shell development company · according to the participation agreement for mps!, the confidentiality...
TRANSCRIPT
~
Shell Development Company A Division of Shell Oil Company
Bellaire Research Center P 0 Box 481
Houston Texas 77001
3737 Bellaire Boulevard
Houston Texas 77025
April 22 1987
Mr Dan Gibson Mr Terry Walden Amoco Production Company Sohio Petroleum Company Amoco Building 2 Lincoln Centre Denver CO 80202 5420 LBJ Freeway
Suite 800Lock Box 03 Dallas TX 75240
Mr Jim Poplin Mr Charles Smith Exxon Production Research Company U S Dept of the Interior P O Box 2189 Minerals Management Service Houston TX 77001 12203 Sunrise Valley Drive
Reston VA 22091
Dear Participants
MECHANICAL PROPERTIES OF SEA ICE - PHASE II (MPSI-2)
Enclosed is a draft copy of the paper entitled Uniaxial and Biaxial Compressive Strength of Ice Sampled from Arctic Multiyear Pressure Ridges by F U Hausler E N Earle and P Gerchow This paper is taken from the HSVA experimental study conducted as part of our MPSI-2 program and will be presented at POAC 87 The final report for the HSVA study was transmitted to you on January 24 1984
According to the Participation Agreement for MPS the confidentiality obligation for proprietary information extends for three years after the completion of a Phase Although work on Phase II continued at CRREL well after the completion of the HSVA study the CRREL work is an independent program and has no confidentiality restrictions Conseshyquently we view the January 1984 transmittal of the HSVA final report as satisfying the confidentiality obligation of the Participation Agreement The Agreement also requires that publications resulting from any Phase acknowledge all participants who wish to be acknowledged We
2 MPSI-2 Participants
will instruct the authors to acknowledge all MPSI-2 participants unless we are informed in writing by May 15 1987 that a participant does not wish to be acknowledged
E G Ward Manager Offshore Engineering Research Department
Enclosure
bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
001-713-663 2581 March 23 1987ANTO Fax No bullbullbull bullbullbullbullbullbullbull DatumDate bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull VONFROM ~ax No~ Hambur~ 69 20~ 34h ~PitPn7~h1Nn of pages bullbull J9bullbullbullbullbullbullbullbullbull
TELEFAX - MITTEILUNG I MESSAGE
Shell Development Company Houston TX USA lNTOt
Offshore Engineering Research Department
Mr EG W~rd zHdAttmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullmiddotmiddotmiddotmiddotmiddotmiddotbullmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
VONFROM HAMBURGISCHE SCHIFFBAU-VERSUCHSANSTALT GmbH Bramfe1der StraBe 164 D-2000 Hamburg 60 Tel No Hamburg I 6 92 03 - 0 Tlx No 2 174 236 hsva d FU Hausler ( _4middot2middot4middotgt bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 6 92 03
--~bullbullbullbull~-------bull------------------------------------vbullbullbullbull~--------bullbullbullbullbullbullbullbullbullbullbull~-BETRIFFTSUBJECT
POAC-87 Paper on Uniax1al and biaxial compressive strength tests
TEXT
Dear Mr Ward
Please find in the following the draft of the She11-HSVA joint paper on the above topic P1oaco t~ancfor a copy of middott to Or poundN ~arlo for approval If there are any changes you or Dr Earle recommendplease let us know as soon esmiddot possible bull The deadline for draft m~nuecrip~ ie Apri1 2 1007
Thank you in advance for you cooperation
Yours sincerely
~~middot~(p_J Franz Ulrich Hausler
middot
Uniaxial and b1axia1 compressive strength of 1ce sampled from Arctic mult1year pressure ridges
by FU HKusler EN Ealle
P Gerchow
Hamburg1sche Sch1ffbauMVersuchsanstalt GmbH fonnel1y) Shell Development Company Hamburg1sche Schiffbau-Versuchsanstalt GmbH
A Gcrics ol 60 comproooivo ct-n9th tec~c has been pbullrlollfted on iebull
sampled from mult1year pressure ridges near to Reindeer Is Beaufort Sea Three compressive stress states have been invbullstigated un1axia1 Ca1 bull O) and biaxial with oy bull ex and ay bull 05 exbull
()
i i I middotj
0
middot middot
Twn t mrwHbull111111PA - laquotnin rate combinations have been studied T1 bull -5 degC exbull 10middot5smiddot1 and Tlbull middot20 degC exbull 10middot3smiddot1bull
Fnr hnth T~~ombinatinn~ thbull coefficients of an isotropic 3-parametar yield criterion were evaluated representing the yield charaeterf stfcs in tho comp-ccion octant of the principal ~tlA~~ ~pbullcbull middotThbull multiyear pressure ridge f ce was considbullred to behave isotropic on a macroscopic scale
At T1 bull -z~ bullc tx bull 10-~s- 1 the tee exh1b1ted br1ttlt1 fre1~lun= UmSer
these conditions the measured compressive strengths were S times (un1shy~x1bull1) ~Pt~ ~o t11111111s (b1ax1ol oxbull oygt htyfnr Uuan 111tT1 bull bull5 ftc exbull 10 5 s 1bull where a mostly ductile mode of failure was observed
The mu1t1year pressure r1dge 1ce s~ua1eci exn1u1ta a w1d~ varle~y ur bull~c
types (snow ice columnar sea ice etc) Ice densities were found from2P1 bull ~u~ Kg m- up to ~1~ Kg m-2bull 1ne sa11ni~y measurea 1n ~ne samplesmiddot
mampltS ranged from St bull Q Up tO 57 Ooobull
middot i
Introduction
The Alackan and Canad1~n Arotio h~c
proven to be one of the world 1 s most imshyportant resources of hydrocarbons Offshyshore 011 and gas exploration and exshyploit~tinn in thi~ r~aion 1c impeded byfee of seasonally varying severity In order to achieve a long (at best yearround) drilling season and at the same time to prevent ships and structures from damage or loss both ships and structures must be designed to withstand ice loads Another important aspect of safe operation is to preserve the senshys i t1~ An L fl env I runment from avo1Cla1gt1e pollution
One of the most hazardous forms of ice loads is exerted by multibullyear pressure ~idgec which arc by 011 mean~ e common natural event in this region Besides the dr1v1ng forces (current wind) and the response character1st1cs of the 1ndividua1 ctruetu~e the foilurc of multi-year ridge 1ce yields an important1im1t1ng condition for the fee loads This cond1t1on may govern the structural design
Multi-year pressure ridges can be deshyf1ned as thick accumulations of broken ice blocks that have survived at least onA mPlt ~~~~nn Tho voids originallycontained fn the ridge from the pile up process are filled with refrozen water from surface melting Thus a multi-year oressure rf doe ntnrPiant( A huobull piocoof massive nearly vo1dless fee Its thickness may exceed 30 m (Kovacs 1976 and 1983 Cox et al bullbull 1984)
Infonnation on thA miarhAn1r~l prnportiasof multi-year ridge 1ce is scarce in literature except the data reported byKovacs (1983) and the results of the investigations by Cox et al (1984 and lgtOi) Tile laL~r dlso contarn tne only strength data available up to now for multi-~xi~l load conditions compressivestrengths under various lateral confineshymAnt4
In the present study plane stress faii shyure characteristics of multi-year ridge ice have been evaluated For this purshypose strenoth tests have baan rnnduetod under un1ax1a1 compression and under biaxial compression with two different load ratios (ll and 21) The influence
i I ~
l I
middot~
of strain rate and temperature was inshyvestigated by performing the test at two di~farcnt tcmpcratui-e ~trcin rote cumlJlshynations
Ice field sampling and shipping
The field sampling was performed byCRR[L perconnol between 3 and 15 April1981 in an area northwest of Reindeer Island 1n the Prudhoe Bay Alaska (seeFig l It was the same field campaign rluino which the camplbullc wo-o co11cctcd for Phase I of the comprehensive CRREL study on the mechan1ca1 properties of multi-year sea ice reported by Cox et al (1984) Their report gives detailed descriptions of the sampling and shipshyping procedures such that here on1y a short surranary seems necessary bull
Tho ~amplec wo~o collcotcd ~rom 10 multi-year ridges of various size The ridges were part of the fast ice belt and were apparently not grounded All ridges showed rounded outlines indicotshy1ng surface melt processes in their hi story
From each ridge four vertical cores were drilled using a 108 mm (425 in) f1berg1ass coring auger (Rand and Melshylor 1985) The cores were drilled in pairs (AMB and CmiddotD on two sites of each r1d~e Oependina on the ridaA~ lti1~ thQsites were from 14 up to 46 m apart Imshymed1ately after removal from the ice the cores were catalogued and packed in core tubes The tubes were then placed 10 1n~ul4tea sn1pp1ng boxes In the field no measures have been taken for sample refrigeration because the ambishyent a1r temperature of below -15 degC was e103e eruJu41Jh cu Ute NoC1bull t~o eutect1c-22 9 degc and because the mu1 timiddotyea-r ice had a low salinity (usually 4degoo)The ice was then shipped by airfreight to CRREL Hanover Nlf where it wcs stored at -30 degC (Cox et al 1984) On 2526 October 1982 a lot of 10 tubes with ice samples was shipped to Hamburg aaain bV airfrPiOht Oulino all airshyfreight shipping the ice was refrigershyated by dry ice In Hamburg the ice core tubes were unpacked from the insulated boxes and stored in a freezing box at
~ r~a ~aml- OG nc t ll I 1t microroclSS l ng
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
2 MPSI-2 Participants
will instruct the authors to acknowledge all MPSI-2 participants unless we are informed in writing by May 15 1987 that a participant does not wish to be acknowledged
E G Ward Manager Offshore Engineering Research Department
Enclosure
bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
001-713-663 2581 March 23 1987ANTO Fax No bullbullbull bullbullbullbullbullbullbull DatumDate bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull VONFROM ~ax No~ Hambur~ 69 20~ 34h ~PitPn7~h1Nn of pages bullbull J9bullbullbullbullbullbullbullbullbull
TELEFAX - MITTEILUNG I MESSAGE
Shell Development Company Houston TX USA lNTOt
Offshore Engineering Research Department
Mr EG W~rd zHdAttmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullmiddotmiddotmiddotmiddotmiddotmiddotbullmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
VONFROM HAMBURGISCHE SCHIFFBAU-VERSUCHSANSTALT GmbH Bramfe1der StraBe 164 D-2000 Hamburg 60 Tel No Hamburg I 6 92 03 - 0 Tlx No 2 174 236 hsva d FU Hausler ( _4middot2middot4middotgt bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 6 92 03
--~bullbullbullbull~-------bull------------------------------------vbullbullbullbull~--------bullbullbullbullbullbullbullbullbullbullbull~-BETRIFFTSUBJECT
POAC-87 Paper on Uniax1al and biaxial compressive strength tests
TEXT
Dear Mr Ward
Please find in the following the draft of the She11-HSVA joint paper on the above topic P1oaco t~ancfor a copy of middott to Or poundN ~arlo for approval If there are any changes you or Dr Earle recommendplease let us know as soon esmiddot possible bull The deadline for draft m~nuecrip~ ie Apri1 2 1007
Thank you in advance for you cooperation
Yours sincerely
~~middot~(p_J Franz Ulrich Hausler
middot
Uniaxial and b1axia1 compressive strength of 1ce sampled from Arctic mult1year pressure ridges
by FU HKusler EN Ealle
P Gerchow
Hamburg1sche Sch1ffbauMVersuchsanstalt GmbH fonnel1y) Shell Development Company Hamburg1sche Schiffbau-Versuchsanstalt GmbH
A Gcrics ol 60 comproooivo ct-n9th tec~c has been pbullrlollfted on iebull
sampled from mult1year pressure ridges near to Reindeer Is Beaufort Sea Three compressive stress states have been invbullstigated un1axia1 Ca1 bull O) and biaxial with oy bull ex and ay bull 05 exbull
()
i i I middotj
0
middot middot
Twn t mrwHbull111111PA - laquotnin rate combinations have been studied T1 bull -5 degC exbull 10middot5smiddot1 and Tlbull middot20 degC exbull 10middot3smiddot1bull
Fnr hnth T~~ombinatinn~ thbull coefficients of an isotropic 3-parametar yield criterion were evaluated representing the yield charaeterf stfcs in tho comp-ccion octant of the principal ~tlA~~ ~pbullcbull middotThbull multiyear pressure ridge f ce was considbullred to behave isotropic on a macroscopic scale
At T1 bull -z~ bullc tx bull 10-~s- 1 the tee exh1b1ted br1ttlt1 fre1~lun= UmSer
these conditions the measured compressive strengths were S times (un1shy~x1bull1) ~Pt~ ~o t11111111s (b1ax1ol oxbull oygt htyfnr Uuan 111tT1 bull bull5 ftc exbull 10 5 s 1bull where a mostly ductile mode of failure was observed
The mu1t1year pressure r1dge 1ce s~ua1eci exn1u1ta a w1d~ varle~y ur bull~c
types (snow ice columnar sea ice etc) Ice densities were found from2P1 bull ~u~ Kg m- up to ~1~ Kg m-2bull 1ne sa11ni~y measurea 1n ~ne samplesmiddot
mampltS ranged from St bull Q Up tO 57 Ooobull
middot i
Introduction
The Alackan and Canad1~n Arotio h~c
proven to be one of the world 1 s most imshyportant resources of hydrocarbons Offshyshore 011 and gas exploration and exshyploit~tinn in thi~ r~aion 1c impeded byfee of seasonally varying severity In order to achieve a long (at best yearround) drilling season and at the same time to prevent ships and structures from damage or loss both ships and structures must be designed to withstand ice loads Another important aspect of safe operation is to preserve the senshys i t1~ An L fl env I runment from avo1Cla1gt1e pollution
One of the most hazardous forms of ice loads is exerted by multibullyear pressure ~idgec which arc by 011 mean~ e common natural event in this region Besides the dr1v1ng forces (current wind) and the response character1st1cs of the 1ndividua1 ctruetu~e the foilurc of multi-year ridge 1ce yields an important1im1t1ng condition for the fee loads This cond1t1on may govern the structural design
Multi-year pressure ridges can be deshyf1ned as thick accumulations of broken ice blocks that have survived at least onA mPlt ~~~~nn Tho voids originallycontained fn the ridge from the pile up process are filled with refrozen water from surface melting Thus a multi-year oressure rf doe ntnrPiant( A huobull piocoof massive nearly vo1dless fee Its thickness may exceed 30 m (Kovacs 1976 and 1983 Cox et al bullbull 1984)
Infonnation on thA miarhAn1r~l prnportiasof multi-year ridge 1ce is scarce in literature except the data reported byKovacs (1983) and the results of the investigations by Cox et al (1984 and lgtOi) Tile laL~r dlso contarn tne only strength data available up to now for multi-~xi~l load conditions compressivestrengths under various lateral confineshymAnt4
In the present study plane stress faii shyure characteristics of multi-year ridge ice have been evaluated For this purshypose strenoth tests have baan rnnduetod under un1ax1a1 compression and under biaxial compression with two different load ratios (ll and 21) The influence
i I ~
l I
middot~
of strain rate and temperature was inshyvestigated by performing the test at two di~farcnt tcmpcratui-e ~trcin rote cumlJlshynations
Ice field sampling and shipping
The field sampling was performed byCRR[L perconnol between 3 and 15 April1981 in an area northwest of Reindeer Island 1n the Prudhoe Bay Alaska (seeFig l It was the same field campaign rluino which the camplbullc wo-o co11cctcd for Phase I of the comprehensive CRREL study on the mechan1ca1 properties of multi-year sea ice reported by Cox et al (1984) Their report gives detailed descriptions of the sampling and shipshyping procedures such that here on1y a short surranary seems necessary bull
Tho ~amplec wo~o collcotcd ~rom 10 multi-year ridges of various size The ridges were part of the fast ice belt and were apparently not grounded All ridges showed rounded outlines indicotshy1ng surface melt processes in their hi story
From each ridge four vertical cores were drilled using a 108 mm (425 in) f1berg1ass coring auger (Rand and Melshylor 1985) The cores were drilled in pairs (AMB and CmiddotD on two sites of each r1d~e Oependina on the ridaA~ lti1~ thQsites were from 14 up to 46 m apart Imshymed1ately after removal from the ice the cores were catalogued and packed in core tubes The tubes were then placed 10 1n~ul4tea sn1pp1ng boxes In the field no measures have been taken for sample refrigeration because the ambishyent a1r temperature of below -15 degC was e103e eruJu41Jh cu Ute NoC1bull t~o eutect1c-22 9 degc and because the mu1 timiddotyea-r ice had a low salinity (usually 4degoo)The ice was then shipped by airfreight to CRREL Hanover Nlf where it wcs stored at -30 degC (Cox et al 1984) On 2526 October 1982 a lot of 10 tubes with ice samples was shipped to Hamburg aaain bV airfrPiOht Oulino all airshyfreight shipping the ice was refrigershyated by dry ice In Hamburg the ice core tubes were unpacked from the insulated boxes and stored in a freezing box at
~ r~a ~aml- OG nc t ll I 1t microroclSS l ng
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
001-713-663 2581 March 23 1987ANTO Fax No bullbullbull bullbullbullbullbullbullbull DatumDate bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull VONFROM ~ax No~ Hambur~ 69 20~ 34h ~PitPn7~h1Nn of pages bullbull J9bullbullbullbullbullbullbullbullbull
TELEFAX - MITTEILUNG I MESSAGE
Shell Development Company Houston TX USA lNTOt
Offshore Engineering Research Department
Mr EG W~rd zHdAttmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullmiddotmiddotmiddotmiddotmiddotmiddotbullmiddotbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull
VONFROM HAMBURGISCHE SCHIFFBAU-VERSUCHSANSTALT GmbH Bramfe1der StraBe 164 D-2000 Hamburg 60 Tel No Hamburg I 6 92 03 - 0 Tlx No 2 174 236 hsva d FU Hausler ( _4middot2middot4middotgt bullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbullbull 6 92 03
--~bullbullbullbull~-------bull------------------------------------vbullbullbullbull~--------bullbullbullbullbullbullbullbullbullbullbull~-BETRIFFTSUBJECT
POAC-87 Paper on Uniax1al and biaxial compressive strength tests
TEXT
Dear Mr Ward
Please find in the following the draft of the She11-HSVA joint paper on the above topic P1oaco t~ancfor a copy of middott to Or poundN ~arlo for approval If there are any changes you or Dr Earle recommendplease let us know as soon esmiddot possible bull The deadline for draft m~nuecrip~ ie Apri1 2 1007
Thank you in advance for you cooperation
Yours sincerely
~~middot~(p_J Franz Ulrich Hausler
middot
Uniaxial and b1axia1 compressive strength of 1ce sampled from Arctic mult1year pressure ridges
by FU HKusler EN Ealle
P Gerchow
Hamburg1sche Sch1ffbauMVersuchsanstalt GmbH fonnel1y) Shell Development Company Hamburg1sche Schiffbau-Versuchsanstalt GmbH
A Gcrics ol 60 comproooivo ct-n9th tec~c has been pbullrlollfted on iebull
sampled from mult1year pressure ridges near to Reindeer Is Beaufort Sea Three compressive stress states have been invbullstigated un1axia1 Ca1 bull O) and biaxial with oy bull ex and ay bull 05 exbull
()
i i I middotj
0
middot middot
Twn t mrwHbull111111PA - laquotnin rate combinations have been studied T1 bull -5 degC exbull 10middot5smiddot1 and Tlbull middot20 degC exbull 10middot3smiddot1bull
Fnr hnth T~~ombinatinn~ thbull coefficients of an isotropic 3-parametar yield criterion were evaluated representing the yield charaeterf stfcs in tho comp-ccion octant of the principal ~tlA~~ ~pbullcbull middotThbull multiyear pressure ridge f ce was considbullred to behave isotropic on a macroscopic scale
At T1 bull -z~ bullc tx bull 10-~s- 1 the tee exh1b1ted br1ttlt1 fre1~lun= UmSer
these conditions the measured compressive strengths were S times (un1shy~x1bull1) ~Pt~ ~o t11111111s (b1ax1ol oxbull oygt htyfnr Uuan 111tT1 bull bull5 ftc exbull 10 5 s 1bull where a mostly ductile mode of failure was observed
The mu1t1year pressure r1dge 1ce s~ua1eci exn1u1ta a w1d~ varle~y ur bull~c
types (snow ice columnar sea ice etc) Ice densities were found from2P1 bull ~u~ Kg m- up to ~1~ Kg m-2bull 1ne sa11ni~y measurea 1n ~ne samplesmiddot
mampltS ranged from St bull Q Up tO 57 Ooobull
middot i
Introduction
The Alackan and Canad1~n Arotio h~c
proven to be one of the world 1 s most imshyportant resources of hydrocarbons Offshyshore 011 and gas exploration and exshyploit~tinn in thi~ r~aion 1c impeded byfee of seasonally varying severity In order to achieve a long (at best yearround) drilling season and at the same time to prevent ships and structures from damage or loss both ships and structures must be designed to withstand ice loads Another important aspect of safe operation is to preserve the senshys i t1~ An L fl env I runment from avo1Cla1gt1e pollution
One of the most hazardous forms of ice loads is exerted by multibullyear pressure ~idgec which arc by 011 mean~ e common natural event in this region Besides the dr1v1ng forces (current wind) and the response character1st1cs of the 1ndividua1 ctruetu~e the foilurc of multi-year ridge 1ce yields an important1im1t1ng condition for the fee loads This cond1t1on may govern the structural design
Multi-year pressure ridges can be deshyf1ned as thick accumulations of broken ice blocks that have survived at least onA mPlt ~~~~nn Tho voids originallycontained fn the ridge from the pile up process are filled with refrozen water from surface melting Thus a multi-year oressure rf doe ntnrPiant( A huobull piocoof massive nearly vo1dless fee Its thickness may exceed 30 m (Kovacs 1976 and 1983 Cox et al bullbull 1984)
Infonnation on thA miarhAn1r~l prnportiasof multi-year ridge 1ce is scarce in literature except the data reported byKovacs (1983) and the results of the investigations by Cox et al (1984 and lgtOi) Tile laL~r dlso contarn tne only strength data available up to now for multi-~xi~l load conditions compressivestrengths under various lateral confineshymAnt4
In the present study plane stress faii shyure characteristics of multi-year ridge ice have been evaluated For this purshypose strenoth tests have baan rnnduetod under un1ax1a1 compression and under biaxial compression with two different load ratios (ll and 21) The influence
i I ~
l I
middot~
of strain rate and temperature was inshyvestigated by performing the test at two di~farcnt tcmpcratui-e ~trcin rote cumlJlshynations
Ice field sampling and shipping
The field sampling was performed byCRR[L perconnol between 3 and 15 April1981 in an area northwest of Reindeer Island 1n the Prudhoe Bay Alaska (seeFig l It was the same field campaign rluino which the camplbullc wo-o co11cctcd for Phase I of the comprehensive CRREL study on the mechan1ca1 properties of multi-year sea ice reported by Cox et al (1984) Their report gives detailed descriptions of the sampling and shipshyping procedures such that here on1y a short surranary seems necessary bull
Tho ~amplec wo~o collcotcd ~rom 10 multi-year ridges of various size The ridges were part of the fast ice belt and were apparently not grounded All ridges showed rounded outlines indicotshy1ng surface melt processes in their hi story
From each ridge four vertical cores were drilled using a 108 mm (425 in) f1berg1ass coring auger (Rand and Melshylor 1985) The cores were drilled in pairs (AMB and CmiddotD on two sites of each r1d~e Oependina on the ridaA~ lti1~ thQsites were from 14 up to 46 m apart Imshymed1ately after removal from the ice the cores were catalogued and packed in core tubes The tubes were then placed 10 1n~ul4tea sn1pp1ng boxes In the field no measures have been taken for sample refrigeration because the ambishyent a1r temperature of below -15 degC was e103e eruJu41Jh cu Ute NoC1bull t~o eutect1c-22 9 degc and because the mu1 timiddotyea-r ice had a low salinity (usually 4degoo)The ice was then shipped by airfreight to CRREL Hanover Nlf where it wcs stored at -30 degC (Cox et al 1984) On 2526 October 1982 a lot of 10 tubes with ice samples was shipped to Hamburg aaain bV airfrPiOht Oulino all airshyfreight shipping the ice was refrigershyated by dry ice In Hamburg the ice core tubes were unpacked from the insulated boxes and stored in a freezing box at
~ r~a ~aml- OG nc t ll I 1t microroclSS l ng
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
middot
Uniaxial and b1axia1 compressive strength of 1ce sampled from Arctic mult1year pressure ridges
by FU HKusler EN Ealle
P Gerchow
Hamburg1sche Sch1ffbauMVersuchsanstalt GmbH fonnel1y) Shell Development Company Hamburg1sche Schiffbau-Versuchsanstalt GmbH
A Gcrics ol 60 comproooivo ct-n9th tec~c has been pbullrlollfted on iebull
sampled from mult1year pressure ridges near to Reindeer Is Beaufort Sea Three compressive stress states have been invbullstigated un1axia1 Ca1 bull O) and biaxial with oy bull ex and ay bull 05 exbull
()
i i I middotj
0
middot middot
Twn t mrwHbull111111PA - laquotnin rate combinations have been studied T1 bull -5 degC exbull 10middot5smiddot1 and Tlbull middot20 degC exbull 10middot3smiddot1bull
Fnr hnth T~~ombinatinn~ thbull coefficients of an isotropic 3-parametar yield criterion were evaluated representing the yield charaeterf stfcs in tho comp-ccion octant of the principal ~tlA~~ ~pbullcbull middotThbull multiyear pressure ridge f ce was considbullred to behave isotropic on a macroscopic scale
At T1 bull -z~ bullc tx bull 10-~s- 1 the tee exh1b1ted br1ttlt1 fre1~lun= UmSer
these conditions the measured compressive strengths were S times (un1shy~x1bull1) ~Pt~ ~o t11111111s (b1ax1ol oxbull oygt htyfnr Uuan 111tT1 bull bull5 ftc exbull 10 5 s 1bull where a mostly ductile mode of failure was observed
The mu1t1year pressure r1dge 1ce s~ua1eci exn1u1ta a w1d~ varle~y ur bull~c
types (snow ice columnar sea ice etc) Ice densities were found from2P1 bull ~u~ Kg m- up to ~1~ Kg m-2bull 1ne sa11ni~y measurea 1n ~ne samplesmiddot
mampltS ranged from St bull Q Up tO 57 Ooobull
middot i
Introduction
The Alackan and Canad1~n Arotio h~c
proven to be one of the world 1 s most imshyportant resources of hydrocarbons Offshyshore 011 and gas exploration and exshyploit~tinn in thi~ r~aion 1c impeded byfee of seasonally varying severity In order to achieve a long (at best yearround) drilling season and at the same time to prevent ships and structures from damage or loss both ships and structures must be designed to withstand ice loads Another important aspect of safe operation is to preserve the senshys i t1~ An L fl env I runment from avo1Cla1gt1e pollution
One of the most hazardous forms of ice loads is exerted by multibullyear pressure ~idgec which arc by 011 mean~ e common natural event in this region Besides the dr1v1ng forces (current wind) and the response character1st1cs of the 1ndividua1 ctruetu~e the foilurc of multi-year ridge 1ce yields an important1im1t1ng condition for the fee loads This cond1t1on may govern the structural design
Multi-year pressure ridges can be deshyf1ned as thick accumulations of broken ice blocks that have survived at least onA mPlt ~~~~nn Tho voids originallycontained fn the ridge from the pile up process are filled with refrozen water from surface melting Thus a multi-year oressure rf doe ntnrPiant( A huobull piocoof massive nearly vo1dless fee Its thickness may exceed 30 m (Kovacs 1976 and 1983 Cox et al bullbull 1984)
Infonnation on thA miarhAn1r~l prnportiasof multi-year ridge 1ce is scarce in literature except the data reported byKovacs (1983) and the results of the investigations by Cox et al (1984 and lgtOi) Tile laL~r dlso contarn tne only strength data available up to now for multi-~xi~l load conditions compressivestrengths under various lateral confineshymAnt4
In the present study plane stress faii shyure characteristics of multi-year ridge ice have been evaluated For this purshypose strenoth tests have baan rnnduetod under un1ax1a1 compression and under biaxial compression with two different load ratios (ll and 21) The influence
i I ~
l I
middot~
of strain rate and temperature was inshyvestigated by performing the test at two di~farcnt tcmpcratui-e ~trcin rote cumlJlshynations
Ice field sampling and shipping
The field sampling was performed byCRR[L perconnol between 3 and 15 April1981 in an area northwest of Reindeer Island 1n the Prudhoe Bay Alaska (seeFig l It was the same field campaign rluino which the camplbullc wo-o co11cctcd for Phase I of the comprehensive CRREL study on the mechan1ca1 properties of multi-year sea ice reported by Cox et al (1984) Their report gives detailed descriptions of the sampling and shipshyping procedures such that here on1y a short surranary seems necessary bull
Tho ~amplec wo~o collcotcd ~rom 10 multi-year ridges of various size The ridges were part of the fast ice belt and were apparently not grounded All ridges showed rounded outlines indicotshy1ng surface melt processes in their hi story
From each ridge four vertical cores were drilled using a 108 mm (425 in) f1berg1ass coring auger (Rand and Melshylor 1985) The cores were drilled in pairs (AMB and CmiddotD on two sites of each r1d~e Oependina on the ridaA~ lti1~ thQsites were from 14 up to 46 m apart Imshymed1ately after removal from the ice the cores were catalogued and packed in core tubes The tubes were then placed 10 1n~ul4tea sn1pp1ng boxes In the field no measures have been taken for sample refrigeration because the ambishyent a1r temperature of below -15 degC was e103e eruJu41Jh cu Ute NoC1bull t~o eutect1c-22 9 degc and because the mu1 timiddotyea-r ice had a low salinity (usually 4degoo)The ice was then shipped by airfreight to CRREL Hanover Nlf where it wcs stored at -30 degC (Cox et al 1984) On 2526 October 1982 a lot of 10 tubes with ice samples was shipped to Hamburg aaain bV airfrPiOht Oulino all airshyfreight shipping the ice was refrigershyated by dry ice In Hamburg the ice core tubes were unpacked from the insulated boxes and stored in a freezing box at
~ r~a ~aml- OG nc t ll I 1t microroclSS l ng
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
middot i
Introduction
The Alackan and Canad1~n Arotio h~c
proven to be one of the world 1 s most imshyportant resources of hydrocarbons Offshyshore 011 and gas exploration and exshyploit~tinn in thi~ r~aion 1c impeded byfee of seasonally varying severity In order to achieve a long (at best yearround) drilling season and at the same time to prevent ships and structures from damage or loss both ships and structures must be designed to withstand ice loads Another important aspect of safe operation is to preserve the senshys i t1~ An L fl env I runment from avo1Cla1gt1e pollution
One of the most hazardous forms of ice loads is exerted by multibullyear pressure ~idgec which arc by 011 mean~ e common natural event in this region Besides the dr1v1ng forces (current wind) and the response character1st1cs of the 1ndividua1 ctruetu~e the foilurc of multi-year ridge 1ce yields an important1im1t1ng condition for the fee loads This cond1t1on may govern the structural design
Multi-year pressure ridges can be deshyf1ned as thick accumulations of broken ice blocks that have survived at least onA mPlt ~~~~nn Tho voids originallycontained fn the ridge from the pile up process are filled with refrozen water from surface melting Thus a multi-year oressure rf doe ntnrPiant( A huobull piocoof massive nearly vo1dless fee Its thickness may exceed 30 m (Kovacs 1976 and 1983 Cox et al bullbull 1984)
Infonnation on thA miarhAn1r~l prnportiasof multi-year ridge 1ce is scarce in literature except the data reported byKovacs (1983) and the results of the investigations by Cox et al (1984 and lgtOi) Tile laL~r dlso contarn tne only strength data available up to now for multi-~xi~l load conditions compressivestrengths under various lateral confineshymAnt4
In the present study plane stress faii shyure characteristics of multi-year ridge ice have been evaluated For this purshypose strenoth tests have baan rnnduetod under un1ax1a1 compression and under biaxial compression with two different load ratios (ll and 21) The influence
i I ~
l I
middot~
of strain rate and temperature was inshyvestigated by performing the test at two di~farcnt tcmpcratui-e ~trcin rote cumlJlshynations
Ice field sampling and shipping
The field sampling was performed byCRR[L perconnol between 3 and 15 April1981 in an area northwest of Reindeer Island 1n the Prudhoe Bay Alaska (seeFig l It was the same field campaign rluino which the camplbullc wo-o co11cctcd for Phase I of the comprehensive CRREL study on the mechan1ca1 properties of multi-year sea ice reported by Cox et al (1984) Their report gives detailed descriptions of the sampling and shipshyping procedures such that here on1y a short surranary seems necessary bull
Tho ~amplec wo~o collcotcd ~rom 10 multi-year ridges of various size The ridges were part of the fast ice belt and were apparently not grounded All ridges showed rounded outlines indicotshy1ng surface melt processes in their hi story
From each ridge four vertical cores were drilled using a 108 mm (425 in) f1berg1ass coring auger (Rand and Melshylor 1985) The cores were drilled in pairs (AMB and CmiddotD on two sites of each r1d~e Oependina on the ridaA~ lti1~ thQsites were from 14 up to 46 m apart Imshymed1ately after removal from the ice the cores were catalogued and packed in core tubes The tubes were then placed 10 1n~ul4tea sn1pp1ng boxes In the field no measures have been taken for sample refrigeration because the ambishyent a1r temperature of below -15 degC was e103e eruJu41Jh cu Ute NoC1bull t~o eutect1c-22 9 degc and because the mu1 timiddotyea-r ice had a low salinity (usually 4degoo)The ice was then shipped by airfreight to CRREL Hanover Nlf where it wcs stored at -30 degC (Cox et al 1984) On 2526 October 1982 a lot of 10 tubes with ice samples was shipped to Hamburg aaain bV airfrPiOht Oulino all airshyfreight shipping the ice was refrigershyated by dry ice In Hamburg the ice core tubes were unpacked from the insulated boxes and stored in a freezing box at
~ r~a ~aml- OG nc t ll I 1t microroclSS l ng
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
i I ~
l I
middot~
of strain rate and temperature was inshyvestigated by performing the test at two di~farcnt tcmpcratui-e ~trcin rote cumlJlshynations
Ice field sampling and shipping
The field sampling was performed byCRR[L perconnol between 3 and 15 April1981 in an area northwest of Reindeer Island 1n the Prudhoe Bay Alaska (seeFig l It was the same field campaign rluino which the camplbullc wo-o co11cctcd for Phase I of the comprehensive CRREL study on the mechan1ca1 properties of multi-year sea ice reported by Cox et al (1984) Their report gives detailed descriptions of the sampling and shipshyping procedures such that here on1y a short surranary seems necessary bull
Tho ~amplec wo~o collcotcd ~rom 10 multi-year ridges of various size The ridges were part of the fast ice belt and were apparently not grounded All ridges showed rounded outlines indicotshy1ng surface melt processes in their hi story
From each ridge four vertical cores were drilled using a 108 mm (425 in) f1berg1ass coring auger (Rand and Melshylor 1985) The cores were drilled in pairs (AMB and CmiddotD on two sites of each r1d~e Oependina on the ridaA~ lti1~ thQsites were from 14 up to 46 m apart Imshymed1ately after removal from the ice the cores were catalogued and packed in core tubes The tubes were then placed 10 1n~ul4tea sn1pp1ng boxes In the field no measures have been taken for sample refrigeration because the ambishyent a1r temperature of below -15 degC was e103e eruJu41Jh cu Ute NoC1bull t~o eutect1c-22 9 degc and because the mu1 timiddotyea-r ice had a low salinity (usually 4degoo)The ice was then shipped by airfreight to CRREL Hanover Nlf where it wcs stored at -30 degC (Cox et al 1984) On 2526 October 1982 a lot of 10 tubes with ice samples was shipped to Hamburg aaain bV airfrPiOht Oulino all airshyfreight shipping the ice was refrigershyated by dry ice In Hamburg the ice core tubes were unpacked from the insulated boxes and stored in a freezing box at
~ r~a ~aml- OG nc t ll I 1t microroclSS l ng
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
iii
Test faci li t1 es
The strength test program was carried out on a triax1a1 closed-loop testing machine (Fig 2) The vert1ral axi~ of th1s device consists of a screw driven universal testing machine of 100 kN load capacity type Schenk-Trebel RME 100 The two horizontal axes are provided hy 3n HSVA-des1gned biaxial test1ng device w1th servo-hydraulfc actuators of more than 100 kN load capacity each The biaxial frame is mounted on the same base-rram11 as the RME 100 Thus the three axes form a monolithic unit Each of the three axes is individuallyelectronically closed-loop controlled Coypl i n9 be tween the QX~~ i ~ IJCU ru1amp~tJ
electronically For load application the testing machine is equipped with three pairs of brushlike loading platens (M1hdotf t 1066) C3pecially desi9oed ror ice compression tests (Hausler 1982)These loadfng platens consist of a clusbull ter of slender metal rods (bristles)arranged n a quadratic orroy They e~shyhib1 t a high rigidity in the load direcshytion combined with a minimal lateral constraint Their transverse compliance allowc the plotcn~ to folluw lo~cral ~~shyformation of the tested specimen Thus even true triaxial tests are possibleAt least the edges of the samples a1ways -om~n acceccbla to ot~oin mco~urin9
devices (L1nse 1975 H~us1er 1982 and 1986) The complete loading apparatus 1s installed in a cold chamber where any t~flHbullbulltu o bbulltwocn ~12 ond - 30 degc can be SPt Flectronic control and data acquishysition devices are installed outside the cold room For sample preparation a secshynnd cold room with an independent refrishygeration system of the same capacity was used It was equipped with a band saw and a lathe modified for milling cub1c ice samples
Data acquisition and recording
Loads were measured by means of HBM (Hottinger Baldwin) Z4 200 kN load Cells For deformation mbulla~uramentc MQMW2K LVDTs were applied which have a maxshyimum range of + 2 mm The primary axis deformation transducer was installed in a parallelogram guide (see Fig 3) wni 1e we other LVDTs were mounted on pairs of bristles on the loading platen sides Since the bristles follow the samples deformation they can be used as
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
p1ck up devices for deformation measurements
The analoguous signals from the transshyducers were on-11ne converted to digital on an HP 21 MXE series computer (HewlettPackard) and stored on floppy discs for later off-line data processing The d1middot git1z1ng rate was limited by the transshyfer rate from the AOmiddotconverter to the d1sc and by the storage capacity of the foppy disc In the present study rates of 100 cps and 50 cps were appliedin the high strain rate tests and of 4 cps in the low strain rate tests
All other data such as temperatures or sample dimensions were recorded manualshyly
Sample preparation and test procedure
Sample preparation was done by cutt1ngcubes of 698 + 01 mm s1de length from the 1ce cores7 This was done by first cutting the cores in cylindrical pieces of 10 cm length and then cutting these pieces 1n raw cubes of about 8 cm s1de lengthThis was performed on a band saw F1g 4 shows the orientation of an ice cube 1n the parental core The raw ice cubes were then milled down to the1r target dimensions on a modified lathe For sample dimension control a high precision stage w1th a d1al gauge of 001 nm resolution was employed In order to minimize brine drainage samplepreparation was performed at a temperashyture of about middot25 degC 1e below the NaCl bull2H o eutectic -229 degC in sea 2ice
The samples were then stored for one day at their target test temperature in orshyder to achieve a homogenous temperaturedistribution within the samples when tested
The work-off order of succession of the potent1al samples within the whole entity to be investigated was determined by drawing lots By this means a random distr1but1on was achieved in the paramshyeters not varied systematically such as eg sampling site (core number) vershytical position of a sample in its parenshytal core or salinity
Prior to each individual test the samples d1mens1ons were measured again
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
and 1ts weight was determined Then the ambient air temperature and the ice temshyperature were recorded The ice temperashyture was determined 1ns1de a reference ice cube of the same dimensions which had undergone the same temperature hisshytory as the sample to be strengthtested
In the next step the sample was instalshyled 1n the loading device The primarycontrol axis actuator wasmiddotdr1ven to a compressfve preload of about F ~ 025 kN (corresponding stress 005 MPa) and kept these under stat1c force control
In the biaxial tests the same was done wf th the secondary axf s but w1th a preshyload corresponding to the target~load rat1o The secondary axf s was then set to dynamf c force control using the primary axis force signal as a dynamicsetting means From here the load rat1o between the two axes was kept constant
In the consequent steps of the procemiddot dure the tips of the bristles used for strain measurements were frozen to the sample by a drop of water Then the parshyallelogram guided deformation transducshyer to be employed as actual value transducer for strain control was puton top of the specimen parallel to the primary axis Its tips were attached to two opposite br1stles such that deforshymation was measured between the loading platens This allowed the continuation of a test even past sample fracture The primary axf s was then switched to closed-loop strain control
During the test the primary ax1s strain was controlled by a dynamic setshyting means such that the raquostrain-rate was constant Test stop condition was the attainment of the target ~strain
After the test the sample or its debris was melted for sal1nity determinat1on
Data analysis
For each sampl1ng cycle the measured forces were converted into stresses using the equation
middot
( l)
with o the stress and F1 direct1on and A1 the fn1tial cross secshythe load in ishy
prrMor1
fl j ~OI
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
t1on area normal to the i-d1rection
Simultaneously the deformations u in i-d1rection were transformed into strains c using the simp11fied equation
1
t z u Ci (i bull 123) (2) 1
with cj the initial distance between the pick-up points of the u -deformation 1 transducer (basis length)
The initial tangent modulus was read from the slope of the a over 1 pound plots 1 Because of the irregular distribution of ice types and crystal orientations withshyin the ridge ice mass the ice was asshysumed to behave 1sotrop1c on macro scale Upon this assumption the ice failure cond1tion can be described bythe isotropic three-parameter yieldfunction (Smith 1974 Reinicke 1977)
2 fa1j bull aJ + 6Ji + yJ - l =0(3) 1 1 where a 11s the stress tensor and J1 its first 1nvariant while Ji is the second invariant of th~J stress dev1ator s1j
s1j bull aij - t 6ij akk (41)
(42
(43)
For a given temperature-strain rate condition the average strength values for the 3 different plane stress load conditions were computed These averagedfailure stress states were then used to determine the coeffic1ents of the yield function (5)
f(aij)bulla(a11+a22+b(a~1+a~2)+ca11a22middotl which is the plane stress case of the three parameter yield function eq 3)
Results
The averaged test conditions temperashyture strain rate density and salinity are compiled in Table 1
The average strengths (stresses at yield or failure) and initial tangent moduli are listed in Table 2
x
The following charactarist1cs have been
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
~middot
observed for the six test series bull -5 middotlSeries 1000 (o1=cr 2bull11 TI 2 -5degCe1=10 s )
The stesses t1me histories exhib1ted a sharp rise unti1 about half the yield stress The yield stress was reached after a period of strain hardening at strains ~Y usually less than ii After y1e1d tfte stresses decreased slightlyAt strains around 4 stresses increased again and reached or slightly exceeded the initial yield level Dur1ng some test runs a reasonable amount of brine was squeezed out of the specimen and formed icycles at the unloaded bottom surface Cracks if any 1 were of minor sfze and were in plane with the loads applied (Fig 5 1012) middot
Series 2000 (o1a2bull21 r1bullbulls0 cpound1z10-5s-1)
The typical stress t1me history showed as in series 1000 a sharp rise up to approximately half the yield stress followed by a period of strain harjenshying Yield was reached at strains c1 of usually less than lS Typically the stress decreased continuously after yield but apparently tended towards an asymptotic value to be encountered beshyyond the end of the tests (pound gt 51)Cracks were of minor size an~ were in plane with the loads s1m11ar to the pattern observed 1n series 1000
Series 5000 (o1a2=lO T1s-5degce1=10-5s-1)
The specimens ex1b1ted an almost ductile mode of failure In general no cracks have been observed Deformation perpenshydicular to the load was large Yield was r~ached at an average primary strain of pound1 = 07~ Series 3000 (o a bull11 zt0-1 1 T bull-20degC~ 3 s- 1 )2 1 1 The samples typical1y failed before havshying ~eached a zero slope in the stress over time (or strain respectively) curve The failure mode was usuallybrittle In some cases the first failure was not catastrophic but was followed by a second stress rise which then was term1nated by final failure Only one single sample broke after a period of strain softening The failure stress (first stress maximum) was typicallyreached at strains of c~bull 025 The samples were usually crusted during the
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
tests The major surfaces of failure wera slightly inclined to the pl~ne expanded by the loads In addition most of the samples were f1lled w1th microshycracks (Fig 6 3013)
bull 3 -1)Series 4 00o (afi2~21 TI~-20centc e 1~10 s
Typica11y brittle failure was observed in the strain soften1ng part of stress history The average ~rimary strafn at yield was poundY =032 The crack pattern was simi1ar1 to the one observed in the series 3000 tests
1series 6000 (a1o2bulllOt r1--20degcpound110-3s- gt
All samples failed more or less brittlemiddot Ty After the tests the samples were filled with cracks or totally crushed In four (out of ten) cases fracture occurred during or at the end of the stress rise Fracture or yield was obshyserved Yat an average primary strain of only = 021 The governing crack t 1pattern exh1bited crack surfaces orienshyted parallel to the load but without orientation 1n the two perpendiculard1rections In some cases parts of the samples failed 1n add1t1on by bucklingThe m1crocrack density was smaller than observed in the corresponding tests with b1ax1a1 loed application Ser1es 3000 and 4000)
The plane stress yield functions (eq 5)for the two temperature-strain rate comshybnations have been determined as follows
a) T1t-51degC pound1bull101 lO~Ss-1 f(aij=l78 MPa~(a11+a22 l
+178 MPa (az +az ) -183 MPa-2a1 i22~~=0 b) oc -3 1T1m-20l ~ 1=099bull10 s
f(a j)~00476 MPa- 1 (cz +a~2 )1 1+00262 MPa~~(a11+a22 )
The str~ng~~6 ~~~~1ti11iai- 1 p~esentedgraphically in Figs 7 and 8 Themiddotupper left halves show the individually measshyured strengths while in the lower righthalves the average values and 99 confidence interval bars are presentedAlso shown for comparison are the reshysults obtained at CRREL for the same ice under uniaxial tension and compressionparallel to the parental core axis (Cox
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
t ~-
(
I
I I
~ i
middotmiddot
et al 1984 and 1985)
Discussion
A compar1son of the data from the preshysent study with the results obtained at CRREL for the same ice (Cox et a1 1984 and 1985) 1s ev1dent
The discrepancy between the uniaxial compressive strengths seems obv1ous Cox
~~os4~Pa ~~8~~er~~~rg1~1deg~o~~i~1oi ~~~3sYmiddot 5~on~it~~~9 (~~~sfo~t~~~ f~4~c~ 026 MPa and -726 + 203 MPa) The main reason for this dffference is the fact that the assumed isotropy of multi-yearridge ice is not fulf11led In their Phase II test series Cox et al (1985)have performed a comparison between horizontal and vert1cal compr4syestrengths at a strain rate of 10 s The vertical was 131 times (-5 degC) and l65 times (-20 degC) as high as the horizontal strength The values of Cox et al (1984) used above for comparison were obtained from vertical tests (parallel to the cores axis) while the strengths of the present study represent horizontal results
Approximate hor1zonta1 compressivestrengths can be calculated by correctshying the vertica1 strengths with the horizontal to4 vrtical ratios obtained from the clO~ s- test Th1s correction yields -179 ~083 MPa for the a51M~a ~~1~~e ~2oe0~~omiddotS1~ae-~~e~t~~=~ ly These values are in reasonable agreement with the results of the preshysent study
The coefficients of the yield funct1ons evaluated strictly speaking cannot represent more than the plane stress failure characteristics under compresshysion Extrapolation to plane-stressesincluding tension 1s theoreticallypossible but 1s usually 1nappropriate due to different failure mechanisms under tension and under compression
Concerning the initial tangent moduli
~~~~d ~e-2~a~ean~middotf~-s~14 is6~P~~~~d by the corresponding modulus 762 +-119 GPa measured by Cox et al (1984T The value 087 plusmn 049 GPa measured in th~
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
-5present study for the -5 degC and 10 s condition appears to ba too low and is assumed to be due to larger initial setting of the brush platens in this case
Comparison of the f3l~le strains show for the -20 degC10 s case almost 1dent1ca1 values 019 + 004S (Cox et al) and 021 (this study15 T~e largerdifference in the -5 degCl0 s case of 038 + 0171 (Cox et al) compared to 071 (this study) may again be explained by the larger initial setting of the of the brush platens
The density and salinity values given in Table l f1t well w1th the data reportgby Cox et al (1985) (JS 891 + 26 kg m and S z 126 + 082 - (Phase I only)
Conclusions
The present study provides horizontal iRd wflolfan1lr compressive strengths of multi-year pressure ridge ice The unishyaxial strengths exhibit good agreementwith the corresponding results obtained by Cox et al (1984) for the same ice This is remarkable since Cox et al bull have used 254 mm (10 in) 1ong nearly cylindrical specimens of 1016 1m1 (4in) diameter with bonded synthane endcaps (Mel1or et al bullbull 1984) while in the present study cub1c samples of 689 mm s1de length and brush11ke loading plashytens have been employed The good agreeshyment achieved in spite of different testing techniques supports the credishybi 1ity of both data sets and allows their combination for subsequent analyshyses
The combined data sets provide an amplepicture of the mechanical properties of ridge ice under the conditions investi shygated Nevertheless as already recogshynized by Cox et al (1985) the broad structural variations encountered in natural pressure ridges as well as the 1ack of knowledge on near to melting conditions demand additional research work Acknowledgements
This paper is a joint contribution of the Hamburg Ship Model Basin (HSVA) and Shell Development Company The support
shy
j
1 i l J f i
middot1 bull
of th US Army Cold Regfons Research
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
i 1
l
1 j
and Engineering Laboratory (CRREL)namely Mrs JA Richter-Menge given in sample shipping 1s gratefully acknowshyledged Thanks are also expressed to the members of the staff of the HSVA ice engineering department who contributed to the present study in particular to Mr w Naper who was in charge of samplepreparation and technical assistance during test performance
Referenees
Cox GFN et al uMechan1cal propertfes of multi-year sea 1ce ~ Phase 1 Test results US Army Cold RegionsResearch and Engineering Laboratory Hanover NH Report 84-9 April 1984
Cox GF N et al bullbull 1Mechanfcal propertfes of multi-year sea ice - Phase II Test results CRREL Report 85-16 October 1985
H~usler FU middot11Multfaxial CompressiveStrength Tests on Saline Ice with BrushType Loading Platens IAHR Symposium on Ice Proceedings Vol II pp 526-539 Quebec Canada 1982
Hausler FU Mult1ax1a1 mechanical properties of urea doped fee IAHR Symposium on lee Proceedings Vol I pp 349-363 Iowa City USA 1986 Hflsdorf ff 11 Bestitmtung der zweiachshysigen Festigkeit des Betons Deutscher Ausschu6 fUr Stahlbeton Heft 173 Berl in 1965
Kovacs A Grounded ice in the fast ice zone along the Beaufort Sea coast of Alaska CRREL Report 76-32 September1976
Kovacs A Characteristics of mu1t1-year pressure ridges POAC Proceedings Vol 3 pp 173-182 He1s1nki F1nland 1983
Linse obull Losung versuch~technischer Fragen bei der Ermittlung de~ Fest1gkeits- und Verfonnungsverhaltens von Seton unter dreiachsiger Beanspruchshyung mi t ersten Versuchen Technische Un1vers1tat MUnchen 1975
Mellor M Cox GFN Bosworth H Mechanical properties of mu1t1-year sea ice Testing Techniques CRREL Report84-8 1984
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
flSVAshy~potf-l=1~middot
ff114)
31
Jlf
Figures
Fig 1
F1g 2
Fig 3
Fig 4
Fig 5
Ff g 6
Ffg 7
F1g 8
Smith MB 11 A parabolic yieldcondition for anisotrop1c rocks and soils PhD Thesis Rice UniversityHouston TX USA 1974
Re1n1cke KMbullbull Plasticity Analysiswfth an Isotropic Three-Parametric Yield Function Unpublished report BEB Gewerkschaften Br1gf tta und ElwerathHannover Germany 1977
Rand J and Mellor M Ice-coring augers for shallow depth samplingCRREL Report 85-21 1985
Map of ice sampling area (after Cox et al bull 1984)
Trfaxial closed-loop testing machine
Deformat1on transducer with parallelogram guided LVDT attached to a pair of opposite bristles of the brushlike loading platens at the end of a biaxial test (Series 1000)
Or1entation of a cubic sample in its parental core
Sam~le 1012 after bh_31a_~ compression test with01bulla2bull 03bull0 at w5 degC and 10 s
Sample 3013 after bllSjCt~l compression test witha1bull02bull abullO at -20 degC and 10 s
P1ane_secttryenss strengths of multi-year ridge ice at -5 degC and 10 s
Plane ~r~f s strengths of multi-year ridge ice at -20 degC and 10 s
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
f
Series Temperature
T1 [degC]
Strain rate at yie]f ei [ s ]
Stress ratio 0 1middot (72
Dens1ty
[~~ m-3]
Salinity s
( 0 oo]
1000
2000
3000
4000
5000
6000
500l l 00 plusmn 0 00bull l0-5
-5l plusmn 01 100 plusmn 001bull10-s
-202 plusmn 01 100 plusmn 009bull10-3
-202 plusmn 01 098 plusmn 009bull10-3
-52 I 01 101 ~ OOObulll0-5
-199 t 01 099 plusmn 011middot10-5
11
2 1
11
21
1 0
10
868 plusmn 36
889 plusmn 15
909 plusmn 4
890 plusmn 23
873 plusmn 31
886 17
18 t lC
13 t 06
l3 plusmn 06
19 plusmn 07
17 plusmn 13
25 plusmn 16
1000 2000 5000
middot51 plusmn 01 101 plusmn OOlbulll0-5 all 880 plusmn 27 16 t 10
1 i I middotI
j I l
3000 4000 6000
Table 1
099 plusmn 009bull10-3-20l plusmn oz
I
Test conditions
all 895 plusmn 19 19 t 12
j
1l Ibulli
Series Primary st~ss at yield 1[MPa]
Secondary ~ress at yield 2[MPa)
Init1al tangent modulus E [GPa]
Number o tests
1000
2000
3000
4000
5000
6000
-2 30 plusmn 0 90
-2 33 0 23
-1337 plusmn l87
-1 l 98 plusmn 2 bull l 2
-140 plusmn 026
-7 16 plusmn 2 03
-230 plusmn
-112 t
-1332 t
-598 t
091
011
186
104
087 plusmn
633 t
049
240
10
10
10
10
10
10
Table 2 Average strengths and elastic properties
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
r ~middot
middotmiddot-middotmiddot-----~____~----middot~-- -~-----------------_______
-4 shy
BEAUFORT SEA
~ ~STUDY AR~~~ N
I
0
ij-~
REINDEER~ ARGO I
NIAKUK IS
bull
Figzf ~on~ple ~tion ~fter ~ 111~2)
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
~AXIS OF PAREMTAl CORE
PARENTAL CORE (1067 MM DIAMETER)
CUBIC SAMPLE (698 MM SIDE LENGTH)
TEj SYSTEM OF SAMPLE
j
3
COOROINA
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
0
I e I
~
1 TEMPERATURE T = -51 degCa1
(MPAJ STRAIN RATE Ibull 101 middot10-5 s-1
O DENSITY 9 bull 880 KG M-3 ----4gt-alD4gta------1--shy
SALINITY S= 16 degloo
-1
-2
-3
-4
i
-s
MEASURED STRENGTH
AVERAGE STRENGTH WITH 99 deglo-CONFIDENCE INTERVAL THIS STUDY HOR COX ETmiddot AL (1984 1985) VERT
-5 -4 -3 -2 -1 0 1 CMPAla2
I
bull -middot~ ~bull _____ bullbulln - 1 bullbull bullbull bullbull middot-middotmiddot~ middot~middot bullrbull middotmiddot middot~middot ~~tbullbullmiddotmiddot ~middot _
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-
4
01 2 (MPAJ
0
-2
-4
-6
-8
-10
-12
-14
-16
-18
-20
TEMPERATURE T= -201 degC STRAIN RATE t =099middot10-3 s-1
DENSITY 9 =895 KG M-3
SALINITY s 19 ooo --ltJKD--O--i()ICKJHgt-----a1-----+-bull-
0 STRENGTH
o AVERAGE STRENG TH WITH 990o-CONFIDENCE INTERVAL
J e I THIS STUDY HORmiddot~
COX ET AL (1984 1985) VERT ( middot~
-18 -16 -14 -12 -10 -8 -6 -4 -2 0 2 a2 [MPAl
---middotmiddot- - -middotmiddotmiddotmiddot-middot middot- ~
- -f~-1
- Shell Development company
- Mechanical properties of sea ice
-