U.S. DEPARTMENT OF COMMERCENational Technical Infornation Service

AD-A032 962

I

STANDARD PENETRATION TEST AND RELATIVE DENSITY

ARMY ENGINEER WATERWAYS EXPERIMFNT STATION,

VICKSBURG, MISSISSIPPI I

FEBRUARY 1971

t

ýy i

AVSC!ýL-4EU PAM R M.7 I-I

STANDARD PENETRATION TEST AN4RELATIVE DENSITY

IL r

I~~DEC9Li

NATIONAL TECHNICALFem I?~I INFORMATION SERVICE

11 PARIfPOT of COMMERi~c

- ~~ & L S A m y #A t* h C i nePIqmltO VA 2k

fb6~ of UL!. Aamv Emgw Vebmwwqs Gaip00 11sbhusmt vkbmmN-

Ab~I &&m bin bw"pw palft mkm nalf ai *b ias~u h haft"

Unclassified

9MUCIMIT CONTROL DATA -r RDAdd~ A"ee a"U.e.. eftt= ý.fe swe be anan .00 410 "" ORNM to eae

I ONIGIW1A TIWO ACIVIY~(. UW "P=~T CAPICAT"M

U. S. Army Engineer WaterwasEprmn tain UcasfeVicksburg, Miss

STANDARD PENETRATION TEST AND RELATIVE DENSITY

4 oiscomovive wOevs 1rno o mo," ed hwmi d~o)

Klaus- Jurgen Meizer

0 ?&T ^6T TOTAL 100. OF PAGES 8 ~ mp

February 1971 7 "'5'G& CONTRACT ON 401ANT~@ "a. O5IEIWAVOWG 01600"T Uwgusem11

& .0-~c-w 1T062103AO46 Miscellaineous Paper M-71-1

*Task 03 101ý OTall Vqm.O ODOM (Ado 00o40. ftfS.*"dfd

I0 04EYMSUT104i STATEAMET

This document has been approved for public release and sale; its distributionis unlirnited.

1I 9UPWL9WMESTARV NOTat ý-aper to be presented a 'SOOIW

Fourth Pan American Conference on Soil U. S. Army Materiel CommandMechanics and Foundation Engineering, Washington. D. C.Ran, Tiln. Puerto Rico, -IS June 19. 0" 1__________________Is AESSTA"TV

* Since ground water greatly influences penetration resistan( e of soil, an empiricalrelation was established between the number of blows applied in the standard pene-tration test to sand below ground-water level and the corres~ponding number appliedto air-dry sand at the same relative density. Also, since the number of blows was

* found to depend not only on the relative edensity but also on the compactibility and thegrain size of the pe~netrated sand, an empirical relation was developed between thenumber of blows and the relative density, with compactibility and mean grain diam-eter taken into account. This relation was verified by results froyn laboratory tests(onducted writh a small static penetrometer.

DD sIw 04473 "a"acs o ea 1AS. 0e4C Un!s

Unclassilied

I.L~Ua Lou" L"Im 41

KyOfw"L.6 W, QOLG WIT OMLa m

Ground waterPenetration resistancePenetrometersSandsSoil properties

Unclaqs0fflod

I (0} •.• a..•..

Dos""y this report when. tolnger noehidecl Do net returnitf to *No or 1ginator.

The fii.iiings in t~ s report are not tco b. construeJ as an of~icialDepartment d th~e Army pcosition unless so designated

6y other autlinoized docutnonts.

FOREWORD

Thi•. paper xas prepared by Dr. K.-,1. Melzer as one of the

I nr1it States',; o ontribut•ons to the Fourth Pan Amerri an Conferent e on

S,11l ,Me( han i s and -ouindat ion Eng ineering in San fuan. Puerto Rico, 1-4-

Is Iit•' 1071. The ,onference was organized as a Regmonal ( onference of

t-kc International Soi tetv of Soil Mechanit s and F.mndation Engineering

,Ind as a Spe iaitv (,onfcrence of the American Sot iety ()o ( 1",11 Engineers.

T'he researt h reported upon herein was based (,n the author's

d,, iral dissertation and data collected in studies ýonducted at the U. S.

Arnvy Engineer Waterwavs Experiment Station OWES, under DA Project

I T0-I' I 03A04(,, "Traffi, ability and Mobility Re:;earch,," Task 03, "Mobil-

,t1 Finodamentals and Model Studies," under the sponsorship and guid-

,,t ,. of the Researt h, Development and Engineering Directorate, U. S.

Arm M iate rmie ( ormunand.

C OL Ernest D. Peixotto, CE, was Director of the IVES during

this study and preparation of this paper. Mr. F. R. Brown was Technical

1) re t to r.

Proeiling page blank

6955-2

A

STANDARO PENETRATION TEST AND RELATIVE DFNSITY

La Prueba Normal de Penetraciocn y la Densidad Relativa

K.-J. Melzer PhD, Research EngineerU. S. Army Engineer Waterways Experiment Station

Vicksburg, Mssissippi, U.S.A.

SYNOPSIS

Since ground water greatly influencea penetration resistance of soil,an empirical relation was established between the number of blows ap-plied in the standard penetration test to sand below ground-water leveland the corresponding number applied to air-dry sand at the same rel-ative density. Also, since the number of blows was found to depend notonly on the relative density but also on the compactibility and the grainsize of the penetrated sand, an empirical relation war developed be-tween the number of blows and the relative density, with compactibilityand mean grain diameter taken into account. This relation was veri-fied by results from laboratory tests conducted with a smaill staticpenetrometer.

Se OPSeS

Debido a que el agua subterranea granclemente influve la resistencia asuelo, se establecio una relacion empirica entre

el nurmero de golpes en la prueba normal de penetraci'n (StandardPenetration Test) de una arena bajo el nivel freatico y el correspond-iente numero en una arena seca (ie. sobre el nivel freatico) con lamisma densidad relativa. Asimisnmo, porque se encontr6 que el nu.-mero de golpes depende no solo de la densidad relativa sino tambiende la compacticidad y el tam a~o del grino dý )a arena penetrada, sedesarrollo una relacion empirica entre el numero de goipes y ladensidad relativa, tomando en cuenta Ia compacticidad y el diametromedio granular. Esta relacion se verifico con los resultados obtenidosen ensayos de laboratorio ejecutados con un penetr6metro est'ticopequeno.

INTRODUCTION

One of the main problems encountered in subsoil e'xploration is insitu determination of relative density and related characteristics ofcohe.ionless soils. The deep penetration test, one of the earliest ap-proaches to solution of this problem, yields results that are used toempirically correlate certain soil properties with resistance meas-urements. In use today are not only a variety of static and dynamicpenetrometers, but also numerous empirical and theoretical relationsbetween the results from specific penetration techniques and the prop-erties of cohesionless soils, e.g. results from the standard penetrationtest and relative density.

The standard penetration test (SPT) was developed primarily forsampling cohesive soils. Its secondary purpose was to mtasure

I

penetration r:sictance by counting the number of blows required todrive a sampler I ft into the soil. The test also is used today ir ,ohe-sionless soils. but primarily as a method for measuring penetr ionresistance rather than for obtaining undisturbed samples.

The author ts xware that there is a cornroversy concerning the ap-plicability of the standard pcnotration tes, (Moretto. 1963; Irelandet a&.. IQ701. Newertheless, a technique will be shown in this paper forevalhuating the relative density of sand from the number of blows ob-tained from the standard penetration test by taking into account thetompactibility and the grain size of the sand under consideration andthe possible existence of ground water. Furthermore, the techniquewill be shown to be applicable to the evaluation of relative densityfrom penetration resistance measurements with static cone penetrom-eters. The concept, based on a few general considerations of whethappens during a cone penetration into sand, is purely empirical, andit is offered only as a beginning and an encouragement for furtherresearch.

TESTS

Results of standard penetration teste conducted under laboratoryconditions in fo'ir sands with different gradations (sands 1-4 in table I)

Table I. Sand Properties and Types of Test

Sand Source of Data Type of Moisture Corn- Coeffi- MeanNo. Pene- Condition pacti- cient of Grain

trom- bility Uniform- Diametereter D' * ity C d **

u m

mm

1 U. S. Bureau of SPT Air-drlr 2.36 '0.0 0.23Reclamation(1953); Gibbsand Holtz (1957)

2 U. S. Bureau of SPT Air-dry, 1.31 5.0 1.-10Reclamation sub- 0S(1953); Gibbs merged

and Holtz (1957)

3 Menzenbach SPT Air-dry 0.62 21.0 0.42

(1959)

4 Schultze and SPT Air-dry, 0.76 2.4 0.55Melzer (1965) damp,

sub-me rged

F, Melzer (1971) Cone Air-dry 0.51 1.5 0.12

6 Melzer (1971) Cone Air-dry 0.59 1.6 0.27

7 Melzer ( 971) Cone Air-dry 0.63 2.5 0.50

4 Compactibility D' = (emax - emin) /emin according to Terzaghi(1925).

:* As defined by Burmister (1938).

2

were used to develop the concept discussed herein. Its applicability tothe determination of relative density from static penetration resistancewas evaluated from . one penetrometer tests conducted recently a& theU. S. Army Engineer Waterways Experiment Station (WES), Vicksburg,MLssissippi, in three different sands 'sands S-7 in table I), also under.arefully controlled laboratory conditions. The WES tests were con-duti ted with a mechanical cone penetrometer: the cone has a basediameter of 2 um and an apex angle of 30 degrees. This penetrometerwas not developed for deep penetrations, but for exploration of the toplaver 40 to 11 c m) of the soilunder tonsideration. The pene- SANDO VI

tration speed in these tests was SOLT S0.03 m/s.

A detailed description of ".-+4 ,,s"*-4the various soils and test pro- ., 1 ....... -. . ..cedures is not within the scope .of this paper; therefore, only I . .. ,,4

ertain pertinent properties of kthe seven sands are listed in .table ', together with the - *....... 4 .sourte. of data and the pene- . .-tronieters used. Grain-size 10 4 1+ + 1

distribution curv s are pre- . - -- 4sented in fig. 1. ". ,_ I 0

INFLUENCE OF GROUND Groin size, mmWATER Fig. I Grain-Size Distribution

In a permeable sand with a Curves of Sands Investigated-iven relative density Dr , thenumber of blows N ot the standard penetration test is srmaller belowground-water level than above the ground water (Menzenbach, 195'1;Rodin. 196,1; and Gawad, 1964). Some investigations seem to indicatethat the magnitude of this difference depends on N and, therefore, onr. , which is directly related to N . But because most results were

based on field tests in which relative density could seldom be meas-tired accurately, there is some doubt as to whether the relative densityabove the ground-water level was, in fact, the same as that below.

To examine the effect of ground water more closely, the resultsof the laboratory tests on sand 4 (table I), which was tested not only inair-dry and damp states but also submerged, were evaluated as fol-lows: The number of blows N counted in tests conducted at a certai:idepth and at a certain relative density with no ground water presentwas compared with the number of blows N' from tests conductedbelow ground water at the same depth and the same relative density. Astatistical analysis of the data yielded a linear relation between N and

N' (fig. 2) for this sand. In fact, this relation shows that for low Nvalues the decrease (in percentage of N) from N to N' is largerthan for high N values, and the difference, therefore, depends on rel-ative density. A similar evaluation of the results of tests on sand 2,the only other sand without silt particles for which results from testsin air-dry and submerged states were available, shows the data points(lustering fairly well around the relation between N and N' forsand 4. From these reqults, a cautious conclusion might be drawn thatfor medium and coarse sands, the relation between N and N' ismore ar less independent of sand type.

3

The number of blows it-. smaller below ground water than

above because the effective unitweight of the sand in a sub-Imerged state is smaller than in

S/a dry or wet state and because-, +-_ the dynamic action of the pene-

trating sampler causes a quick-S/, sand effect, at least in very/Standar loose to medium-dense lands,

ae ,tion resulting in decreased penetra-tion resistance. Therefore, theSinumber of blows measured

Jr below the ground-water level inI medium and coarse sands thould

N W+ VNWLt21 be corrected for these influ-Consitoton co.fficient e.O909 ences by means of the relation

S. Send io, 37 in fig. 2 before an estimate of

2Send S relative density is made.

"RELATIVE DENSITY EVALU-S0ATED FROM NUMBER

Number of blows N' botew 9gmund water OF BLOWS

Fig. 2 Influence of Ground Water on In recent years, compari-the Number of Blows in the Standard son of relative density valuesPenetration Test evaluated from the number of

blows by various existing rela-tions sometimes led to contrasting results (Doscher. 1967; Tavenaset al.. 1970), possibly because nearly all such relations weee developed

from results of test. !:onducted in different types of sands. These de-viations are not too surprising: however, the influence of the sand typeon the number of blows can be taken into account by relatively simplemeans.

When a cone or a standard penetration test sampler penetrates acohesionleas soil, the grains are displaced. The forces required fordisplacement depend not only on the relative density but also on com-pactibility in that the grains in a highly compectible soil can be dis-placed with less difficulty than in a soil with a low compactibility butthe same relative density. Thus, penetratior resistance is greater forthe latter case. Earlier investigations with ptnetrometers support thisreasoning (Kclbuszewski, 1957; Muhs, 1969). On the other hand, pene-tratior. resistance is greater in a scil with large-diameter grains thanin a soil with smaller grains. For example, when a gravel and a sandwith the same relative density and com,-ctibility are penetrated. pene-tration resistance is greater in the gra, ... Thus, compactibility andgrain size, the latter characterized by the mean diameter, influencethe relation between relative density and the number of blows when thestandard penetration test is used.

The general form of one proposed relation (Schultze and Melser,1(q65) for the determnination of relative density from the number ofblows, with overburden taken into consideration, it:

Dr = aI log N - a2yD + a 3 (I)

where Dr = relative density in percent; N = number of blows per 30cm of penetration; y urit weight of the overlying soil; D = depth of

4

L ...... .....

the point of the penetration test below'%-he soil surface: yD a *ffe,:tive over-burden pressure in kg/cmZ; and aI , sad / nd

2 , and a3 are constants. For Sand San

easier interpretation and comparison,it is assumed that the tests were con-ducted at the soil surface, which leads r-to VD = 0 , and equation I becomes

Droal-tog N. a)Dra a, log N + a3 (2) 0300

If two sands are assumed to have a, .tonfl0different compactibilities and a 3 & 0the relation between relative density m A b oand number of blows can be plotted asshown in fig. 3a. _ Relohoi between aj and c'

At the same relative density, thenumber of blows increases with de- sandcreasing compactibility. Angle p , f (d, a nd dravelwhose tangent correspondsa to constant C dna1aI in equation ?, can then be seen toincrease with increasing compactibil-ity. On the other hand, if two sandsare assumed to have the same com- Dr aItogN apactibility (0 = constant) and the same Ias' toi'l const.relative density, the number of blows I G3 0 0increases with the mean grain diarri-eter (fig. 3b). Thus, the intersection , __.

on the .elative density axis, which is Number of blows tog Nequivalent to constant a 3 in equation b Rtation between a3 and dm2, deereases with increasing meandiameter.

The above considerations were Fig. 3 Influence of Compact-"ealidated by using the results of tests ibility and Grain Diameterwrith sands 1, 2, 3, and 4. If equation on Number of Blows2 in its general form is valid for allsands (this point is not under discussion in this paper), correspondingequations for sands 1, 2, and 3 can be established. Constants aI anda3 for the relation between relative density and number of blowsmeasured above ground water are given in table II.

Table II. Constants aI and a3 (Equation 2)

and a4 and a5 (Equation 3)

Sand aI a3 Penetrometer Sand a a5 Penetrormeter

No. No.

1 46.1 31.1 SPT 5 71.Z 53.9 Cone2 38.3 38.2 SPT 6 75.5 45.0 Cone3 30.6 42.5 SPT 7 77.2 35.2 Cone4 31.7 39.2 SPT

Plots of compactibility Dl versus constant a (fig. 4a) andmean grain diametýr dm versus constant a3 (fig. 4&) show agreementwith the general considerations concerning the influence of D' and

A51-i

dm shown in fig. 3. As happens oftegi, there is one point (sand 1, fig.4b) that diminishes the validation. However, if the fact is taken intoaccount that the data came from three sources and, therefore, maycontain some scatter, it is surprising that only one point is an outlier.Thus, ef least the ge-neral trend of the observations concerning the ir-fluence of compactibility and grain smie seems to be reasonable.

RELATIVh DFNSITY EVALUATED FROM CONEPENETPATION RESISTANCE

To confirm the a.bove trend and check whether the general con-cept is applicable te tlhj relation between relative density and resist-ance to penetration of static cone penetrometers, the results of testswith sands 5, 6, and 7 were analyzed as described above. A statisticalanalysis showed that the relation between relative density and average(0- to 15-cm depth) cone penetration resistance q for a specific sandcan be described best by a function of the general form:

Dr a a4 log qc + a5 (3)

Constants a 4 and a for sands 5, 6, and 7 are listed in table II.Plots of compactibiliky D' versus constant a4 (fig. 5a) and mean

-- o Sand I A Sand ?

I Sand Si IISanSand 3Ifuno osat

30 0 t4 o Constant 046 0Constant ol 0L Inf1uence of V an 0%

9 InfLuence of ( on a,

E Sandi1 ... Sand 2

Sand 2_T • 4 Sand4 4.3_

S e $and I0&0 & Sand 0,L2-i sa Sand 7

30 34 1 4 46 0Constant Oa Constant a0 o 1

IL Inftuence of dm on o3 JL InfLuence of dm on Sg

Fig. 4 Factors Affecting Rela- Fig. 5 Factors Affecting Rela-tion Between Relative Density tion Between Relative Densityand Nunmar of Blows and Penetration Resistance

i .. .6

grain dia-neter d,. versup constant a5 (fig. 5b) show the same trendobserved in the resuits with sands 1-4, even though the overall varia-tion of D' and dm .aas not as broad in the cone penetration tests asin the standard penetrati'.r, tests; soil selection was limited in the conetests by the capa'ty of the pressure measuring device. Furthermore.the cone penetron%,.-ter, because of itE greater PensItivity, respondedmuch more to a change in the mean grain diameter than did the pene-trometer in the standard penetration test.

CONC LUSIONS

In the standard penetration test, the number of blows measuredfor a given relative density is larger when the test is conducted abovethe ground-water level than when it is conducted in submerged sand,at least in medium and coarse sands. Thus, before any estimate ofrelative density can be made, the number of blows counted below theground-water level must be corrected for this influence. The correc-tion can be made by means of the emp)irically established relation infig. 2.

The number of blows depends not only on the relative density,but also on the compactibility and the mean grain diameter of the con-sidered sand (fig. 3). Ba.- ed on a qualitative explanation, an empiricalrelation can be used quantitatively to take into account the effect ofcompactibility and mean grain diameter on the constants of a given re-lation between relative density and number of blows (fig. 4). Compact-ibility and mean grain diameter can be determined from disturbedsamples taken from the borehole in which the standard penetrationtest is conducted.

The cone penetration resistance in the static penetrometer testswas influenced by compactibility -nd mean grain diameter of the in-vestigated sands in qualitatively the same way as the number of blowswas influenced (fig. 5).

Further research should be conducted to confirm and extend thebasic empirical. concept developed. A real standardization of the"standard" penetration test would be useful so that evaluations basedon the results from various research agencies would be more valid.

REFERENCESBurmister, D. M. '1938), "The Grading-Density Relation of Gr•anular

Materials." Proc,.edings of the American Society for Testing Mate-rials, Philadelplha, Pennsylvania, Vol 38, Part II, p 587.

Doscher, H. D. (1967), "Relationships Between Sounding Resistanceand Soil Properties," Journal of the Indian National Society of SoilMechanics and Foundation Engineering, New Delhi, Vol 6, No. 3,pp 313-337.

Gawad, T. E. (1964), "Study of the Natural and Mechanical Propertiesof the Soil Forming the Sides of Suez Canal," Report No. 21, Engineer-ing Department, Research Center, Soil Mechanics and Foundation En-gineering Division, United Arab Republic Suez Canal Authority, SuezCanal Authority Press, Ismailia, Egypt.

7 69542

Gibbs, H. J. end Holtz, W. G. (1957), "Research on Determining theDensity of Sands by Spoon Penetration Testing," Proceedinge of theFourth International Conference on Soil Mechanics and Foundation En-gineering, London, Butterworths Scientific Publications, London, Vol I,p 35.

Ireland, H. 0. et al. (1910), "The Dynamic Penetration Test: A Stand-ard That Is Not Standardized," Geotechnique, Institution of Civil Engi-neers, London, Vcl 20, No. 2, pp 185-192.

Kolbuszewski, J. (1957), Discussion contribution, Proceedings of theFourth International Conference on Soil Mechanicb and Foundation En-gineering, London, Butterworths Scientific Publications, London, Vol 3,pp 126-128.

Melzer, K.-J. (1971), "Measuring Soil Properties in Vehicle MobilityResearch; Relative Density and ,Cone Penetration Resistance," Tech-nical Report No. 3-652, Report 4 (in preparation), U. S. Army EngineerWaterways Experiment Station, CE, Vicksburg, Mississippi.

Menzenbach, E. (1959), "Die Anwendbarkeit von Sonden zur Prtifungder Festigkeitseigenschaften des Baugrundes," Forschunusberichtedes Landes Nordrhein-Westfalen, No. 713, Westdeutscher Verlag,KMin.

Moretto, 0. (1963), Discussion contribution, Proceedings of the SecondPan American Conference on Soil Mechanics and Foundation Engineer-ing, Sao Paulo, Vol 2, p 533.

Muhs, H. (1969), "Neue Erkenntnisse giber die Tragfihigkeit vonflachgegrmindeten Fundamenten aus Grossversuchen und "hre Bedeutungfiir die Berechnung," Die Bautechnik, Wilhelm Ernst & Sohn, Berlin,Vol 46. p 18 1.

Rodin, S. (1961), "Expriences with Penetrometers, with ParticularReference to the Standard Penetration "'est," Proceedings of the FifthInternational Conference on Soil Mechanics and Foundation Engineer-ing, Paris, Dunod, Paris, Vol 1, pp 517-521.Schultze, E. and Melzer, K.-J. (1965), "The Determination of the Den-

sity and the Modulus of Compressibility of Non-Cohesive Soils bySoundings," Proceedings of the Sixth International Conference on SoilMechanics and Foundation Engineering, Montreal, University of'-ronto Press, Vol 1, pp 354-358.

Tavenas, F. et al. (1970), "Etude des sables submerges par echantil-lonnage non remame," Canadian Geotechnical Journal± National Ra-search Council of Canada, Ottawa, Vol 7, pp 37-53.

Terzaghi, K. (1925), Erdbaumechanik auf bodenphysikalischer Grund-lage, F. Deuticke, Leipzig.

U. S. Bureau of Reclamation (1953), "Second Progress Report of Re-search of the Penetration Resistance Method of Subsurface Explora-tion," Report No. EM-356. Design and Construction Division, EarthMaterials Laboratory, Denver, Colorado.

8