6585845005

ESOPT

State-of-the-art report

Penetration testing in USA

J.H. Schmertmann

Because of Its versat1l1ty the Standard Penetra-t1on Test (SPT) has evolved as by far the most common type of penetration test In the U.S.A. — desplte the quant1tatively crude results obtalned from the SPT and a comnon lack of controlled re-search to support the var1ous correlatlons often used 1n design. Only the quas1-stat1c Cone Pene-trometer Test (CPT) of the Dutch type at present shows promlse of supplementing, on a large scalę, the use of the SPT for soli explorat1on and pre-llmlnary design. The report concentrates on the pervas1ve SPT and Its foremost contender, the CPT.

1.    GEOLOGICAL CONDITIONS

The U.S.A. Includes a very broad spectrum of geologie env1ronments of deposltlon and forma-t1on of soli and rock materiałs of vary1ng sult-ab111 ty for penetration testing. In glaciated areas, about 18% of the U.S.A., solls vary from hard boulder t111s, generally unsultable for penetration testing, to large varved deposlts such as underlle Albany, N. Y. and parts of New York City, generally sultable. Also sultable are the compresslble clay tllls that underlle such gladal Great Lakę cltles as Detroit,

Chicago and C1eve1and. Alraost the entlre east and south coastllne areas, about 7% of the U.S.A., have very sultable, often weak and compresslble, Coastal plaln soli deposlts. Wind blown deposlts cover about 10%. Resldual solls cover about another 33%, and have var1able su1tabH1ty for penetration testing. Another 10% conslsts of a11uv1a1 and lakę deposlts, usually sultable.

The reraalnlng 22% conslsts of generally moun-talnous areas, whereln only the val1ey deposlts may be sultable for penetration testing. See W00DS. et^ al. (1962) for a morę detalled sum-mary of engTneerlng soli condltlons in North America.

In ■yiew of the great d1versity of geotechnical condltlons 1n the U.S.A. 1t seems remarkable that any penetration test could ach1eve almost un1versa1 U.S.A. acceptance and use — as has the SPT.

2.    THE STANDARD PENETRATION TEST (SPT)

2.1 Use In Practice

One of the key reasons for the popularity of the SPT 1n the U.S.A. 1s that w1th th1s test the en-glneer can obtaln soli samples and a penetration res1stance index, the blowcount va1ue N, 1n a very wlde rangę of soli condltlons. The SPT equ1pment usually allows the englneer the eco-non\y of sendlng out just one piece of equ1pment wlth good assurance that 1t can perform the ex-ploratlon asslgnment no matter what the dr111 crew flnds at the s1te.

SPT N-values have been correlated agalnst many foundatlon design parameters, such as footlng bearlng capadty, pile capaclty, relative den-slty, clay undralned strength and sand frictlon. See, for examples, PECK ęt^ al. (1974) and FIETCHER (1965). After consTHerlng the many yarlables knovm to affect N-values 1n the SPT, as also dlscussed 1n 2.2 hereln, many engineers 1n the U.S.A. now recognlze that the SPT N-values prov1de, when used together wlth sample classl-flcatlon, at best only rough quant1tat1ve data for preUmlnary design purposes. However, some engineers apparently wiłling to make finał design recomnendatlons on the basls of SPT data. Others conslder the SPT as merely a convenient method for obtalnlng stratlgraphy and samples for classlficatlon and place no design rei lance on N-va1ues.

2.2    Poorly Controlled Yarlables

Desplte the fact that this test originated 1n the U.S.A. about 45 years ago (MOHR, 1966), and ASTM Standard D 1586-67, recently rev1sed, covers this test, engineers st111 wldely use this test wlthout proper control of several yarlables known to affect slgnlflcantly the blowcount, N, obtalned from the test. Some use caslng for hole support and water as the drlll-ing fluid whlle others use an uncased hole and drllUng mud — the latter usually produces a greater N. Rotary dri11 Ing methods for advanc-Ing the hole are now comnon, but chopplng and washlng are stlll employed — usually produclng lower N. We have great divers1ty 1n the welght and power of the dri111ng equ1pment used — ranglng from heavy equ1pment capable of drlll-ing over 300 m and uslng heavy N-type sampUng rods to portable tri pod Systems uslng relat1vely very llght E-type rods. Shorter and Hghter sampUng rod strlngs usually produce greater N.

By far the most common method for lifting the hammer 1s by tlghtening a ropę wrapped usually at least twice around a power dr1ven, horlzontal drum or "cathead", and then qu1ckły slackenlng this ropę to drop the hanmer for the blow. Common practice usually includes no mechanlcal control of how high the operator Hfts the hanmer. Ind1v1dual operator preference and techn1que determlnes the type of ropę used, the number of wraps, and the qu1ckness wlth which he slacks to drop the hanmer — all of which can greatly affect the rope-cathead frictlon reductlon of del1vered hammer energy. The rev1sed D 1586 stlll does not Include a mandatory free-drop mechanlsm, nor does 1t limit the shape of the harnner or speclfy cushion and capblock require-ments. Common practice omlts any cushion or capblock, and thereby also inereases N.

For further dlscusslon of the yarlables affect-1ng SPT N-values, see for example FIETCHER (1965), MOHR (1966), and SCHMERTMANN (1971).

2.3    Research and Recent Developments

Cons1der1ng the thousands of SPT dri11 r1gs 1n use throughout the U.S.A., and the many years of its use as a soli englneerlng explorat1on and design tool, 1t seems Incongruous that this usage lacks the foundatlon of a large body of controlled and documented research.

The most widely known controlled research study uslng fuli-scalę SPT equ1pment 1s that by the U.S. Bureau of Reclamatlon (GIBBS & HOLTZ, 1957) correlatlng N-values agalnst relatlye denslty 1n sands. The results from this work emphaslzed the need to correct N-va1ues for thelr 1ncreas1ng value wlth increaslng overburden pressure — which has now become conmon, but stlll not