HRVARD SOIL MECHICS SERIES o. 88
LIQUEFATION AND YLI DEFOMATION OF SANDS A CRITICAL REVIEW
b
Arthur Casagrande
Presented at Fifth Panamerican onference on Soil Mechanics and oundation Engineering Bueno Aires, Argentina, November
1975
Pierce Hall Cambridge Masachusetts January
976
Reprned (with correctons)
anuar
19
IUFCTION YCLIC DEFOTO OF SS
A CRITICL RIEW rhr Caarande Profeor eri Haard Unieriy Caridg
M US
SURY i paper reiew ineigaion o wo baically differen phenoena or which in l i er a re he ae er LIUEFCTIO i ed To prn onf ion he e CTU LIUEFIO i ed in hi paper for h repone of conracie looe and ha lad o lo of rngh ich can cae flow lide and he e CYCLIC LIUECTIO or he repon of dilaie dene and whn beced o cyclic laboraoy e Topic coe rd: Hypo hei of cri ical oid ra io and i early e ing Liqefa c ion lide in or eck Da and hypohe i of Ineigaion of aca liqeacion wih load conrol flow rcre riaxial e definiion of dilaie and conraci one lin and l iqefa c ion po en ial Ineia ion of repone o cyc lic loading in ario ype of cyclic laboraoy conclion ha cyclic liqefacion in e pecin i caed by rdiribin of waer conen which i eneraed by echani ha noaly ar n in in i ha cyclic liqacion oally canno deelop in i deneand and Becae i i nliely ha laboraory e can be deied o eliinae he r re radien in peci n and o reprod ce h nifo ree ha ei in a ypical eleen in i he ahor belie ha h gra ap bwe abo oy and in i repone o cyclic loadin will rqir coprheni ield ineigaion of fll cal e ha nare i proing in igly eiic region or eiin he in i cyclic repone of di dene and dene and an inri procdre i ggeed ing cclic riaxial e on anioropically conolidad pecien
CTENTS
RDUCTION
I
DFINITIONS OF "ACTUAL LIQUFACTION" D "CYCLIC LIQU CTIN AS USED IN TIS PAPER
III
LAOATORY INE STIGATIONS O F ACTUL LI QUFACTION D POTESIS OF CRITICAL OID ATIO (19598)
2
IV
ISTIGATION OF PTIAL FAILUR OF FORT PECK D AND POSIS OF FLOW SRUCTU
V
U LIQUEFACTION PRODUC IN TRIIL TSTS WIT MONOIC LAD CONTROL
V
CONTS ON POTENTIL FOR ACTUL LIQUEFACTION
0
6
VI
CCLIC TRIIL TSTS PROFESSORS SEED L
2
VI I
IS TIGATIONS WI GTR S APPATUS
4
COARISON OF CYCLIC STSSS INDUCED IN STU AND IN OTORY TESTS
X
CRIS ON OF RSPONS TO CCLIC LOADING OF DENS SDS IN SU AND IN OATOR TESTS
20
XI·
TORY TSTS FOR DES IN P UPOSE S - INTERIM RECONDATIONS
23
XI
S R OF PRINC IP CONCLUS IONS ND CONDATIONS
24
XIII XIV
2
ACOWLEDGNTS FRNCES
2
LST OF FIGUS
YPTHESIS OF CRITICAL OID TIO EIN Y NS F DIRECT SHEAR TESTS
2
C OSS SECTION TOUG FL SLID IN FORT PECK D AT STATION 2+0
CAISON OF TR ISOTRPLLY CONSOLIDAT R TSTS S T ST U S ING DE-LOAD INCNTS
4
SSS CIRCLES OF ISOTROPCLY CONSOLIDAT R TEST S "A" i Fi 3 5)
CARISON OF T TPS OF R TESTS WIT ACTUAL LIQUFACTION
6
CAL OD TIO F LI N) FM CTS OR CCLIC LOADING
7
CAL OID TIO c LIN) FROM R TSTS USI STI L LADING
i TESTS
USN D-L
8
VRIBLES CONTROLLING POTEI FOR ACTUAL LIQUEFACTION OF ING SAND Baed Dta in Ref )
9
LIEFACTION IN LOOSE SD JACENT TO A WATERFONT
1
STS OF TYPICL CYCLIC TRI TEST ON DENSE S 10
1
ISRIBUION OF RELTI DSIT IN CYCLIC TRIL MN (Ref. )
2
GS ILLUSTTING MCNICS OF a) TORY SHR AD RECIPROCATING SHE PRODUCED BY OTATING OF GYRATORY S PTUS
13
SCHTIC SECTION OF GYTORY APPTUS LE HF SLIDING
4
T FOR GYRATORY TESTS, RIG HALF FOR RECIPROCTING TESTS SCTIC PL OF SLIDING PTES FOR GRATORY ND FOR IPROTING TESTS
1
CRISON OF TYPICL ISTRIBUTION OF RETIE DENSIT I IPROCTING GTORY TESTS
16
ISTGS OF RTIE DENSI DISTRIBUTION IN SPECIMNS AS RPD AND ER IOUS NUERS OF CCLES IN CIPROCTING STS
7
LETE RECORD OF REDISTRIBION IN MDI LS SPECIMN 2 RCIPROCTING CYCLES
18
RCIPROCTING TEST ON DENSE BNDING SAD
9
ICE PORE PRESSUS HORIZONAL DISPCNS S ER CLES IN GYTORY TEST ON MDILOOSE BING SD
20
CLETE RECORD OF RISTRIBUTION IN TORY TEST ON I
2
S BDING SD SRESSES IN ELNT IN SITU BENEATH HORIZONTAL GROUD F SUBJECTED TO CCLIC SHE STSSES
22
SESSES IN NT IN SITU WHICH IS CONSOLIDAT UNE STRICL STRESSES CAUSED BY OERLYING LO WITH I SHR FORCES SUPERPOS
23
STSSES IN ISOROICLY CONSOLIDATED SPECIMN SUBJCTED TO IC TRIIL TEST
24
IC STSSES ON THE SURFACES OF A SPECN SUJECTED O IPRTING S TEST N SULTING DISTRIBUTION
2
loding the s te ben to e sed for spcfc response of nd in cclc trixil tss (Ref l In n ffort o prevent cn fio b he se of e sa erm for entirly differnt pnomn infol coie of severl colls nd self red n 1969 t fin noter te for e cyclic rsponse phenon n lbortor ts Finll we reed on the ter "cyclc mobilit" Heer b thn the ter liefction w lredy so well entrenced n lrtre fo e it te cclic trii tet tt it proed imrctl o ere to ts gren; nd e confson contned Wle I wold ch prefer not to e he er "lqefcton" for penomenon tt tr i no lefcton I decded relctntly o se t for both henoen in ths pper bt o dfferentite between tem wh orite djcives s folls: 1 C QAC wht ws smpl clled "efcion" eore the deelopmnt of cyclic lborory ests It s the ponse f loose strted snd when sjected to strins or socks tht elts in sbtntl loss of strenth and in extre cs leds to fl sldes
2 CYCC QAC e response of test specmen o ditve snd to cyclic lodin in tril test wen te pea pore rere rses nril in ech ccle to the confinn pressre
A strong inority on tt 1969 coee spported te erm "trin softenng" Sperfci lly the proressve sofnn wic evelos n cyclic tss on dense snd my resele strin softenin Te trin softenin however is prodced by stress cycles wh eentl nform dsrbton of sresses wtn the mter t the softenn of strtd snd specn drin lodng is ced by redstrbton of the wter conten whcyclc sbsntl oosenn nd softnn of certn zones in e specin wle oter ne re bein cocted
III Y LOTORY INVESTIGATIONS O ACTUL LIQUEACTION SIS O CRITICL VOI TIO (1935l938) rt b obering the ol cnes of dense nd looe snd rect sher tests nd in prt by nttion I deeloped n 935 te hothei tt when loose sd is shered it decrese n ole it orct nd entll proche ted ste vol (or oid io dity) nd tht t he s snd in dene tt ncrese in i lte til it l reches te s sedy tte i the loose snd Ti ste I clled he "criticl " or "crcl oid rtio" In m first lectr on thi b i Neer 935 (Ref 1) I w till nder th iression tht iticl deit i ndeendent o the nol str B lter hen erforing tet ing greter rnge of no I nclde t the crtcl deit fcton of th tre Bc irec er tet did not perit fficienl oion l chges I desie n 193 for the s ngne ii rt for instigion of h i o ri n d or th Frnn Fl D; 2
anation of the concept of critica oid ratio i facilitated by efein to the relationhip as deried fro direct hea tets, Fi bu pottin the nol tr on lo cale, wil be ud i ssequent figue In i. (b) the vertical cale i the oi a r reatie density, and the horizontal cale the dipacent. e h the contractive repone of looe d; cre ' th ate repone of dene and; d the horizonta ine ' repreents citical void atio or critica denity which both ape would e if heain coud be continued fa enouh, or if the shea test e starte at that void ratio. In ig. la) are plotted on he etca scae the applied hear tress ad horizontaly aai the ispacents. Cue L shows a tead incree of displacmnt se sd with inceainBut hear ti eentuay an utite senth s eached for tre the dene d cure deeop rt a pea strenth d then, with further hearing and futher ncease, the resitace drp and th m utimate hear trngth i eaed as fo the ooe sand Cure represent a tet stated e criica oid rati o that in thi et pecin theoretica n o che shoud deveop, indicatd b line ' in Fi. ( . Fi. (c are potted the oid ratio (or denity) on the etica scae, and hoizontaly on a o scale the applied no te n. F reference, cure L c, c a c repreent consoidaio e a obtaine in onedimensiona conolidation tests The diect hear tet plotted in igs. la) and (b) are asued t e perormed ude a confining preure of one kg/q cm. Therefore, points x d z on the conolidation cure in i. (c are the stain point of the direct hear tet. By projecting thee point izontaly onto the vertical scae of diagr (b) , we fnd the
x zy'repreent tatin points and z' ofthe thecritical void ratio dipacement s point z and voidvratio for this cur. and en conoidated under a normal tre of one kg/q I epeat the tets at higher norma trese, he startin o atios, including the critica void ratio, woud all be ower. in seera eries of uch tet undr different nor presure, e fin the critical void ratio line E, the heay line in diagra (c ich i santiall teepe th the oid ratiopresue cue c i tt r of oid ratio. tn te critica void ratio line, idea the tst ho e peod on sturated pecins without alowin chane of a h pore pesure au rents; bt h teting eqi pnt a n t een developed 0 years ao erefore te quetion happen when atuated sd i ujected to hearing at conta a seed indirectly, foows Suppose the sand i ited ratio wod a normal preue point , . c. Ifthe thevoid pecin be heared with repesented volu chane e, the oi ratio would decrease verticaly dward, at constat ti ts, ti the critica id ratio ine E i reache at t H if no olu chge were perted, the san wod il t e i ol; t bcaue this i no psile t p ferin tess from the grain stucte to t po horionta at conta vo til we reac the
cal vod rt io l ine a po in q e int il effet stres of aou 4 kg/sq t o int wld t be de to alms oneh f he init illy ppl id ts te e wter wud r a he enre ppl ied stress S lge reun in effetve res orespon ing sa streg aes t lqefat ion e he and ges fm a soli te i in a s w fls k a vous fl i en sjete t ll er strese B eeath a level g sfe, nt loded by stres, a l iqf ie o fl Te onl of te lquef ie stte wol be ols a he srfae produe b pwd ranage f xess wae Hve a rre erlan b l iqef ie snd wld be i er ole Fro Fg l we c also nle th satae sd e he cral v i rao lne wl e fe aga inst tl lac ion; tha s sd s expose to ser foes, i wuld delop nega ive pe pesse, e, te effe ive integlr sese woud e inrsed a te sn w brae itself, to a, ganst flure ne ig its streng ve t wh ee befoe the sear tresses were ppled Bu av iatn wld l h nrease in streng to ne tspee neg ive pore psre e hptes is f r ital vo i i te l rest le to te fllng nl is: All onat ions f vd at d effet ive l strs (1) ch are ld belw, r t the left, of te it il l ie E" F ig l, psent ttes w i t wold eelp ilt ie rspone n re safe ga inst [al l iqeft in (2
All p int t te g r ve te it ial le
"E"
wd repesent nt ions wld eslt ose I er t e fl lta ie, nd nt lyinntat l fiel i ite dnss, te strt ig nt wol ae to be sbtnt ill t te rgh f te ital i at i l ine s tt in te l iqef ie tte the effet ive stsses wl rp t a smll fat io f tse t eed at te start f sear
e greter te efet ive nf ining pssre, eg, te greater the dept f a s stat, the ler i te rt ial o i ao or, in oter wods, t enser be e s t be safe agans [atal] l iquefat ion B wen eav il lade, een e s be sep ile to [atl] lqfat ion Eeall these were vs when I preented he eond lee on lqefat in Je 38; Rf 3 e nts lte, in eee 38 a maor l iqfaon falr evelped in te F Pek d n Mnta w w the narng t le in
5
V.
ESTIGATIOS OF PIL FAILURE OF RT OF OW STRUURE
PK D
YOTSIS
I Fig. 2 is plotted a typical cross sectio throuh the slide, th tre scale i the ler sectio d with the vertical cal five ies elarged i the upper setio. It ca be ee that the mass that ed s alst lvel after the lide. he major mvemet occurred i ot 3 mite. So portios of the upstream toe moved to a maximum de of about 400 meters with the sped of a rapidly movi river. ot illio cubic ters of sads i the dam d i the foudatio rti iated i this slide. Corp o Eieers carried out a major ivestiaio (Ref. 5 7) ich icluded the takig of lare diater diturbed sample the sds by freezi colus of sad i situ ad the corg them h Calyx drills. From the results of may triaxial tests, uig hpothesis d testi procedure, it ws coluded that these sads od ot have liquefied. Probably iflueced by these test results, jrity of the cosulti board, which was appoitd to ivestigate t slide, cocuded that the failre as ot c aued by liquefactio; f 5) A miority, icludi self, cocludd from th peed of the moet ad the toporaphy of the slide mases aftr the moveet, t sd at cosidrale depth must have liqufid. I as was orced to the colusio that sad located bel the critical void rati li "E", as defied i Fi. 1 ca alo liquefy othr ords that the typ of test, which I used for detmiig the criical void ratio (Ref. , did ot defie correctly the boudary bte sads tht re safe d usafe aaist actual liquefacto. the years folli tee Ft. ad Peck s damliqueid ide I developd gdually the hypothesi tat ad i actually fli, it must hav a strcur e diffret from that whe the ad is satic; that duri flow eac ra is cost atly rotati i relatio o ll surrodi rais so as o offer a miim of frictioal resce . I termed ti flow tructur". I postulated that sh a structure 1 sprads by a chai reactio, exists oly drig fl, d (3 that i the mot fl stops, he rais re rge themselves ad rever t ito a stati sructure which, after the ces ater has draied, ill be slghtly deser tha the stac tctre befoe liquefactio occurred hus, a ss of sad that as lieed, ould ed up wit icresed safety agaist actual lique io I a ypical alluvial sad with little or silt, the staic tre before ad after liefactio d also flow structure ar of the igle-graied type. A sad fill costructed by duig d i ist state ca have a hoeyco sructure bcause the pir force betw the moist rais resist the sd frm lig ito sigle-raied strcture such "bulked" sad o strated, i is particlrly proeWe to liquefy; but fteards ed up ith a sigle-r aied static strcture. We the ture f a sesiie cla collapses i a chai rei, liqu o d fl side ill lt. e srcia srucur f th ever, i irrvrsibly dstroyed. Expriece h sh tht eclidtio ch s is o lg ssii o li
6
e following observation will assist in visualiing the chracter of the fl stuctue of sand W obseving hydraulic fil plcement of gavel I once obseved a boulde that dischaged fom the h ralic ipe line and apeaed to be larger than the diamete of that pipe Obvioly thee was one osition in which it could avel hough e pipe povided that the mass was flowing fast enough to maintain the ole in a osition without ouching the wall of he ie and ad j cent pebbles. ried several tis to achieve in th laborator a fl rcue but failed I beg to susect that the fl structure can evelop d exist on l within a large mass Then, out 11 years ao adate studen from Chile, Gonalo Castro asked to suggest a doctoral eseach topic I ecided to make one moe attet and I ecibed to him this roblem, wih the waing that evious attempts ha failed. I ntioned that one reason be that during flow in nate the driving foces are maintained essentiall constant, ie ta natue is aplying dead load diving forces not stain contol ha tied Load contol ensres that whe a fl structue is inning to develop, the driving force does not elax as in strain conrol B keeping the driving stess contant, the ss acceleates faster and faste and more d more of the sand grains will rearange eelves into a flow structure. Casto w willing o detake thi difficult task. Now I will explain his incial findings which ae conained in his dissertation; (Ref. 8 d as sulemented late b additional tests V
U LIQUEFAION PROUCED IN TRIAXI TSTS MONOTONIC OAD CONTROL
asto perford a majo ortion of his investigation on "Banding Sa", a ifom, clean uart sand wih su oded to sgular rin, a coefficient of uniformity of 1. 8 and with about 1 b weh smaller tha .1 n i 3 ae plotted the stressstain cuves of three tical R consolidateddained) riaal tests in wich the aial load wa incease with deadload incrents. In addition, is shown for corison the stressstrain cuve of an S test (drained test. In a or tests the specimen was first isotropically consolidated under drostati ressure of kg/sq m Then the aial load was in creed with small deadload incents applied at out oneminute nervals on a hanger. n ts "A" the specin had relative density fter consoli ion of Drc 3% Afte 15 mnutes of incrental axial loading, tresstrain iagram Fi 3) ofeced a eak en at a din stain ato stess ofcurve about (pper 2 kg/sq one ercent r te net smll loa incent the ecin sdenl liuefie in a s fraction of a send strained o ot 25% when he opped automiclly t out % strain the effective or tes ha droed to £ .3 /q c a then it re ined ontt t this vle rin frthe fl In the ler
7
igr cue "A" shows how the pore pressure rose a reache urig th stte of flow the costat valu o 3. 85 kg/sq wth the correspig effective ior pricipal stress 0. 15 kg/s q c r the relatioship si � f/(2 3f + f the agl o tl frictio of this loose s specie ca be cput o b 30 grs Tis agrees with the results of the S test carrie out t he s relative esity a which is sho by the ashe stressstri curve i the upper igra by te correspig stress circls i Fig. 4 for both tests However it y b coiciece th the stresses urig the sate of flow reflecte the sae frictio S test frictio agle of relatve 30 egrees as i the the effective I a iclie believe a ul esity agle itote flow tht strucure i sller th i the static structure This is also supporte by the observtio tht i Fig 4 th effctive stress circle t th peak whe liquefacio starte reais bel the 30 egree stregth lie.
e relative agitue of the various stregth circles i Fig 4 is goo illustratio for the eoous ifferece i stregth which the s specie of saturate very loose sa ca isplay after it has bee csoliate uer a hyrostatic pressure of 4 kg/sq c. I test i. e whe pore pressures are ot pete to evelop the or pricipal stress ca be icrease to 12 kg/sq i. e the stregth (eviator stress t failure) will be 8 kg/sq c If o rai ge is peit te urig aial loa icrease th specie evelop s a pa stegt of oly 2 kg/s c with uc pore pressures of 2 5 g/sq c. But this peak stregth reprts the thrshol of the chage to a flow structure which is particulrly proe evelop whe th applie loa will ot relax with cotie stri i ote wors whe a ea loa is restig o the specie Fig 4 a also urv "A" i the upper iagra i Fig show tat the chage to the flow structure reuces the peak stregth of 2 kg/s c to a stregth urig flow of 0. 3 kg/sq c I the liuefie stae te sa i this test ha a apparet frictio agle of oly about two egrees. Theefore urig a flow slie a ss ofsuch sa woul sprea out util its surface woul slope oly a fw egrees Fig 3 test speci B" has a relativ esity of 44 at the start o axial loaig The upper iagra shows that it evelope a pea stregth soewhat grea ter th the loose speci A the it liqueie straie to 18 i 0 4 seco a the all o veet stoppe T ffect aitioal striig the eviator stress ha to be icrease agai with ea1a icreets I this test urig flow the stregth ha ecrease oly little as copare to the peak stregth evrheless this specie actually liquefie. The iuce pore pressures i this test are sh i the ler iagra which together with th effective eviator sress (stregth) urig flow i the upper igr reflects first cceleratio a the eceleraio with the speci "freeig" at 18% sri Fro the short stay stat segt lw oveet oe c copute th ratio f/ (f + 23f 053 si � fr which � 32 egres This specie was substatially ser h specie "A" a reter efective frictio agl urig l is rasoable I this tes th efecive ior pricipal stress
8
deread to aut on-fourth of th srss r hih w nsolidat isotropically Thi arg cr is edent frm te fac tat uring ta of fl h a intaind Should n scrib t rnse B "mt" atual lquation?
h cimn nt idial y pk strngth of spcin
e hrd t, "C" in Fg 3 wa carrd ou at 4% rlat den aer onoliaion, i, ighly dser than in tst "B" e rrain ue apard to vl int a ak at a slighly rear diar srss t tst B", th a suggsin of ienng a qacton Hr t hn h strtran r rrsed t aur d th s rngth incrd raily a rsut f a rsne, prssu b sen r by the inced stn n the dlati lr dagr The or to zropor at areurs srn of I th tst had bn coninu ngate pr rure d ha dvelod If th tet in th ur dagr had bn re a he s reat nity the intersction o curs and "D d ha ocurrd at ut 14% strain, hn the pr preur dred o r But th test wa mad on a losr in or a dener sn cur "D oud b hir nd the intrsti uld are ter i h 14% strin crrspnng o ero pore prsur n 5 ( in Rf ar card thr on loe semns ith rlativ nsitis ou atr cnslaion der a hydrostati prsur of 4 kg/sq c Tst A" th st A in Fig 3 d 4. h lading rocurs in the othr o tsts we dierent, dcrbd blw en ! irst iorpically conlidated ith a mnr
na o 4 tr kg/q s for e o tts a ajorstres prinCia of (th 8 g/q wihothr a principal tr ratio o 20 Oly sall adlad inrnt r apld n t simen thn it liqfi at abou 05� train hi e dentrats hat loos nd hic is ustible t atual qatio, ill re aily liquy ajacnt step sl h eeath a orial surac en inta coliat orialy 4 g/sq d e sjtd to yc loading ith a iatr rs g/q e first four cycls devl alt elt
on, wth t hteri op patcally cinn, the ing l in 5, d th strain a sl rctin one ern Heer, durng the 5th the ra need a roded p deopd at u ne perent trn d e se aay lqeed
deee ea at durng he ae fl he tde o the rdul trnh f d the rdu et D rna re were not l ontnt durn , u ee denia in re After a f tre eed, the d had ls al or o re s trin rere, t i re t U th the structur dea t r e
ali
I 6 are potted (1 the startng condtons d he cotons during fow of a tests erfod by Castro in whih acua lquefacton deveoped (Ref 8 d tests perfoed sseuenty e verca scaes are vod rato and reati ve ensiy; and the orthmc horonta scae is the effective mnor prncip stress e arrs pointing to te eft ndicte the sart conditns; and te poted point to the eft of each arr s te eective minor prnpa stress uring the stte of fow The are, heavy circes re et on isotropica11y consodated specns whch deveoped quefacton with rge drop in strength The sma crces re tess o sotropca11y consoidated specns during which ter onl revey s mal drop strength d uring f; the ries re et tes aisotropa1y consodated the suares repreen wth cycic oadng e specins horzont d dstance frm the trg arr to he effectve stress in the uefed stte is the reduon in the effectve stres tha deveope by actua quefacton or ee, n this pot the three tests with the oosest specimens a reative densty sighty greater thn 0% One test was ooated uder 4 kgsq e ad durng f ts effectve minor prncp stress dropped to 00 kg/sq e, a reducton by a factor of 20 o ther tests at the sa void rato were consodate under 0.3 kq e an durng ctua quefcton they aso deveoped about .02 kg/sq m effectve mnor prncpa stress, a reducton by a fator o ony 15 Fg 6 shs that no matter to what minor prncipa stress a specn was consodated initay d no matter ether t was consoidated sotropca11y or anisotropca11y or hether t was cycicay oaded, the conditons durng faure a ended up on a fairy accuratey defned ne whch I now ca the le the etter F stadng for crtca vod rat n whch ueaon wth a fw trctr deveoped en I aske astro to perform tests wth tr ctr oadg d the resuts are potted n Fig 7 with the arrs on the riht agan ndicatin the startng condtons The crcuar points, whch represent he condtons during fow, e cose to ne wich is ispaced rom the P ine (g 6 by a factor of out 5 n terms o te effectve mnor prncpa stress I w refer to ths ne as e le e reon for the are difference n the strength dur wen using oad contro ad strain contro oadng is the re o stran The constt drivng force o a dead oad produced n ese tests a rate of stran out 0000 times faster than that n e tests with stra contro oadng The reativey s rn i the ater tests caes ocay groups o sand grins to oe eorary their f structure I suspect that one coud eve a de re o c lines wth wdey dferng rates of tr coro e I ke astro t perfo triaa S (draned tests sm1ar o ose I used 40 yers ao, d whch were ued o investigate the queion poteti o he ands in connection with the invest on ter the lde n t eck D Wth a fe eeptin, e reul ell onsder aove the ineie even urther r e ne, r te ie n d the
10
ctered widely. I retroect, it i clear y the ivetios of the esitivity of the sad i d blow th Ft. Peck D elde d stregth bout ti greater t trixial R draied) ig dead loads. Te reso is ily that a flow trutu�e develop i S tests. VI
S ON POTI FR AUA IQEFAI
ferrig to Fig. 8, where thead F lies are reroduced from . 6 d 7 it see reasoable to exres the liuefactio pl, L, by a raio relatig the iitial effective mior pricipl he F lie.stress I to e the effective thi ratiomor eual ricial to zr tres alog theFf o le I e the exresio L (i E. g. , for the iitial oditios at oit Mi, the coditios actual liuefactio ar fd by drig a horizotal lie to its itersectio M f o the F le, rreodig to a liuefactio otetial Lp 1 26 r y startig coiatio o the F lie, the liuefactio otetial d be zero; ad below te F lie it would be egative, i. e. , whe he sd is straied bel the F lie without allig volu chage, egtive ore ressures ad additioal tregth would be mobilized. the area bel the F lie, which I desigate the "dilative zoe", or the "D" zoe, the sd for which this artcular F lie was determed cot actually liuefy. ove the uer lie with the short hes, the Eu lie, this sd will alway develo a ditictly cotrtiv resose wh straied; ad i that zoe liuefactio is liely 8 a reult of y raid straii g or hocks. Betwe e the E u the F lie eiter dilative or cotractiv resos is oible deedig o te rate of strai, d alo deedig o ow cloe to the F lie the tartig oit i locaed. Muc remi to be leaed out the reoe of sad ad the characteritic of the flow ructu withi that zoe. E. i trai cotrol tet it i pssile that fl tructure will aear d diaear i mall poets or lees; that as soo it develo i oe ocet the res i that zoe will dro by redistributio d the tructure i h pocet may revert to a ol structure while i oter oct m chge ito a flow structure. The overall resitace of a le urig strai cotrol would the be average resitace of a ly chgig atte of distributio of flow structure withi h th stregth durig ctual liuefactio drops oly slightly l he e streth, e. g. , i test "B" i Fg. 3 the lique oetial rges betwee about 2 d , i. e. , the starti p would be locted i Fi. 8 aroximately alog the Esc le. pte f h drop i por resure by a factor f 3 to 4, th th rp oly litl l i f ri iterest for egieer ig plicati . h m reset le ad udt, I cosider th tire the F li to be a hard with repect to ctual lquefactio ot o d ort structures i hily sesc
es Bel the F lin e atual liqu fati n is nt pss ible , althu sl l s tains a y develp the F line in Fi g 8 we an quikly estite the stength f Assig a fiti angle f is patiula sd in liquefied state duing fl the effetive 30 degees and a elative density f pinipal s tess wu ld be aut 0 1 kg/s q and e deviat stegth aut kg/sq e Heve at a elative desity f the n efetive pinipal stess duig fl wud be abut kg/sq 3 gs d this wuld equie a deviat stngth f abut 6 cespnding t an effetive j pinpal stess f abut gs q heefe , in liqueied s t at thee i s an e us dfeene i the f al the pu sa sand, at densities f e d F stength all pati pses thiselative s and with a elativ desty f culd t develp atual liquefati Is it eivale that a ass f let this quesi sd lated well belw t he F line , s ay at p int A in Fig 8 whee s safe agains t atual liquef at i uld epa nd dil ate by ul auses t pit A 1 well abve the F line whee it wuld th e e p tibl e t atual liqu fa tin ? Let e des ibe a atual phen wi i the upe Aps is n by the nae asially lage ses f athe dese gula tal will liquefy d fl d a valley aig gea t des tu ti It is well d cuented i lteatue that just pi t suh avalanche the u nt ai b k w h ee ges f the te the talus dep si t , s tps flwing he aive ppulati in tai valleys have k and heeded suh aing si gna ls f etuies I nside it p ss ib le that a cina i f he avy aifa ll sn el t pe haps ined wi th a thi k zne f till ad elatively ipevius talusass adjaent t thelage slp, ill isefzen the gunate levl in the talus d eate sh ea st ess es whih au se st n g dilati n ie , a substatial inase i vlu whih sas up lage vlus f wate Suh anges uld ceivaly hange the psitin f lage sses adjaent t the tal s lpe f b elw th e F lie t wel l ave the F line e haps the fibility f suh a hais culd be investigated at elatively all e en se by e s f lat y tes ts iqacti slides sties eate the iessi f a suessi e a s lides Ass a ive b ak i lse sand, Fig a d that elent "A" s distae f the slpe is suject t a safe cipal stss ati but that a eleent B" nea the slpe is ject t a uh geate piipal stess ati whih ndes it e ele t liq uef ati his ied wi th p gess ive s teepeing f he slpe by esi a c ause a li ite d zne t liquefy as ta te d i i g a) the ha thed wedge f s ad fls u the ipal ati will i n adjacet zne ic i stess additi, als thiee seepagequikly fes will be geate weinv b a int th e ive b a Fig b I ths ae liq u- cti ca pgess baad d a lage vl sd y fl i ve leavig behid a s lpe su fae wi th a ve fla t g le tha t ets te lo segt the liquefied sd Heve, sll t el the d sufae d ext t te fae f th slp ie
1
t ones not shaed in Fig . c , the san is so lightly loadd that i t l ie bl the cri t ical F line w here it can not li que fy . O the o s ur face obs erver y see only a progr s si on of shar crac s , oveying the iresion that the entire movent consists of a cs i o of ordin ary sh ear s lid es , wher es in fa c t the major p or ti o as is mvig li a isco mas wi thout sear surfac s . Such fl lies usually start along a short stretch of the ban. lique fction progress es ldwar , the a re a widens suc h that in pl the fia area appear lie a fl with a arr zone through which the if ie s has fled ou t . VI I I T I TE ST
SSRS SEED D LEE
BY
n fially , af ter so my ye ars , be lived that wi th as tros help I ha solved the problm of liquefaction of san in a satis fator manne rofessors Seed an Lee plished their well-nown p aper ntitled LAC SAUD S DUG CCC ADG; f . la ro fes or See an h is i cip les had the already carrie out iortat pioneer wor on the response of earth masses and trc tue to ear thquae . To app ly their theories they nee de s oi l arters that express the response of sad to earthquae loading. I t w logi ca to expe rime nt wi th cyclic t ri axial t s ts . There by th ey dicovere that when a triaxial specimen of saturatd san i subjected to cclic loading in undrained test i such a manner that the ricipal stresses pass in each cycle through a hydrostatic state of res wh ich mea ns that all shear s tress es dis appe ar , evn a dense higly dilative sand will develop cyclically high pore pressure efoations . A tyical test result from the paper by Seed an Le is reproduce i Fi g 10 . Th s an ecim en was p aed at a rel ativ de ns ity of 8%. In the upper diagr are plotted verically the axial strains and oint all y the ner of cycl es on a log scale . Th hori zon t al line i the middle represets zero s tra n . The horiz ontal lines oe the ero lie ar fo r 5 , 10 d 15 co ress ive s trains ; d blow the zero ine for 5 , 10 15 exten s io. In the l ower diag ram are p lot te ricaly the inuced poe pressurs and horizotally again the nuer ye o a log scale . The ini ti al hydr os ta tic conf ining pres sure oe g/s cm. The a xial eviator stress wa s pl /minus 0 . g /sq e • the ve r tical prin cip al tress was cycl ed betw een 1 . an g e, while the lateral stress rema ind cn t t at 1 . 0 gsq ing the first 10 cycle the aial strains remained negligible, the por pressurs increase decreased with eac cycle, in higer higher vaue til after about 12 cycles the pore ure reace the confinin pressure of one gsq c in the nt a princip tresse equal all strese in he cin ero . he nwre the por e pr s su re shear equalled th conf ining ue for the firt time See d Le dfine thi "iitial fction it aditinl cycles the xial strain ept inraig e i th uppr i agr. ilure wa efine i n te of the r of ccl we the trai litue in o cplet cycl e a cert it c b ee i n Fig . 10 th r
13
pessue ope sutantia th scimen iae a tong iative eone wheneve the inca te tio oe to a mim But in ite of an enta ta atte of ccic oe esue nca e n eceae th ccic eak tains of t e pcin ke t inceaing Fo exame at 2 cce th e to ta tain itue wa 1 out eua ivie been coession a eteni on om suc tests, ee an Lee w the fo ing concusion: nitia iuefaction be ouce b ccic oaing aso in i ense ense sans n i not imite to ooe as 2. e iquefaction otentia eceae with inceaing con fining pesue
he iquefaction otentia ecease wit inceaing initia 3 pincia stes atio an theefoe n ajace nt to a soe is ess suscetibe to iuefaction t beneath a hoionta uface t tt ti I i not aie tat te te "iquefaction in thei pae ha a entie iffeent aning tan the phenomena that I a awa associate with that tem Theefo e thee thee concions aeae to contaict a ast eeie nce n ao the esuts of Cato' invetiation Since I cou not fin an cue in thei as fo thee tating concuions I ha no choice but to out uc ccic tet obseve caefu evething that hening in thee tet I esigne a sime aangement that co be attace to Castos tiai a aaatu to ap ccic ato oas ofefo sinuoia howe ae thatAea the tothe of fist a ecimn few ccic was etting tets oft which that obvious a aica eitibution of the wate content eveope within the ecimen; tat gaua the top of the pecimen fo b ateate necking an bugin, that otimes even a in e of wate aae on the top of the pecimen in the mont n a hea tesses beca eo a the oe eue euae e confining esue We mue this eitibution of the wate ctent b feeng te st ecimen This uie that the secimen econsoiate une th confining ue b oening the vave eeb a cetain t of wte fom the soft top potion wa ne off an then the ame cou be move a foen of the ue eistibution of the wate content, eese i t of eative ensit, i pesente in i 11 fo Ref 8 ing the tet, the aveae eative enit of the ecimen wa 7% t e econoiate at the en of the tet it incease to ag of 78 he iffeence was ue to wate aine fom the ft p potion of the pecimen , theefoe, the actu eativ e sit of the top mt av been conieab smae tha 0 e e eativ nit in the e potion ha incease to 80 to 90% i efect te ubsttia inceae in ensit of the e f of t pecimen Thi aca eitibution of wat content f coeponing eative enit is te coine eut of the ffct t oa effect whi caue etition a in
1
o toic lly l oded sp ei es w as me ured le dy my yers go t t e Wtey s Ep erime t St t io; Ref . 11 . (2) the high pore resure d softeig which deelop the mot the spec s le throug h the s t te of hy dros t tic s tress; d (3) id of ig ctio which see to dr wter to the to tht is freed i The th ts whe cycli g throu gh the hydros t tic st te of s tress ortt questio whether such redistrbutio i test specimes is reresettie for wht hppes i elet i situ will be dis e lter
etiled iestigtio of the redistributio o the wter o i cyclic triil tests i which redistributio is produce eroitio of seerl differet chis wold be difficult e rig . erefore I decided to tc le first th e redis tributo ro cd by oe m echis oly ly th t p roduce d by oifo tress dis tribut io o the specime b odr ies or thi purpo se I cocetrted o the deig of gyrtory pprtus i which we could ero gyrtory well cycli c direct sher t es ts . The desi g I h i id would permt freeig the specime t the ed of thetest the prt Not e At th t time I c ould o t thi of reltiely simple desig for freeig the test specime iside trxi l ce ll . I th e while I h e de e loped desi gn which ful i lls h i reu ire met . I he s m t te i t to the Wteys Experi t St tio icsbu r g Miss i th e he tht it will st imulte detiled iestigtio of the reditributio i cyclic trixil tests VIII
IESTIATINS WITH RY SHE TUS
Iestig tio o f redis trib utio i re ciproct ig she r d gyr tor sher tes ts would he t he fol lowig dt ges: 1 It would ermit iestigtig the bodry effects more clerly tha i cyclic trixil tests i which it woul be difficult to sort out the ifluece o ( bodry ef fects b o the pumpig ctio d c of cycl ig (2 it would be esier to through the hydrosttic stte of stress; freee the tes t spe c ime i the ppr tus nd wi thout the eed of re olidtio; d (3 he redis tributio could probbly be ies ti g with greter ccurcy. rgill I iteded to cofie the specimes i wire-reiforced ber memres of th tye deeloped by the Noegi Geotechicl It itute . But I soo relied t h t it wou ld be ecess to destroy r f or ec tes t whe cut tig the fro e speim d tht u ct tig wol d be d iff icult . The it oc cur red to tht we could ort the rber mre by tpe of flt coil sprig which is 8 t b th of "li y" . Thi proed excel let lti o . W coe red the urf ce of the fl t co ils w th te fl o to uc frctio; bt prcipll we relie o slight sertio of t oil by teorril isertig spreders" before the strt o hig r r to p p r tu I de ie d w i th th hel p o tr o
B o or rerch effort I eded other itelliget l. ( i wt I clle m doctorl c did tes oly hl-joi g ly .
The
5 ar in ac a mde kid o saver because hey slave fr sveral ar fr h chrished h D al ho h ey ay n o h ae read a nl bok abou philoop y ) Ju s ime ud suale cand a akl i Red , a ic ude who as wel ndowd endur ance a quali par cularly im pr a or his proj ec a h may chages in desig ad i sig ehique ha a d eopd by ra an d rro r
I d scrbe e a ara us nl y brif y I d gn d o r cl ic dre shear e s d for gyra shear es s T n o e gyra oy es s is i lus r aed i n h e op o F A nan rznal frc s appied o he op f e spcie b a ar ngar prig moued on a roaig a Fr brev I es" na is yp e In he lr diagram o i son h mhans f h cyclc dc shear es fr I n rdc d h na me "re cir aing shear es " r b ref l "X Fr b y o es h a sys em is u se d r appng rzonal forc exc a n he reciprcai ess i ra r d a rr aing rce f nu s da shap e 3 ow a s ema rs s e n o appara T ran a sh on e lef mned a pair Negar sprigs ranm a nsa rzna rce o e o he semen, ri zoal fr e rem an cns an regard ess o ma n ud r za sp lacm en The b shig and pis n are m ne n a dng pae whc r g yr ay s , so on h ef d ds bew ee gyra o a l bear in gs ; and r e repr an sar s sh wn n e rh s d , m g b ack d or een 3 se s baris s i de o ad bo om ud barigs e eimn i ecsed n a rbr membrane whch in u p r ed by h "s linky" s pr ig Whn p acig he s pecime, h c ls e fa spring ar kep separad by coi sraders wic ar r u or a es s ared Thi sures ha h s lnk n carry any ercal oad urng h es
I n decri sch deal as e lerca dsaceen d re prsr ransdcer eqpm fr ecrdg h upu of ransduc rs , h ype m r used f r ro ag h a rm e n and eqme r aran e spcien pror h e and an pran dals rlad h echque frezng e and u ig h em no 64 gm n A hse dea ar ad 13. rd reerenc per dagr n i 4 a pan e e grary d n ae a d d la cen randrs , ne e d r a n a n dir c on hr zn a for c and nd a 90 degre a en ung fo r calb ra n s on rbe r pc ien he d lc in a alw ay r rc io f h e ran a H er f or and p ien he vcor lagd ehn direcion of a a nad nl A e na rcion in h a a a nd i i n an al l bra in . er dar n
6 v o f t h e reci p 0atn and v L atera
g sld
plate
th
s
h' buo h F i S w f d v e ops i n such t e . ts alteri g l y s o ncoored J ) ! f . A t e t
o these t w � nd � a T v : se n prctcab p w j t rocg yes h ests. d p l d g t n t y z w cu int 6 4 m, whh hd dd a ard o c d u e t i v t gn. g 15 y the rsult o a ct e t he p h o g aph ; : each e e me t o m h 2 eow e v1 (' w cos (afe a d n g the amn xeled b fz g) ; and he com n s h a h e d the wr dags a ndatd ae nd t e zones as nk Both deveoped op c e d zons d j aen o he c n d the b ase n the lef spl e h mimu v d n y i n the top laer was 3 n the bo om in 7% n th t smple the maxm n the was % and h o om T g r e t e t g o the snd ve op e d n th d d e p, th the e vue 20 n e e me and ih am e . Ho ee , n he riht uppr rner of h ef n < t shed mns 3 % whe s ome fe wter hd uuated tp f the semen and wc ate reezngws uded a s p ar t o f ha lem ent yp e o Zs wi h rea tive densi e f les ha n % ths s d ere p q e f ed lea t p ar n y cas the i y o was obvousy th e looened zs . hi l eectd by the derons he and l aye r sen in p h o t o g at t h t op , whi a e mos t prnou nced th roug h u he h pecme n and in t he ep rocaing test n he f p h o og aso aong the sde whee e oten ound the loosest
gyato ts ts th loset egm e ent s ed long th id-ple.
ents ere
ofen
e ate o 'edisiton is pobably ontanty chaning sige yle exept n the compaced p and bttom ne that reain sntialy b shape , b ut bec ome mor om pa ct ih eig nmbe o cycls Ther e adcal dference in disb hh veop n ecprocating and y ests as n he la ter th e greatest water p e d to cc x tests n ac u tE n the top of i n t aly u n m ts t specimn s v whin
e d s r b ution of the I Fi . shn tpica histogra o te ntent nd reltive densty p ecim e s as plaed and e s t s he let upper (2 the ed o t o a tyial rbo n tet n as p r e p a b I hs m the watr ontet a g ed betw ee n a nd 4 % , co rrep ndin e sty btween 50 nd 62%. Te ditrbuon t ne o relate d pproxd a d tb on cu rv A our specimen were at aout the l h ly ee densty, han 0%
;I
7 n th er e f tes t th on tw es of rero ctn he d vope subtt redtruton to e orn e mnt he ur h t rm ho s t red u t o 5 c ce, th r f ret dnt fm 3 % n the oe right the redtrun o e e 6 i s red m 9 to 4% nd t he th stor o oer resebes m a s tr t o u ve ig. 1 7 ur the redstron eem o 64 mnt for tet ater 2 ces o re rt sr Vsua th distruton s usttd us tree r o rete b res fr te ooe t e t reat siti: (2 the tte es for the rne eten nitis s thn 48% 4 8 an %; d () th htchd res o the enst zones th ati nste greter thn 59% h tot ne o retbutio i thi test s beten 33 nd 74% ret denst. P pressure n dispaement trnduer reords are rdud i ig 1 8 or a recirotn shear tst dnse sad ith % hih hnged n 7 1 cs to rage aag reatve et of 7 t 87% In th top digram the ore resure ru cased reach a mximum n cs; nd fter tht most th a patt e repate its n a cc w th th e ek or psur n eh eachn bout 95 of th e 2 0 kg/sq cm vertca ci ig presur O thee sg mnt s of the tota rcor r pr u e : te rst 23 s cs t 34 d s 8 to 71 n th ower recrd t an b se tht the hornt deftons I� fa t en th icasd m uc h mre s o tha th or e res urs p prssur rah d ts m ax m um af tr 2 ces , th df e t on s w on out on n ach directo; but the ket nesng amot thoughout the test dveoed td defctons duing th at 1 d cce of butresnb 7 mm ec h d rect on this ts t th t e ffec tve fnng resur s 2 00 k/ sq th shear sr s ed eteen ± 0 2 kg /sq m th frequ nc 012 cces / s trsducer reord f a grtor tst on secen wth a a tie d t of 50 i sho wn n F . 19 nd the rdtrbutn c i su a ri d n F g . 0 he ffe ct v vert st ress t the at t he test 2 0 kg / s m and th grtor her stress 022 g q m thu s / c 0 11 h frequen s f 00 ces / c
t th t ig . 1 e e that th nu re pesure tad t a ximum equ t bout 2/3 f the it ffcti c fining prsure a d t h n it r em an r onstnt . hs ica r ator tsts perfr e d s te of th s res ns hich i tr rad or cic axia fo ts tsthe rcpotng hih th otest re pe ssur r nd se evn eve ntu t th connng rsure durg th mont hen the c dvatr test at st as t ases thugh r obsrv n rtor ctv rdtrbuton nd ver softn of the tst spmens as i cpcatng he and cc taxa tests
18 go es the eflet be easurd to dicios t I g 19 prodc ef th f ct h recods o h spen ucs ch mu e lcos h f th o l the o his ts h g a g \ 4 gee s n e t mu ch d nser sp ns h g g s sl h h I f s ag =
T rdsbon h nd ths te te cles s h i F g 2 he pa of r bu on s s h silar ha i recpog tts atl h oace zoes ae o th s e an d he ap u he o os ' s concen ae i his st chefl alng h ss ad th e of he ec his ogra m of e s iuon s eal uns cl he s a nda ai s 0 = h developet f pre pessur al o h cofig qefacio, psu w us as a cirio or he oset f cclic wul ocd h ccuso hat canno achv cclc l i aci on i cp rocaig shea a d in g a o shear es s hle ss , ra cal rdis ribu on o h a cn a s ev ig h es sp c ins dev lp o onl cclic riaxal s als o rcp rca i sha a nd esp ecia ll g a sh a Th gar io se s e par cuarl e ffe c i pcig disibuon o water ce and esi in cjuncon wh aivl hgh ffctv ssss
A aisical aalsis h redisriu i all recipcaing rr s s o Ba d g Sa d sh d ha he s andad dvia on cass ith ceasing elaiv esi ad approxmael he amgd pa e of the ss ts hen ohe v ales kep 2 un cha The for to ab oh l rapes g f dard dvai o agesa re b te Spc s a s prepa rd no su c e d shar f al l he w ge o 2 3 ad s spe cins ha d ol s lih l grea ad a viais v a c cl s X
O RSON O CYCL IC STR E SS S NDUCED N S ITU D N AB ORA TOR Y STS
h ses a elmn siu, beneath a lvl go surfac i a omall cnslidae sad staum are llusad i Fg 21 iiall h prcpal stsses ar vercal a hrial, h a pi cip al s ss rao o a u 2 . sho b elemen (A in h l pp cnr an d n h Mhr diagram h s ss c c l A w s ppos a rthqu ak ducd horon tal cclic sha orce s as h i he e lm n des gnaed (B &C Th is ca uses h pnc pal s sing lik a pedlm btwe th pss shw he c l d lm ns (B an d C Th corrspn dng s rss ccles i ci h sngle sress crcl dsgnaed "B&C hich s lager h cicl ad con cnric . Dung ah ccl , all s rsses ae p a ll cc t c rcl es be thes wo c cls p ig m A o B A o C t A h ssss h elet pp a ch a hdoa c s a s r ss , i e a sa n wh ch
9
hr res o a s
i u ne o u y ar ru o
zo.
aso ubjcte i e n a su h niia y o c h r hl p a s e n n he fu d i u T u e s h e cycic hr s r s s s a s u " o y n e r I" I IV " s r s i h eng se b h l e me w vr appoach a e eem su wh e ss of a aoy pien any en ss on us f
e
n an ad n s are or a c unfory diiue he ero rs e e e p ero. v in a s ra fe an w rs f ffer o s n he and shea ess ha are he rfa f ayr w unfory su vr an rea of a few uar ich a any vn mmen un cy oad. w w onsier in Fi. 2 3 h ess on h ounar and n h n ri o f a ria xia s ec e which s fs s o ro pca11y cidaed and hn sujece in unrne sae psve an iv veca dev iaor s res . h r r for h e n i ia i he in H on h horzoa ax n he ohr dara uprpiion of dvao sress ann w and - � cau rss res o emere on oh of o wh a imu dia How w a es san rs o suh cyc di ? houh h s a is s ony ai ve nev ehe ev r i i i cyced hrou h h hydr s ai a o f sres , i d veop ihy oaciv rpons ov a sa ae of devaor r d ih o re reure w iu c u mpn a c o of h ia cy forces nd o ove he ess war owar he op o h pm e. hs ac o, o m ne wih he ffe f ea es aens nued y nouform suon o sresse on h dar ca us s r uon of war onen a o fenn of h p h sp e i . h y c as of he o re rs urs ep i ncin an fna y r e o ai y o h onf n n re ure c im he s re yce hou h he hyros ai c s e o r o ue nray ued eraur for hi henoenon he cim ufe r " iuefc io " areay ex aine h ro du io h resose ars no raosh o acua quacn of d and o diiuh wen hes wo fnanay ffre phna am sn in hs er "yc uefain" whn referrin h phenmeo of reruon a sofenn in a aoraory pcis and when h cyc oe su monariy euas he cnf ini n ressure . Exay how y n hrou h he hyros a s r h pum aion a h yc nea sress raens omne o c h iv yic ufaon n suh s
s re man o e n ve aed .
hans of rediruon n reiroatin re shear e i uc h aie r orhen I i . 4 are show he nonunform ha are en e rae n suc h es s or om parso n, n he uppe ef ha cor r show aan he sresses in an een i iu i whih caio of yc horona sar srses
0
tically mobilize the same sear stree o vert ical ple . tresses o these ficttios boarie o the ele t , a well s in he te ior o f the eleme re i form t ay gie moet . I t ar cotr the b o daries o f eve y te s spec ime re se f or onir stre distribios the oures ad strs gradiets in e iterio r The rbber me r ae ra smi s to th e cyl idric l Because the shear dar y o f the pecime oly orma l st esses . ree o t his verical bo d y are zero , the hea r treses alo g e edge o th e hori zota l face o cp a b se re lso ze ro ere ore he app lied shear orce i s distribte pproimely pr lica l y sho w i Fig . 24 . Bt the ver tig mo met of these ear is verticl ot lced by conterome her ri m, rce orce o the b aries . Thereforeprdced to ma itai byeqilib e ceter of gravity o the orl treses trasmited by te cap d e e m be ecetric d th i ect r city m st cycle back a d This highly oim ad cyclically chagig patter of r dar tresse ad the reltig severe itel tress giets end t o co mpac the sand jacet to the cp a d the base a d prodces e pe o reditribtio tht we hae measred i my reciprocatig d gyratry she ar te st ; Ref 1 ad 13 X
OISON OF SP NSE TO YL IN LABOTRY TESTS
LODIN F ENSE SD S IN S ITU
pendig o he ype o cycli c te t oe or mor e f te fo llwig ree mechais are the major case o redistribtio o water et ad developmet of cyclic pore presres ad cyclic strais a raor tes o ee s ad 1 cig throgh hydrotatic tat o stre.
Ppig acti o by veri cally ap p lied cyclic lo ad .
3
Str gradiet withi the test specime geert by o ir st res dis tribti o o the seci me b res . i mechanism is probably ggrvae by the small sze o e e pecime s i co ve t iol tes ts .
n d dit io a l igt rearr gem et o grais ge er ted by b rea kig of ed e a d corer of grais , ma mak e a ior c ib ti o to the de elp et o pore pres sre a nd t rai i l ab rt or y tet . Thi cim i orally the ol e tht prce pore presrs ad ri withi a eleme of dee s i situ. B i labortoy e thi mechism i completely overshdowed by the efects of the ree pricipal mechani listed aove. cae o te basic iferece betwe the chims tht rdce cyclic pore prere d trais i the laboratory ad i it ider it vir tall y im pos ib ak o etrapolte r et imate r c labo ro ry te t with the he lp of eiric l relat ioship clic pre presre and rai will develop i it.
21 e n ver o ne ma kes u e o em pi ri al el at ionships , there must e E g , th Atte beg lim ts ar e useul n soi l a oon denominator canis or estalishing eral relatnships beause pastiity d various engineeing poperties are related to the ntents o clay articles an d their mi nera logy ) I nnot nd a oon eninato teen the prinipal means that ntrol the yli response in rat or y spe im en s an d the esponse o an elemnt in si u . In loratory speimens the deveopment o pore pessurs and ylic de rmations is the result o radial redistriutio whih in turn s roduced y mechanisms that normally do not eist in an element in situ I leent in situ normally there is no redistriution and the te trses r ema ho tmo eo , ional i , no s t ress graial dietests nts e velp thin sp im n nIn hegen on vent ycli tria cyl ic wlique applyi ng a tion i l eelop eve n in very den se s ans , when u ccl e ; u t in s itu y li li queati on c nnot de e lop , ee pt rha p qui te loc ally ad j acent to er tain rigid o undaries suh a s ie or ir s
Is ere a possiility o developin a labratory test hich wil duicate the stresses within an element in situ? It is oncivale that e coul d sha k a larg e ma ss o sa nd, e g . , a 3 mter ue with a ins trum ented e le me nt in t he cente r o this ma ss to me asue li ore press ure s an d s trains I t oul d be a c ost ly resear ort u orth serious consideration otr aproac ould e the appliation o the inite element to d T o r edue t h nume r o varia b les to a min imum, I sugg s t coaris on o the twodim ensi onal s tress dis tri ution o a n eleme nt in itu it that in a long rismatic test specimen sujete to reci rocatin shear i th zero shea r s tres on the verti al sides This ould e equivalent to ssuming that n situ a sand stratum ontains 1 ric ti onl ess , vert ial , p arallel pla nes hi h ar e alig ned no a l t o te cycli ores , d 2 rou an d ine tensib le hor in tal planes t il l requie e pe rim enta ti on i th v arus assu mpti ns o the s and roperties to achieve redistriution o ater content that approximates at e av e obse rv ed in r ecip roc ating shear tes ts hen I i rs t osidred this approah, I concluded that it is too ompliated or t a aila e tehniqu es But the vis ion o s ucess ul tr eatm ent is ve attractiv that I oul lik to persuad te most resoureul cialists in this ranh o applied mathematics to lend a hand Suce o this approach ould not only esult in a onining dnstration o the enormo dierenes teen the respons o in itu ements and laoratoy specimens to yli loading, ut it ould up the possility o investigating the eects o loal stress ocn trations prduced by rigid odies that penetrate into a sand n su j ec td to eart hqua ke load ng
e o clearly realistic aproah to this entire rolem area ist o comreensive investigations o the ull-scale eperiments ic nature has made and continues t peror in highly seismi rons o te orld But reliale inormation will not e derived ro uc investigation le tey are preeded by etensive ri na y eorts to shar n ou r t ools so t o spea k ; in partiular to 1) or onductig irv available tods ad to develop n mthods
i i tu sur ts of th d of d ss of s ds , d (2 for derii th lti dsty o sll scs cut f u diturbd blck sls tht r tk by h f lwg th rou dwtr Wh r s bl ls h of d t t shl b d trd bcus h dicts tht yu lluvl sd r uc or susctbl ctul liqufc o lot of l trais th ol sis h h s b c lry to y evr r hq ks cu of t h r ud d s t f such opr hs v t o d for th sos , h y woul b st b with o s drtk s coori ffos by svrl cutris uct to svr iic ctivity ot publishd fild ivstit o ffcts of rthquks o d dposi ts r dfi cit d k itrp t t di ff icult U arrtd rlic is plcd o th stdrd dyc rtio tst ich is f r fr s t drd s rct i cd oy E du i orehsiv ivstiti of rth ifr st of ch sd dpoi ts fo r th fou dti o of uclr r lt four dif frt orig orgiios d bors dpdtly t th s loctio, ihi 3 rdiu s Wh copri thir rul t s , th bl owcou t s dffrd by s uch s fct or of thr! ) As b s is fo r s itig e sit u rlt iv d si tis by crrlt io with blo wcou ts , thr is a tdcy to us th v rg b low cou fo r ti r s tr tu is d of pyig scil tttio to lyrs or ls with th lowst blow o ts I lluvil dos it s o o f t cou t rs lr vritios i ds t is b tw d ct o s i th s d st rtu Dt ructiv ots t roud lvl y b cusd by ctul liqufctio of oo rl ily thi l yr ht c b si ly o vrl ook i such rud uurfc ivsttos; cth lyrs rogly b l fo r hvi usd rlyi th o uch t s sr W oredt i l ad or ccurt i sit surts th i possibl by dyic trtio tst, rhs cotuus sttic co trtio sur ets o idti fy th loo s s tht r rspo sbl fo r lg oets Ufr tutly t h id t if i cti o of such os is riculrly difficult r ipssibl wh th lqufid sd hs ld out ltrlly or thrugh crks to th surfc Sud of dscriptiv litrtur of rthquk ffcts o lluvil epos its covicd tht lr d dsructv rod ovt r ausd by ctul liqufctio or v softi of cotrcti sd lars t so dh I bliv tht th liqufcio otil of u lyrs could bst b dtifi by s of triil tsts o dis turd s ls us g lo d cotrol tsts f h ty dvlpd by t ro (R f 8) . or typicl lluvil sds cotii ot or th w prct o i lt iz th u p r li it of rlt iv d si ty for whic h I o sidr tual lqufct io to b posibi li ty , lis h r b tw out 0 d 50 , ddi g o cof p ur Mdiu - loos sd s i te rag btw 40 d 0 y b slhly cotrctiv or slihtly dlat iv ; d i s tu t hy y s od t c yc l c lodi wi th s trs obj ct io l g i tud , ut rrl y with ctul liq ue fc tio trogl diltiv aisotropiclly cosolidtd sd situ, with
3 relatie desities greaer ha about 70% cosider t oay iossibe fr cyclc ore essures o aoach or qual the cofiig ressure becae datecy wl autaicy cause e gai sructue to ffet oss of rg y bacg sef" so to sa ruirig ly ute s a a dou ha duc ed ore ressures il ormaly e as hgh as 50% of he cofig ressre cotra t la bo ratr y sece s of dese a d h e rdis ti uio f aer cote deele of sft es ad cycic lquefact cao be reed by oiiaio f a trog dlate esose Secia codiio that ay cause deloet of high ore ressures d excessie strais dese sad i situ ae the follwig: 1 djacet to igid oudaries whee seee stess gradits ca d elo a d roduce redis trib utio of water coet t would eit estigatio whether such esose could ubstatially reduce the ari g c aacity of i les or iers dug earthqu a e (2) f i t e r e ossible to aitai i situ a ass of sad a isotroically coso dated st ate du rig c ycli c loadig w hi ch dou b t th e c ycl c liu e actio i dese sad ould be theoretically ossible X
OATORY E T OR DE UO E - NTE RM O NDATON
U i suc t i e h e ore rel ia b l d ata wil l be aailabe o the i situ resose of diatie sads to cyclic loadig hae recoeded rece ears a a te orar exediet the use of ccli c triaxial sts o aisotroically cosolidated secies cut fro ndistured s l es de e de t ly ot er i es tiga tors ha e also te de d to clude cyclic triaxial tests o aisotroically cosolidated secies S e g the cor ehe sie reort ANALYSIS OF THE SLIDES IN THE S E DAMS DURING THE EATH -UE OF FEBRUY 9 9 (R ef 14) n recoedatio ad teretatio f such tests o guided he fo ll oig o s eratis ad assu tios :
1 n cyclic tri axia l tes s redi s tribuio a d buildu o f cyclic or ressures becoes articularly seere wh the ea ore resures rise el l aboe 50% of t he cofiig re ss ur Thref ore use as a es arete the ume o f cycles N O required fo the ore rss ure t o rea ch 50% of th e c ofiig ress ure At that oit he cclic s trai s are ge eall y s ti ll sal l
e he follwig arbitrar assuios hich cosider o e el l o the s af si de ( a) hat the echa s which causs the ildu of ore ressure i situ (breaage ad r rearrageet of rais is resosible fr oefourth of the cyclc ore ressures duced i laboratory tests with the oher hree-fourhs roduced by chais that caue edisiutio of aer cotet ad softeig o h es t secim e s ; ad (b) that i si tu h ore res sure wil l rac h 50% of th cofiig ressure which i j udget is a uer f or ediu des a d dese s ads u dr the ors codi ti os 3 ore res sure i th e tes ts ris es aoxi ae ly i direc t roorio to th uer of cycls u to about 50% of th cofiig rs sur n coiig this ith the recedig assutio it folls
4
tat te ner of cycls wic wold be reqired in te confining presre wod be 4N 5.
sit o re 5
4 an e timte of te cyclic s tri n i reqred it wo ld e neer to crry ot precise easrements f te cycic strin b asi f or dg ent , eeping in t nd e these strins nd tat only a frcion of these strns re csed y he mecnism t a cntrol e deveopment of pre pressres d strns n sit gin I wi to ephaie that the stris wic evelop in cyclic bortory tes en the pore pressre ri e cnining prere (i.e. wh en cyclic l iqefcion i rache , w ell as ll frer inc ree in cyl c str ins rin i inl cy cl ing ber n o re l tion hi� to te t rain ht eveop in sit r il ive snd in y g e nt . nider te e of te bet poible distrbed smples reerably ndct block ample) n essetil reqirement for n ingf reslt . I also reqire te e o te type of lbricted nd ic were d eveloped by Profe ss or P . Ro we nd L Barden (Ref . 15 ) . n en eral te te t specime n old be precosol ida ted t p rinci pl trerati o 2.0 r averag allvil and it ilt fraction not exceeding a e er cent d w it a r ela tive deni ty in te r ge be twe en a bo t 4 n 6 0 e deal it tran it ion rnge beteen loose s ds tt c an actally liqefy nd medim dene snd tat re afe gainst ctl i efacti on . In tt r nge it is oible t trixil tes ts wi t notnc l d incre s e ing dea l oad i ncre mens wil l prodce age s trin t n cycl ic loa ding. T eref ore , in t t ran I reco en ls o perf orm anc e o mon tonic l ad co nt rol t e sts on spec ime ns at are aniotropically cnsolidted with a principal stres rtio of . 0 . S t et s also per mit bet ter differe nti in be tween dil ve d ontra ct iv e rpon e ta n cyc li c te ts . Unles t e rel t el develped dilative response I advise performace of dditinal t to define te F l ine Fig . 6 a nd 8 X SUY F PRIC P CNCLU N C AT IS
1 enral To dif fer entiate be we en basicall y dif fere nt nna wic are bt called LIQFACIN n li terre te term L LIFTIN i ed in tis pper fr te respone of contractive e) and ic relts in sbtntil increse n pore pressre a nd o tren gt t c n c e a flo w s lide; nd te te YLI LIQ i ed for te te pore reponse of dilative dense) snd cyclic iaxial tet en presre rie momentrily in in ec cyle t cninin prere.
IN
. ri tcl vo id o Te concept o te critical void rio v ope d mre an 4 year go ( Fig . ) h to be mified on te ai invetigation of e fl lide in Ft. Peck da and Cstr' vetn (ef 8) ring actal liqefactin te and gan
rr range teelve nto a lo trctre c enre a n r tan ce to o B ng ea oa n ( n ran e) traal tet leacton can be aceve at greater elatve ente tan t ran control t t an te tpe o t e t ear ago gave rl t eve n ore on t nae e e e rlatonp between crt cal vo rato an eectve nor prncpal tre obtane t o tropc al onoate pc en ( 2 ) anorop call 6 conol ate ec en an 3 wt cclc oa tet g
3 Rdsrbu f Wr C Cy Tss e pro reve nceae o cclc pore prere an otenng n tet c en n varo tpe o ccc te t n n cc q eac ton n raxal tet , are cae b racal r t rbton o te wat r co ntent c gnerat b ecan tat are noal not actve n t -
4 Becae t d fr F d el tat a laboator tet can be ve w wll proce n tet pecen te tpe o nor tre trbton tat ext rng cclc loang n a tpcal eleent n te ator beleve at clong te great gap betwen laborator an n t repone ll rre rarl copreenve el nvetgaton o an l-cale t a natre a perore n gl e c reon c n egaon l ave to be carre ot t te bet pobl nvt gaonal tool an t etco attnton to etal le t l l crat e leang no at on -
ss - Fo r et at ng 5 rm Rs fr n repone o e ene an ene an e an w c ar ae agan t actal l qeact on an nter procere ge te c ba on peroranc o cclc taxal tet on anotropcall conola te en ct ro n trbe a ple
Wenever a qeton are weter a an trat col eperen actal l qeac t on t rcoe ne to carr ot tet wt ea loa ncre ent a e velope b a tro (e 8 , ng te bet oble n trb e a p le XIII
KEGEN
Te octoral reearc b onzalo atro an Frann enon wa ae po ble w t a t ance o an or gan a ton an nv al Detale ac nowege nt wer e ncle n at ro te (8 an ll be n cl n n on te I a artclar nebte to tanle Won or crtcal revew o a rat o t paper an a n valabe gge t on
26
XIV. RFENES ()
asaga nde, A. , " haateis is f hesi nless Sis Affeting S ii ty f Eah i s , Jurnal f the Bstn o Civi Engi neers , Jnu a 19 36 . ei ed n "TRIBIONS
(2)
L CHANIS
-
Bstn
f ivi
Obe 1940.
asaga nde , A. and Wats n , J . D . , " patin
Tests nd iia
nsiy Inestigains f hesinless Maeias f anklin as Dam , Aend ix BII in t the U S sn, 38.
3)
asagan de , A. , The Seai ng Res is e f Sis nd it s elain te St ai i ty f Ea D" , ee din gs , Sils and Bstn, Fodtn nerence the U.S s. , ue 193 8.
()
Wats n , J. D . , "S ess-D efm ai n h aaeis is f hesin 1ess Sls f m Tixial essin ests" , S . D . The i , Haa d vsity, 10.
s f Enginees , U . S . Depaten f he A , " Rep t n he ide f a t in f th e Upst e Fae at t e k D. " . Gvernt Ofi ce, Washi ngt n, D . . , Ju ly 1939 .
6) d dl e k s , T . A . , "F t ek Slide" ,
CE ransacti ns , .
7, 2.
(7)
as agan de , A. , " le f e aulaed isk i n Eah a nd dati n ngineeing" , Jurl f the Si Mechanics and Fod tns Di visi n SE, July 196.
8
t , G nzal , "Liqu ef atin f Sa nds" , Hrvrd Si Mhac re N . 81 , a nu a y 196 .
()
ast, Gnza, "Liquefatin d yi Mbiliy f Satuted ds", u f the tecica Divisi , e 15.
( a)
Seed, H . B . d Le e , K. L . , "Liquefa in f Satuated Sn ds duing Lad ing " , urn f the Sil Mhnics d undatin ii AS, Nee 1966.
()
Lee,
( )
K L. nd Seed, H . B . , " yli Stess ndi i ns a using qu ef a in f Sa nd" , url the Si Mhai cs ad ot Div s i n ua 196 .
S kley , W G an d Alin, R G . , '�n -Ui f ndi in s in a xia T es t Sei es" , eearch neree n Shear o Cohe Sil s AS, Bude, ad, 1960.
27
12 )
1) )
asagan A an ndon , F , " i oing n d Gyoy Sh Aaa - Dig n s in g od nd on Sd nd ," o o WAEWAYS EXPERIME STATI iksg, ss , 1976 ndon, Dotoal Diaion Un ity, 9 6
in atio n , Ha a d
S d , H Bolt on, K L . , d i s , M. a d Md i i
l,
"alsis h th q k fofh F Slid 9 19in1h " San Fnando D ding eearh Ctr ot o EE 7 -2 Un iit o f a ifo nia , ly, alifonia, Jn 97 ( 5)
ow , an d Badn , L , " oan of F End in Tiaxial Tsting, rna Si Mechas an d nda tns vs n AS E , J an a y 96 4
