the core by susp ending the chip-sepa ra tor plug (which nor­m a lly rides on LOp of the core) by a string from the top of th e core barrel and drilling with the co rc"degs retracted. It was known that cutters with rounded faces ~1ad been sugges ted as reducing corc brcakage (Gillet and others, 1984-), so one se t of cutlers had their inner cdges res haped with a ra dius of about Imm. It was al so thought that fricti on be tween the core a nd the co re ba rrel (and in p a rti cul ar any protru­ding sc rews) might tend to twist the core, so the ba rrel was honed to gi\T a hig hly poli shed surface. Nonc of these tech­niques or modifications had any eITec t on the co re quality.

It was linally decided that, despite the ri sk of conta mina­ti on for trace-chemical measurements, fluid wo uld be used to lubrica te the drilling process. Twent y litres of kerosene was lowered dow n the hole in a bladder and released a t the bo ttom. The first drill run in fluid produced no noti ceable difference in the top sec ti on of the co re, but the bottom sec­tion showed some impro\Tment. A second ro re showed m arkcd imprO\'em cnt, wi th onl y one break, but subscqucnt cores were not as good, and four co res aft er the fluid was placed in the hole, co re quality was again unacce pta ble. At this stage, most of the fluid had been brought back up from the hole with thc cuttings, and the chips were almos t dry aga in. Adding a no ther 15 L of kerosene produced only a sli ght imprO\'ement fo r two co res, and three co res la ter qua l­it y was again unaccepta ble.

It was noticed tha t while the top third of most cores was rubbl e or \-ery thin "pucks", the bottom two-thirds was co n­siderably less broken. II was therefore decided that instead of trying to keep the weight of chips 0 [[ the core, a weight, which simulated the top section of a co re, would be placed in the core ba rrel a boye the chip-sep a rator plug. Th e first test, using a 1.8 kg weight, resulted in a core with just two breaks. As a check, the foll owing co re was drilled w ith out the weight; its top third was again badly broken. From the n on, a ll cores we re drill ed with a weight in the co re ba rrel (except [or another check at 146 m where the top third was aga in badly bro ken ). Adding a furth e r 1.7 kg did no t appea r to res ult in any furth er improYement, but to make a more

Alor.r;an and Olhers: Cones/Jolldence

convenient system fo r the remainder of the drilling down to 270 m, a 3.5 kg weight which took the pl ace of the ehip­sepa rato r plug was made. Ice co res still appea red to be \'e ry brittle, often breaking into sc\"C ral pieces a ft er remO\'al from the drill ba rrel, but the complete di sintegra ti on of co re sec­ti ons tha t was ex peri e nced without the weight no longe r occurred. L ate r exa mina tio n showed tha t co res drill ed \\ith the weight, a lthough unbroken, still had ex tensive interna l fracturing. This does no t a fTec t isotope-ra t io and perox ide­concentra tion measurem ents, but makes the cores unsui ta­ble for trace-chemica l a na lysis and proba bly not usable fo r trapped-a ir studi es.

In 1998, a weighted sep a ra tor plug was tested dur ing drilling on Devon Island , Canadi an Arctic A rchipelago (personal communication from i'd. D. G eras imoff, 199B). It was obsen 'ed that although the " e ight did imprO\ 'e co re qua lit y, drilling in fluid was e\"en more eflce ti\ 'e. The drill used on D evon Island incQt-porated a boos ter pump to assist raising the fluid and chips up the spira l flights between the inner a nd outer tubes.

Ice cores fromjust belo \\" close-oIT, esp ec ia ll y from high­accumula tion sites, a rc \'e l'y fragile. Crysta l bonding is no t ye t well de\'e loped, and fo rces due to air pressure in th e re­lative ly la rge bubbles a rc high. A possible explanati on fo r the appa rentl y greater fr ag i lit y of co res from high-acc umu­lation cores is that there ha s been less Lime for conso lidation and crysta l bonding compa red with low-acc umulati on sites where th e ice at depth is much older. Drilling resul ts in th e abrupt remO\'al of the overburden press ure ri ght at the point where the cutters a rc m a king line hori zonta l groon's in th e co re. Our experiments show that applying a load of onl y 4 kg (on a n 80 mm di am e ter core) by a free weight in th e drill-core ba rrel signili ca ntl y imprO\'es core quality witho ut the use of drilling fl u id. Th e i m prO\ement is surprising since this load is considerably less than the o\ 'C rburden pressure which, for compari son a t a depth of 100 111 , is equi\'a lent to a load of 350 kg on the co re cross-sec ti on.

Anlarctic GRG and Australian Anlarclic Dil,isiol1 , Bo\" 252-80, Hobarl, Tasmania 7001, Auslralia

27 AUgllSl1998

REFERENCES

VI X M O RGAN

ALA"! ELC I-I E IKH

R USSELL BRA:-I D

Bla ke, C. w .. C. P. " 'a ke anrll\l. D. G("ra s i mon~ 199RTh e E C I.IPSE d rill : " fiel d -port able illlcrmedi a te-dc pth ice-coring d r ill. .7 (;I(fciol" 44 11·(i ,.

175 178. G illn. F , D. Donlloll. C. G ira rrl, A. l\lanOll\T ier, c:. R aclo a nd C, R icou,

19R+. lee CO lT qua l it, · in c icnro-l11ccil a ll ical drilling'. CRRFJ ,I/m'. /1"/1. 8'f-3+, 73 RD.

"ar ita. H . 199,1. Status 01' sh a llow cl r ill. SpITi,, 1 scssio n report. ,\i,II. 111 ,,1. Po/a I' Res . . 111'111 .. ,S/w. 1\) 11 1' -f9, +O() +01.

Schwanclc f', J a nd H. Ru lli. 1')9,1·, E k nronH"l' il ani ca l cl rill in,'i; "I' a :lOO III

eorc in a d ry hole a l Sum l11i t, Gree n lancl" \il ll. 111.11. I'n/ar RI'" .1""11, . .\/11'(,

Iss lle +9. 93 98, Shoj i, 1-1 , 199 k Sta tus ol'core q ua lit y. Special sess ion report. ,\i,II. ftll /. I"i/ar

Res . .111'111" .I!m . lss({e 1,9. 10-1 ,1·05,

673 Downloaded from https://www.cambridge.org/core. 05 Feb 2022 at 07:10:42, subject to the Cambridge Core terms of use.