SCRIPPS INSTITUTION OF OCEANOGRAPHY UNIVERSITY OF CALIFORNIA
CALIBRATION OF THE EMERY SETTLING TUBE FOR SAND ANALYSIS
by D. M. Poole and W. S. Butcher
Work done in vart under ONR Contract N6ori-1il, As @ yee vie Beach Erosion Board Contract W-49-055-eng-3
Submarine Geology Report No. 9
April 1950
ec @ee¢chOO TOEO oO
MAO A
IOHM/TEIN
CALIBRATION OF THE EMERY SETTLING TUBE FOR SAND ANALYSIS
D. Me Poole and W. S. Butcher
CONTENTS
MUSE HOn MU oUTCSi eles aie ess elm iwle © \\s BAVIStGNS Cone let slive ie ehflsitteili fale (a. (entte. (ei ll orinallie PNGO CIC tL OR itsitile allio tall te) el im lei isi ay ie)! le) 18 Method of Investigation . . . »- » + « « « » ING CUCU Va ilar ay hep eilan) Cain vce se) lier allliey | (iw!) eile) Iriel are Settiine ube: Brrorl) (esl 6 Js «ss 6 SUC UE ITO s lial iatilivilslay ll aillieillleilitin| ie Effect of Samole Weight .. » « «+» =o
Influence of Particles Coarser than 1 mm. OnMmuner Medraneh ramen cin ciel eiiley ellis iie iis
Influence of Particles Finer than 1/16 mm. on the Median Diameter « » - « » « « «
PPO CSOUME Sie ie) wal le wie fienbiel) te! ver viel) /elinis
Splitting the Samole ......«+
Method of Introducing Sample into Tube .
Reading Height of Sand... .. +. - SUMMA ieee tests epitomize. eile) val Nie, bshinen he Acknowledgments . « 6 + « « «© © «© se © + 6
TWIGnr Shetokarslse MeO EC OOH CH CNC oe oan
Page
Co I nko ASS FEF WW
‘ | Celibration of the -2- Dee Me Poole jac:
Fnery Settling Tube
W. S. Butcher
for Sand fnalysis
LIST OF FIGURES (all figures at end of paper)
1. Sample data sheet for settling tube analyses.
2. Ten settling tube analyses of the same sample.
- Six sieve analyses of the same sample.
3 he Settling tube analyses of sixteen portions of the same sand sample. 5
. Splitting procedure in obtaining sixteen portions of a single bulk
samole by comb
Ge Ae
b.
5j ning alternate quarters.
é i
a)
Settling tube analyses of eight portions of seme sand by combining elternate quarters,
Settling tube enelyses of eight portions of same sand, alternate
quarters not combined.
7. Settling tube enalyses of four portions of same sand split to different weights,
8. Comparative anclyses of same sample with snd without material coarser then 1 ma.
Ss) he be 10. a. b.
Percentage veriation of the median diameter for a given percentage of fine meterial.
Percentage variation of sorting for a given percentage of fine material.
Conparetive enalyses of same sample with and without fines (material < 1/16 mm = 7.7%).
Comperative analyses of same sample with anc without fines (material < 1/16 mm = 22.5%).
Abstract
The accuracy of the Emery Settling Tube for the analysis of sand particles has been investigated. As pointed out by Emery, this method is more rapid than dry sieving and gives equivalent, or settling, diameters rather then geometric
diameters.
It was felt thet a more exact knowledge of the errors end limita-
tions of the method would be valuable.
ae elgats 3 Me moc eR \ cue setod tap ein .
fe
Caegee ae amse to anc.irieg ‘ithe to aseubsas dust gabldsod Part oy | Biase to serial
i. Inisstan tuoddiw bes Atty efquse enae te wow avid
tt
gyetasoieg devig 3 tot todemoib usihac ; tlt to aod
b colt ere MLS ‘Ad Ee foe: $2 outa to Bory, RSS = an Ay
ik foattada a
aed
asfotdisg basa to elaylane edd cot odoT aniissoa bhogeae una te. “pss 4 cats Bigqox otom ah bolton etdd ,yroml yd Jao “badmiog a nixtemosg gait sector etodomiks antl, tow ‘te tee Lavoe cee ‘howe ~od iotif bas gros eid to oghalwondt dtonxe stom Bp talt tfot gaw dL pe saidariav ed SLoow sissies ho
Calibration of the -~3-= D. M. Poole and Emery Settling Tube W. S. Butcher for Sand Analysis
It was found that the settling tube anelyses for materiel between 0.062 end 1 im. had a reproducibility or probable error in medien diameter of 0.8%. For the seme sand the sieving probable error was found to be 0.7%. The method is thus approximately as accurate <«s sieving. The errors occurring during splitting of the sample to the proper size were investigeted by several proce— cures, but the results are not conclusive. The maximum splitting error was 6.2%. The effect of material finer than 0.062 mm. in the sample wes investi- geted end it was found thet no significant difference wes produced where the fine material wes 5% or less of the total. The effect of material cocrser than 1 mn. in the sample was also investigated end it was determined thet ell coarse material should be removed before enelyzing.
A recomaended procedure to be followed in making such en enalysis is in- cluded.
Introduction
Emery (1938) described « repid and accurate instrument for the mechanical analysis of material of sand size. In the original peper there was not suffi- cient information to indicate the accuracy of the method nor procedure to be followed in making an enalysis. This paper further confirms the reproduci- bility of the results obtained from the settling tube, the close correlation with sieve analysis, and gives a detailed recommended procedure. The equive— lent diameters obtained by this method would appeer to be more indicative of erosional and depositional features than those obteined by sieving, perticu- larly where there is a high percentage of micaceous or plety meteriel. The time saved by this method over dry-sieving is of great adventage where there ere numerous samples to be analyzed.
The Emery Settling Tube is essentially e plass tube of 21 im. inside diameter end 164 cm. length. At the bottom the tube narrows to 7 mm. inside’ diameter end is closed with a stopcock. The narrow portion of the tube above the stopcock is engraved with milliliter divisions on which to read the cumu-—
lative heights of sediment. Emery (1938) gives a figure of the settling tube
oe vi LLtdtosborger hodtem ed? EV. od io bio? 2aw torts s otdodeny Baith uabwmuosc atorie oT .galvele 22 ots -sovtd Loravea YW Rod syiteovat sraw este seqorg: dt Gow torte gattdliqs mumixu: of? |,ovkentonoy ton | gtaovek enw efgmea odd mk «tat S000 goat r94,t2 1 ott exedw besubeswg eo eodorsttib dnaskiiogke on Y / qgansa6’ Lobted sa ‘to ‘toatic of? gldtot odt to abel aa Re: ffs garit banlurstoh anw tL Soa bedagitesvni oats esw efqmas oft . | egntayl, ns omcted hewonet ad Shvone
Pin oe al aleylons na dove oufalas at hewakio’ od a ee bs
: “{oolnadsom sat tot iegiretedt afsiipos bins ba ty 3 baltroeeb eres) -ittwe ton enw otedt deqay tsakgito oft oI eesfia base to Lawes sx to ad od stuheso1y ton boillten sdt to yostupss sit atuotbal of woldasr102 ae odd anor tins toddaut, veqag. a Lat setayfsas aa paidtom at
yp a agols eld dint suitlotee aif By beadaddo et toaon me -—s¥tupo off sotubasorg bebromnoset beflatab 6 any hres eievtaan
to evitsofbat svat od ot taeqqs sama batitan ality yd sain cinch aba sir
mug die 3q «anivela yd healsido eo, Ad nodd aortwtso" thins tdeoash hiss t
6dT «laivedsm ye ahi TS avosonoki to sy atuooweg Ay ld get stadt roth
eteht aii wgsdesbs toon, Yo ef ic a hove sith nbat e be ebientk «ti [S to odut each) 2 bya tiuonus HL ocet ganas acts oat avods sdb odd Yo -cobtteq wort ott
aumio ont adet of dokiw mo, enotadvth tod 1 saaeaiil dw houega of stooone i ‘
_ edu s “eithrde 4 git to swykt es sovls (weed) cient Ldnoakhas ‘te adits od bgisan
Calibretion of the -4- D. M. Poole and Emery Settling Tube W. S. Butcher for Sand Analysis
used by him; the one in present use is similer.
The aim of the investigetion wes to determine the probable errors inher-— rent in the method, to determine the splitting errors encountered in preparing the sample, to evaluate the effects of particles greater than 1 mm. end less than 1/16 mm. in diameter, to evaluate the effect of the weight of the sample used, anc to compare the settling tube analysis with that obtained using Sede Sieves. These effects are discussed separately in the following sec- tions. A recommended procedure to be followed in making en analysis by this method is given at the enc of the paper and is based on experience gained in
using the settling tube end the results of this investigation. Method of Investigetion
The Emery Settling Tube was used, in all the tests, in the manner outlined in the section entitled "Procedure." Briefly, this consists of splitting the bulk sample to 3.5 - 4.5 grams, introducing this small sample into the Fmery settling tube, and reading cumulative heights at times corresponding to the settling time for a given size material in distilled water at the observation temperature (see sample data sheet, fig. 1). For purposes of comparison, a sieve analysis was made in certain cases using the Tyler Standard Sereen Series. The shaking time for the sieve analysis on a mechanicel shaker was 10 minutes.
The cumulative volume percentage for each grade of the Emery settling tube and the cumulative weight percentage for each grade of the sieve analysis were plotted on logerithmic probability paper. From such a plot the median diameter (50 percentile) wes read. Where required for comparison with other tests, the
standard deviation of the median (s = (Dev. of Md \ wes obteined anc from sal
es a RECS aywris | sash ott aired os “ seol Sas mitt £ asdt tetaoks eatoltaag, to atootte dt etauLove 0
oignae. eae % Sigtow old ‘to soeTte ont otadhove Oe: ptotenaib ak sm gadgur bentatdo tadt dtiw eteyfoas odut pablssoa edt arsgne 9 ot : mop gaiwolfol odd nf yletotsgea beaasoetd 8 atoatts oasiT seoveds, | pt aids yi eleylsas oo pation ni hewollo? sd ot omuoootg hebroumoooe A ; . at Seaksg sonelregxs mo bsasd ak be ‘taqay ort te bao ont ta new by a Oe ae - stobtagitaovad aidt to ativeot odd bas sdédt uttrios «
sealeul ios
add seitttige to atetesos eit wlio le * otrhogorg* boititae ned Receives edd ada cleus Lowe elit qntoubowtat er Qa = Gk of ead oF pithnoqaereel gomid tn -edigiod avidalnao gaibaot bis osu
aoisorscado odd fe tadsw belliteth ot Lsineden saie movig 3 10? ord
a7 »eotuain Of aaw. todedg Lnoiacdoom » no aleylecs ovete att SOT oad ‘aie
odud yotittes yrsmd edd to obota done tol egatueotag opunlov ovikts fumwo ost
totous ib assets et tolq 2 dose nord stedsq vw ilidedorg git bekake 19 mine
edt qaveut wadto dtiw noaitsquioo tot betinpos atedW «bost esw (elétaaotag oa)
Wee. . ae Sas heatatds asw ( mea ely = ») aus Hav eit to auitaiveb fasbe 238
Calibration of the -5- D. M. Poole and Fmery Settling Tube W. S. Butcher for Sand Analysis
this the probable error (P.E. = 0.6745 o ). For further comparison between different tests it is necessary to have the probable error expressed non—- dimensionally. The non-dimensional form was obtained by expressing the proba- ble error as a percentage of the median diameter.
As shown by Krumbein (1934), the probable error may be separ- ated into any number of component errors. This separation is shown by the re- lationship: E= Veer Bless ere Since the probeble errors are ex- pressed in per cent (see above), they are non-dimensional and apply to any test. The total error (E) in the experiments anelyzed in this paper corres- ponds to the "leboretory error" of Krumbein's paper. The "sampling error" of the cited paper has been eliminated by using one bulk sample split into the desired number of portions. The total error (E) hes been divided into split- ting error and settling tube error.
Settling Tube Error: If the same sample is run through the settling tube a number of times, the probable error of these runs is due to the errors in running the sample through the tube, observational errors, and errors in timing. These may be considered as the error of the settling tube itself or of the method, since splitting and weight-of-sample errors co not enter. Hence, the same sample was run through the settling tube 10 times and the results gave & pereentage probable error of 0.8. For comparison the seme semple wes sieved 6 times, using a different split of the sand tested in the settling tube. Here the percentage probable error was 0.7, end thus is approximately the same as that obteined in the settling tube. It should be noted that the third quartile shows a greater spreed for the sieve analysis then for the settling tube (see figs. 2 and 3). The median diameters from the settling tube averege
0.131 mm. end from sieving, 0.136 ma. Although the sand was a reasonebly clean,
ne a oat @ owoie sf solfotsqes elaT ators Saoaogas to ted a sodas i in “9 ot azote aid adorg odd aonkG . ee $0, bie Soo" + oy = a
neowtad jr) pe tof 4 ( D 8AT9,0 = aa) ae hie ot
anh
mfiO% heaaoraxs Torre eral: said evod oF ot ih st iy ig $
i sal of Mp bas Lonofanemtb-aen 913 yrs: yee 992) tnso Tog ab
ont atat tifce eLyaise Wid eno gaiew yd betentuife mead esd wysq hed } nthige otek bebivts seed aad (4) rowre Lstot oAP .enodttog to tedium & :
sdut gaifttes edt dguorad nut at seinen anae ott 21 seciuae tal fag ators ath of eu) st aay asedd Jo tore eldodosy odd (aenl¥ to 1
sme aorta bas etorze fsantdavicade .sdus art rauserecit ‘edqmae’ add 3 10 tLoatl -adist iatissos ad3 to torta edd as bersblasco sd you saad?
2onsH «tedne ton ob eroTte sicmia~bo-ddaiew fis gaidtilga sonics Hodtox
aaw elqnss orice ef? soatragmos tol 28.0 to torre oldadorty ‘euedtacotsq a
dud snifttes add ot boteat hace ot to dilge tnowtits ¢ gnieu .aomit o be
orsa af? ylotomixonyqs ef ait bus .V.0 asw torre sidsdotq epiinsotag gid cary & freidt oct Jadt Sodom od bieoda ¢] sadist galldsor ond at ‘nsnistds todd es gaifties sit tot madd ateyluas aveta alt 167 Seotge radsoug & ewode ofidarasro
ex vievs edudt gniltdee odd mot? atetemakh nstbem odf «(€ bas S pags? 9aa) odut
eisels yldancanst 3 eaw basa odd dyvodttA ox O€L.0 .ynivele cotk bas .aumt LEL.0
Calibration of the -6- D. M. Poole and
Emery Settling Tube W. S. Butcher
for Sand Analysis
round sand from a drifting dune near Yuma, Arizona, it is likely that this dif- ference is a difference in the actual diameters as comared to the equivalent, or settling diameters.
Ludwick (1948) collected data following Krumbein's (1934) procedure on several southern California beaches, and analyzed them using a composite samole of 8 to compute the coefficient of Hs et Werke From these data the percentage orobable error in median diameter due to laboratory error (the total error of this papor) can be found. The values range from 0.5% to 1.59 with an average of 5 beaches giving 1.0%. The median diameters used in the tests in this paper are smaller than those of Ludwick!s work which ranged from 0.189 to 0.400 mm.
The total probable error from Ludwick's data is generally smallor than that found in our work. His data show no consistent relation between total probable error and median diameter of the composite samolc. The scttling tube error mist be less than or cqual to the total error since the total error is a combination of splitting and scttling tube errors. It is probable that there is always some splitting crror but its amount depends on the homogeneity of the sand. Ludwick'!s composite samole contained sands from the area covered by his grid on the beach. The different sands varied little in median diameter and sorting and thus a composite sample would be reasonably homogeneous. Conse- quently his total error is small, probably because of a small splitting error. The sands used in the tests of solitting error in this vaver were considerably different in median diameter and sorting. The composite sample will thus be less homogeneous and more likely to have a greater splitting error than the more homogeneous mixture. Further, it is difficult to see why the settling tube error should vary significantly when the settling tube is used with care by experienced versonnel. The error due to the settling tube from Ludwick!'s
data is thus assumed to be of the same order of magnitude as was found here;
nea _* Sas ofoot re a. ; Sanam uae
A
Nie
sousd atde ak Sauer ont a boat Brot oustd asta at Woes setts nad? tollens yilsronoyn nt Go also tied nent ToOITg aidedore’ fatot into nsowtod noltsfot Shope tedied ou woda ofab al eitov to ra odit gai itt: od: .ofumee otfeommes oft Ye rotometh asthon brs m: & ek novo Lstot off oomte torre Ledod od ot tie "2 utd seol od bie lea dats oldadorg at 3l .etotro odgt snifdton faa gnittiicge to not nett ‘to Blonegon of sdf mo ebrocoh davone ett ted torr anti? ce oaoe | aid. vd horaves sein eft mott ehane fantetnos aferaa ottsocaes ohio teint Situs metenath acibam at ofeehr Boteav ehose tnetsttib addl® wines ent: “62000 sewoeston aod yideroneet od blow, sdk ie iain ie a aud? Ans etorrs anietifoa L[iame 8 To sessed videdory Line el iov19 fatod abt vidaersblanos sre recent eiat x2 torte wtitditoe re) ay aod oct ak beeg. wags
ed aud? Ifiw ofcmse wt Leogron oft esiiivioe fs notonsth asiban m4 taoae
azss déiv fever ei add aniiites rid nety ultaasttiagie orev Leon torte 5 ade ‘
vray
6 'oo twhsal eer odet antlidse ost ot orb tonse off .fannosted beonetrecxs wf
yetsd bret sav on ebrdingan To nobt0 omea ent ‘Lo of of bomvean emft 2f stah
Calibration of the -72- D. M. Poole and Emery Settling Tube W. S. Butcher for Sand Analysis
Iter y (ede.
Solitting Error: In running a sample split in 16 different portions (see fig. 5), the splitting error will be given by extracting the square root of the error of the settling tube from the total error of the test. In this case the total probable error was 3.9% and the solitting error 3.8%. Figure 4 is a gravh of the cumulative frequency curves of the 16 samples. Ina similar series of samples (not shown) where the splitting procedure shown in figure 5 was not followed, the splitting error was 6.2%.
Because of the large discrepancy in the value of the splitting error in the above two samples, the svlitting methods were checked again. Sight splits were taken from a sample by combining alternate quarters as shown in figure 5, and eight splits from the same sample without combining. The splitting error in the first case (combined) was 1.0%, and in the second case 0.8%. The graphs of these two samples are shown in figures 6a and 6b. Further tests with 32 combined and 30 non-combined samples (not shown) gave splitting errors of 2.4% and 1.8% resvectively. The sand used for these tests had a larger median diameter and was more nearly homogeneous than the sand used in the other testse It seems obvious from the discrepancy between the series of tests that the splitting error has not yet been fully investigated.
The lacl: of correlation between splitting method and splitting error is probably due to insufficient data. As was pointed out in the comparison of Ludwick's work and the results of this vaver, the svlitting error probably de- vends in part on the degree of homogeneity of the sand. The combination sand used in these tests is a non-homogeneous mixture and a larger splitting error would be expected than in a normal beach sand. Since the probable error is a
measure of the variability of a series of tests, we would also expect a greater
— vnoist0¢ tagasthts OLat otfeg efone 2 a pabaaw: P
te toon ezaupe eta gattontizxe wf nevis ed | ithe , is Pore edite at ataes edt. 0 notte fater oat moet, eins | s at it oxi ba owns oats tie ont — bese em watG satiate ripee vookemae at edd 0 worn pment 9 av atone yi vie ak meoda | stuhonoe gtltttiqs add evady (awoits tox) as:
08.3 ot torte gohtt tee oid bovoLfot
ni rove gatttiles att 0 eurisy eci9 ak vorneetsotS, satel odd to 9 otto. sieht etlans beslneds stew abortem anitilios od¢ eeLomne- owd fia out? ai moda eg sretisep stamteatis aninidaon yd sfomsa aan roves ates bige ot spatntdaoe trrodt iv atomae anse odd not aftiga sit «8800 5829 faoee edt ai bar ,SOet sav ‘Bont derog) geno +2 atest aortd uf «2 fue sd eowrsit at avade evs selomse ows seedd % eTorrs Salk iten OED, (aworla tor) pelonna hontdmoomion OF Bas beatdea | aaa s bed etast dows “ot Heer hase ad? wlovitoeqsen 38.2 bas
to neinee adit noveted "ans sgeTDSsh antt met? avofedo-aveea tL sets
- sbotesiteeve! whut tend tay ton eat totte anette orld tid 4 teat to ston fgraros ond aL tuo hadator arr ad waitab dueksttinven! of spt sors guittifqe tegisl « fas omdixio avosnononodt-non & al steed aactlt al eer
8 al rota oldsdore said gonke hiss toned) temton s md. adh betoomes ad bisrow
. Satna, = Josexs oals Sr ov yeteot To asiissa e to vtitidalvey od te ottraon:
Calibration of the -8- D. M. Poole and Emery Settling Tube W. S. Butcher for Sand Analysis
difference in the individual splitting errors from the mean of the serics. Probably more completc data would have shown some correlation between eure ting method and splitting crror. It is felt that the homogeneity of most sands and the time saved by the less comolicated splitting procedure obviate the necessity for the use of the procedure of combining alternate quarters.
All the samoles (except as noted) used in determining the svlitting error were made uo from a mixture of sands from the beaches around La Jolla, Cali- fornia. One-third of the sand was from Cove Beach (Md 0.7 mm.), one-third from Windansea Beach (Md 0-35 mm.), and one-third from Scripps Beach (Md 0.17 mn.).
Effect of Sample Weight: It was thought that some significant error might be introduced in the settling tube analysis if the weight of the material were not the same in cach case. The error might come from the increase in density of the medium and from the increased tendency to advection currents with increased material. From Owens! (1911) data, it can be estimated that 5 grams of material in the 515 grams of water in the settling tube will affect the settling time by about 205% due to increase in the density of the medium, if the sand is considered to be in solution. As the sand is obviously not in solution, the error introduced must be considerably reduced and probably can be neglected. Calculations based on the formulas of Rubey (1933) indicate the same order of magnitude for the error introduced by increase of density of the medium. It is of advantage to have the sample as large as vossible within the capacity of the settling tube, because a large sample gives a greater change in cumulative height for e given volume percentage of the total samole. Most of the error introduced should be due to advection
currents rather than to increase in density.
) : Caines ot te sos ort? nom
wrote ofanrod La, a % - omshosorg mt to on ot x0
7 orto sabta tice ost rialarotos ite boew (bovon, as $eeox0) eofonna ode im a hy bled ci al Aegon: aodonod od mort phat bs ountzta 3 Snort cn ob
a ‘torte freoltinaia smes tadt ve ‘pew " ; dnt ee ont * to ddgiow of? tk eteyfane odist ant lito att ab pie
at asroron odd mort onod. dagin torte ot 698ND ono al omne ott
simorsero notaagie of Nomobnat fossoronl add mort ban suck pom odd to
ae tet bodsattes od ase tf estsh (rer) “‘onowd mort. “obroten Soa iooxas [liv odnit Wetplos att int istaw *o amenm Bf2 ott at ‘pelea tna 3 ett har: Bs) to yudtenoh’ oft al sanotont ot ob 02.8 trode vd salt ents ak toa ylewotvdo at fees oft ad anntaetoe at ng ot howabianon al Bre fs “ldodote fas booubes wider bleaas se 4 eas honubatts «i totto one sto ff
snothnt (feor) vouus to nalomot add no hoesd anotisivolat ahots
8 oot. eisie agtal Saas pod sritites : of to nee Enc “) add ot “9 earinootog onan ov sorts GM de defy Loe! od atansro: at one z9HA9 toltoovbs of oh od bivarle boobed! vows att to dealt " olqure tnsod bi
eviiecos af oanovant of nadt conten ataoaiw
Calibration of the -9- D. M- Poole and Enery Settling Tube W. S. Butcher for Sand Analysis
Four splits of the same sand were run, each having a different weight, 2, 3, 4, and 5 grams. The total probable error of the median of these runs was only 06h. Since the tube error is 0.8%, it can be stated that there is no appreciable effect caused by differences in weight of the sample used.
Figure 7 shows the results of these runs.
Influence of Particles Coarser Than 1 mn.
on the Median Diameter
To determine the effect of varticles coarser than 1 mm. on the median diameter of the sample, two splits of the same beach sand were prepared, one having the particles greater than 1 mm. removed. Figure 8 shows a plot on logarithnic probability paper of the two runs and also a sieve analysis of the same sample for comparison. To make the curves strictly similar, the weight percentage greater than 1 mm. has been added (at the 1 ma. grade) to the samole having the coarser material removed.
It is to be noted that the samole containing the sand particles coarser than 1 mm. shows, with few exceptions, larger diameters for the same percen- tage of the total. This difference can be interpreted as a result of the carrying down of the finer particles with the coarser. The difference in the median diameter of the sample having the fraction coarser than 1 mnme_ removed and the sieve analysis can be attributed to the entraining of the finer varticles by the coarser in the tube analysis. In addition, the differences are due to the difficulty of accurately reading the scale divi-
sions on the settling tube when the suspension-sand interface changes rapidly,
setbor:
in tow Briss dosod onset att to ofca Ho
u A
to tole # ewoda 8 ghee: ehovoiret eam £ at tot. 3 vofodtang Ey
lad i cohen, ent £ ond dis) ‘bobbs good act ostnt if ane ators pus agri a ta aod ea soaneo ode aattvid
TOS TS0. eofoltsse hase odd ac inbedston ofmag ott todd Bot on od of moto aging ont rot ashore nogun at .snatdqeons wot, ftty “seeoda son oft “to tison v er * hatoteratatl od man anamin tain elit alates add %
MS oonoretith off .toatsos odd Ad iy eolnttene oni att te avob Sf. v wit t srastd ash koltonr® odd anivad’ o Corse an to stone th aot
orf? $0) surdadneders ote od fod.ire Tatts Od f.88 5 ehaubane pa tt ine havo
ot ,toltiapa nl .eteylem odart td wi toessen, ott vd nolo tenn x a
~fribh ofsee’ ott
na
} : Ate ye an sitive: wLotaunoos %o cet Ler LYBLE ocd ot ovb ote scenorettih
Whiowr aogmado soAtiotal Aanemtolaaocare nett mothe ogi giiftdoe oft mo anole
Calibration of the ~10- D. M. Poole and Enery Settling Tube W. S. Butcher for Sand Analysis
and to the total error. The difference between the cumulative curves of the sieve and settling tube analyses for the sample having the coarse fraction removed, is within the limit of error of the method.
The conclusion reached in this test of coarse material is that the varticles coarser than 1 mn. should be sieved out before attempting a sottling
tube analysis in order to obtain accurate results.
Influence of Particles Finer than 1/16 Te
on the Median Dianeter
The influence of fine material (less than 1/16 mm.) in the sample run through the settling tube has been investigated by a series of test samples containing varying amounts of fine material. A graph of the percentage variation of the median diameter for a given percentage of fine material and a graph of the percentage variation of sorting for a given percentage of fine material are shown in figures 9a and 9b resvectively. The scatter of the points on these graphs is so great that it does not seen vrofitable to draw a best-fit curve. It seems probable that if fine material in the samole exceeds a total of 54, the error in the cumulative curve due to the presence of the fine material may be greater than the total error. It is therefore advisable to sieve off all material finer than 1/16 mm. unless the amount of such material is less than avproxinmately 5% of the total. An example of the sinilarity of runs with a small vercentage of fine material is given in figure 10a, and of the dissinilarity with a lerge percentage of fine material
in figure 10be
be tonne ott atk ( emat: ar\r reat eeor) initodan ott to onmnertttt aolcnne Jeot to aoltsa s wd hodeghteownt good endl adit gatiddoa on SAF oo TOC one to dort A sLekvotan onlt to edavons gateier Bas Tabvotioe “baht to onatnosiog novis, A 0% tavomahd bation ont ‘to te Sait to i aici sovis s 20% naitdtos to seas sastacrn od | _ ont to sodtnoe eft. loriisecsoy de far ae BOTS, it at pee weeth ot pleat itor goa # oft agub de feds trots, as ve to oe ‘ogee odd at Lebrogan ante tt tnct9 ‘ibaltenés see one a2 lowe Sotesste odd of os orto ovttatimuro: ahh XE torts ost one to intod oF orotorads ai #1 ror tador old hue totnety of ee Lateoden: ont to deubnaate odd avofmr son SA on a wel sontt Le ftotec: ifs tho ovata of S alt to nlenare tt alndod odd to rte Ylotac hrotonn edd eo! al tition nt rove, et Lalmegiec see 20 sgavnaogtod ina £ she act To yd tn Istrotar ait te oystasetod optal a Ad be ‘dencuensenty old <0 bam AOL, jeri §
efOk oneal at
Calibration of the -ill- D. M. Poole and Emery Settling Tube W. S. Butcher for Sand Analysis :
Procedure
The preparation of somples for enalysis is discussed et length in Krun- bein and Pettijohn (1938). The only procedures considered in this paper are the preparation of samples for analysis by means of the settling tube, and the method for running the samples through the tube.
Splitting the Semple: The sample efter disaggregation is first passed through a 1 am. sieve to remove all particles greater than 1 mm. The percent by weight of the sample greeter than 1 mm. can then be calculatec. If there is more than 5% material less then 1/16 mm. in the sample, it should be re- moved by wet sieving. Then the percent by weight of the sample less than 1/16 tau. can be determined.
The sample is next split to a weight of approximately 3.5 - 4.5 grams for the settling tube analysis. A Jones type sample splitter was used to split the sample down to a weight of about 20 - 25 grams. The "Otto Microsplit" was used to split the semple further to the correct weight for analysis (see above) by the settling tube. This chenge of splitters is merely a matter of conven- ience in handling the sample. Tests showed that the type of splitter used, introduced no appreciable error in the analysis.
Method of Introducing Sample into Tube: The method has been scmewhat mod— ified from that recommended by Emery (1938). A centrifuge tube (2.75 x 13.5 cme) with its bottom cut off is used as an introducing tube. The bottom of the tube is closed by the thumb, the sand poured in and distilled water added by means of a wash bottle so as to remove the grains sticking to the sides of the introducing tube and to cover the sand about 3/4, of en inch, The sand is
stirred thoroughly until no bubbles remain, and any perticles floating on the
“ a ots £) eold Yodeorty avlodsrny La: Vvoret of ov - otodkt QI .betoluotse od asdé avo stu f aod neds0%§ oleae eet 90, fg: wot od biyode th yolquan edd at om an\r godt aod sitesi a ont fads seol oLjmae ant to titan wi dneotey ost aos?
aot @eery tod ~ 2.8 ylotauixonygs ‘wo digtow & ae ie. exon, ab ofgen | thie ot bows aow sattifge elymee erg seavb: AU sataylons, edit we sow "Silgeotold of SO" dT .amotg gh + O8 tuode to tog tow Sot mwo's (avods 68) BLlaylans tot ddedow tosttoo odt of teriieu't kine ont snevito® fo 43¢fort + ylotan af avett ily te semtaito eit vod boas eee to ogyd odd chackt | parents aveot -otanse oy atom ‘A
7 -ateyions end ni mie, ‘alsotoangaa Laat
“hom totwemoa noe d end Dootlt se oAT ‘ai ot
(BeEL * ONS) edit exntiantaeo A (Beek) ian 5 Sebneassoat toate sont
yi Sebbs tetew belfiteath bas at ute Onise ont gditehr att vs borols ak ¢ ! oat to eobks odd od yatdolte aed Om of} ovonet of aa og elitod dlasw i) to som j af-Saze odT wigs ms ‘to Ne duods base oft tevos of baa slice yabsubordad
ont me genbss oft apLoidasg yrs das yotunot ge fi dete on Lede hay ve caret
Calibretion of the -12- D. M. Poole and Emery Settling Tube W. S. Butcher for Send Anelysis
surface cen be made to sink by touching them with a wooden pencil. The sand can then be released into the tube. Care must be teken when introducing very fine sand that large density currents do not form. This may be accomplished by slightly tilting the introducing tube and allowing a portion of the sand to enter the settling tube slowly, followed immediately by the bulk of the sample. Tapping the upper pert of the tube will help breek up any density currents thet form.
Severel runs were made with very fine sand samples using disaggregating agents, sodium oxalate (Nag020,) and sodium hexametaphosphate (Nag(PO3)¢)> to wet the sand before introduction into the tube. The results were not signifi- cantly different from splits of the same samples which were wet with distilled water.
It was noted at times that the cistilled water added to the tube contained fine bubbles. This bubble formation occurred when the distilled water supply was low enough to cause somewhat intermittent flow. Flocculation, by adcsorp-— tion tc the bubbles, occurs when the bubbles are quite smell end numerous. Since such adsorption makes the analysis erroneous, it is advisable thet no runs be made while such bubbles exist.
The temperature of the distilled weter is measured by running the water from the outlet through a bottle containing « thermometer and then into the settling tube. At present, the water is led in end out through two holes in a cork fitting a smell wide-mouthed bottle. The thermometer is held in e third hole so that its bulb is bathed by the flowing water.
Reading the Height of Sand: The stopcock stem, gradueted in milliliters, is not tapped as recommended by Emery (1938). Compaction was sel om observed
in the sand column if the send particles greater than 1 mm, were removed, but
hes hae M ia : pseu ce al
Wis
yas _Doittgasaos me gon alist i seek ton ob etasrtee oot ogtat
~Hiiagin ton stow istaicie ah eoded ott oni aobboubontet sie soe ; peliiéaih Atiw dow stow doinw aolqase emsa odd ‘to atiiqe nite tiers Ttbs
elyuue totow bofkiteio edd nedw bertuaol° noitsertoi elddud aid? »ealdd | maton “ws mone i soll Sasddhetstnt Sadwscoe 93039 eats
e ot selon owt dquotdd tuo fas at bel ek wadaw ont ytaspetg tA adit 40 tind dock Sod gk totomegrodt adT saktdod bedtvomobiw Liace a gate? 3 etotow gaiwolt sit yl boris ef diud eth dodt oe of
sATOT ELLE Lb: ab Sodowbary yoete aoooqods oi? ; 2 Jdutel yly apd!
bovrerdo mo Hoe gow noldosqnad (BEL) vreau yd bobnesmooet 25 heggat Fou ok
dud) ybavomot tow «it L asd? te¢se%y aeloitiay hase ‘edt tt mofoo base oft ak
Cz libration of the -13- D. M. Poole and Emery Settling Tube W. S. Butcher for Sand Analysis
with coarse sands some slumping mey occur. The inaccuracies in reading the sand height due to the rapidly changing interface, plus the error due to siump- ing, indicate thet the settling tube should not be used in analyzing samples
containing an appreciable amount of coarse sand.
Summary
The time saved making mechanical enalyses of sand by meens of the settling tube contrasted to sieving hes already been pointed out by Emery (1938). This paper shows that the probable error of anelysis is about the same for both methods, if the perticles greater than 1 mm. and less than 1/16 mm. (if more then 5% of total sample) are removed before running a sample through the set~ tling tube. The settling tube analyses have also been shown to be reproducible with a percentege probable error of 0.8 (settling tube error). The error due to splitting the sample to the correct size (splitting error) for use in the settling tube is undetermined at present. The Jones type splitter may be used alone, or the "Otto Microsplit" may be used efter the semple has been split to
a weight of about 20 - 25 grams.
Acknowledgments
Acknowledgment is made to Mr. D. L. Inman for his constructive suggestions during the course of the investigation and his criticel reading of the menu- script. Dr. F. P. Shepard kindly read and suggested improvements in the manu- Script. Mr. J. C. Ludwick, Jr., allowed the use of unpublished date collected by him and discussed the problem constructively with the writers. Mr. D. B. Say-
ner is to be thanked for drafting the severel diagrams.
add anthedt az aliens with “sH0090 yom gage Hie 08 ob sors ond uy wooxtradnt gaigasdo ylbiq a - roe gpiyinns at hoes od ae cine
ae | yabiten oat to ensom sa oe base to 5 nomgcans tno asian gata on
elt? .(8c08) tanh yd duo ‘hodantion good) vbrorls’ nad a of : tidod ‘yo art aid tyods ai a Yo torte oldodory okt Sada atom $i) oom OL\L nods easel bas one L asd totsetg aolobinsg oat u “tea edt douvords olqoae 2 animurt oto'ted bovomst ats (oLquse Latat to aldissLetqat ad ot awoda seed oels ovad aoaytons sdiv gniliton ad. vod oub ‘sorrs off .(t0Tr9e odut aildtoe) 8.0 to torre oldadorg ogsdnaditeg .
of ot eu tot (torts ymittilga) cede toortos edt of efiqmae ald syeetd hous od yor tottilge eqyh senot onl wtneasrd ts beaiotetenau ad ‘edist | ot iige noad asd elgmne edd tedte boaw od yam "tilquots iM a3d0" ond x6 : spimaTy 2S + OS duoda 10
at penmabie lwamiol, oh
ano tsteoagie evitouttance eid tot micml wl 6G otM of aban et sasmmbelwomisa | ’ mfioe sit to gathsot dovitito ei ban rotioaiteovat odd to cetion cit git
~uned ads ot edersmovorgmt botsaoygsea Sas bso ghntai braqod2 oh oh ot | Jookioe stab bedakftuyas to vas odd bowolLs yoxl yodwbial . .L la sdqiton
yok «H «@ .tM ,arediow odd dthw ylovitowstenos mpiiforea add beaavoalh bat. is eo
enmatasib Loxsvea edt giidis wah tol powliunas ad of ak tan
: “
Calibration of the -14- D. M. Poole and Fnery Settling Tube W. S. Butcher for Send Anelysis
i.
4e
De
References
Emery, K. 0. (1938). Rapid Method of Mechanical Analysis of Send, Jour. Sed. Petrol., Vol 8, pp. 105-111.
Krumbein, W. C. (1934). The Probable Error of Sempling Sediments for Mechenical Analysis, Am. Jour. Sci., 5th Series, Vol. 27, pp. 204- 214. :
Krumbein, W. C. and F. J. Pettijohn (1938). Manual of Sedimentary Petrog- raphy, Appleton-Century Co., Inc., New York.
Ludwick, J. C., Jr. (1948). Unpublished Data on File at Scripps Institu- tion of Oceanography, La Jolla, California.
Owens, J. S. (1911). Experiments on the Settlement of Solids in Water, Geog. Jour., Vol. 37, pp. 59-79.
Rubey, W. W. (1933). Settling Velocities of Gravel, Sand, and Silt Par- ticles, Am. Jour. Sci., 5th Series, Vol. 25, pp. 325-338.
aie such os © stot Cotestonl to bat a.) me: stant ginifgns® to a | “SOS TS 4LoV oe age
ae ‘ Luge . (8e62) nila Lt8d0% " : ltoY wot eda 00D yrs
| nv ion t aasi2 $9 of! no ated beads ss esiatotitsd, aaa
arr ak abifo® to toomeftied odd 9 Btran, x ae ie 99 ‘ =
ege GES bas .pnse fava to webhipabey atidsee A Geeer) a BREESE G9 2h -LoV epattas age ede 2ttol oth <
Sample No.
Wt. of | Dish & | Wt. of t Cum. | Total Dish Sample Sample Wt. p Cum. 7
Faas
in| e
| 4x | Total Withee + Cum. ie '
CE Mii 24
r
1/32| .03125 | 10 45 OWS Se atten LO
Total
Figs 1. Sample Data Sheet for Settling Tube Analyses
pene os
F ilaben | xX ]adeeate esaue| (002 = mili) | A wane >t bs Bilan 999
med Ne i em
rhemenne rates
| H : ie SRE! Se RDS ARO | : 5 i ae menses o a eer : i t fae as NGEIIINS UNION , '
ee
eee ee ed ey ee reece
‘ot fr a Reif Mae RSLE00 | SE
Hi arpa me, ee ene Mapes inal pcp rate cme ‘
teste?
aosylesd odiT satiido2- aot Jeed@ wield eloaee ef salt
FIGURE 2
eee
— Cit Gia
PENSE elNG SUB ERAN AIEY SES OF THE SAME SAMPLE
DRIFTING DUNE SAND FROM NEAR YUMA,ARIZONA
MEAN OF THE MEDIAN DIAMETER = 0.131 PERCENT PROBABLE ERROR OF MEDIAN =0.8
HOr-8 36 4 “2 .| 08 06 .04 02 Ol MEDIAN DIAMETER IN mm.
FIGURE = 99:99
SIX SIEVE ANALYSES OF
THE SAME SAND SAMPLE tae t+ 1! TV | | | a DINEAR YUMA, ARIZONA Hh ff | MEAN OF THE MEDIAN DIAMETER= 0.136mm. : 95
‘PERCENT PROBABLE ERROR / OF MEDIAN=0.7% 90
AL) | | a
ae ee if coe ae
fee tet
CUMULATIVE PERCENT
re & i :
Hina a LY maps fog oc i He
CE ea ee Tee
MEDIAN DIAMETER IN mm.
399
fe) i<o) ice)
\
<
‘\
~~ = > SS —S> ~
a ee ‘alec SS:
al
as
(eae)
Els NE
SS
\
TIS
my Hany A i/o V; SETTLING TUBE ANALYSES OF SIXTEEN PORTIONS OF THE SAME SAND SAMPLE
CUMULATIVE PERCENT
COMBINED SAND: 1/3 LA JOLLA COVE 1/3 WINDANSEA BEAGH 1/3 SCRIPPS BEAGH
sO)-2S8) 16 4 «3 se .| O08 .06 .04 .02 Ol MEDIAN DIAMETER IN mm.
R
ij
OTTO MICROSPLIT
le §
10 ee PRaeetnae IN eerie SIXTEEN SpRIGNS ae A SINGLE BULK SAMPLE BY COMBINING ALTERNATE QUARTERS
ius
Aina
IGUIREMS 99.99
|| A-ALTERNATE QUARTERS |. COMBINED AS SHOWN IN ee
TOTAL PERCENT PROB- | B. ALTERNATE QUARTERS
| ABLE AHO OF THE NOT GOMBINED 99.5 | iweoran: = ff | |
MEAN OF THE MEDIAN Fey, ERROR OF MEDIAN= 1.1 ae
DIAMETER =0.391 H/ MEAN OF THE ME DIEN 98 PERCENT SPLITTING | DIAMETER = 0.39 | ERROR = 1.0 | PERCENT SPLITTING 95 ERROR =0.8
De) oO (o) (o) CUMULATIVE PERCENT
ih ae W ‘ : (ne ie ee
OS ee ee ee
SETTLING TUBE ANALYSES OF EIGHT
PORTIONS OF THE SAME SAND
1.0 Ash BS 4 we POR SB 6 4 2 i MEDIAN DIAMETER IN mm.
i 5
> en rp
u
nr. vu
an 4 fname °
oe
FIGURE ie
CE
9.99
oc) Se a nc ia — | =: ee ai pa fa : 60w 50 w 40 og 305 = 200
Ke)
5
2
SETTLING TUBE ANALYSES. OF FOUR |
PORTIONS OF THE SAME SAMPLE SPLIT TO DIFFERENT WEIGHTS MEDIAN DIAMETER = 0.127mm. 0.01
6 4 2 1 08 .06 .04 02 .O| MEDIAN DIAMETER IN mm.
al
: * as 7; ‘ be rie i
(Md+fine)—(Md-fine) (Md-fine)
VARIATION OF MEDIAN
NVIGAW JO NOILVIYVA IN3043d
FIGURE 9a °
PERCENT FINE PERCENTAGE VARIATION OF THE MEDIAN DIAMETER FOR A GIVEN PERCENTAGE OF FINE MATERIAL
FIGURE 9b
Sv
IWIYSLVW ANI4 3O S9VLINS90YSd NSD s Ve dOae OINTESOS aOe NOREVIEV A Sv ENS ose
3JNtd LN3S043d Ov SE O€ God O02 GI Ol S 0)
+
Co)
©
oN
a
si
oO ON'LMNOS NI NOILVIMWA LN3ONSd
Dew hss
=NOILVIYVA ONILYOS IN30Y3d
ico}
(@ul}-0S)
(9UlJ-0S)-(9Ul}+ 0S)
[
COMPARATIVE ANALYSES OF SAME SAMPLE WITH AND WITHOUT FINES (MATERIAL < |/I6mm. = 7.7 %) SAND FROM LA JOLLA BAY
TOTAL SAMPLE ANALYZED BY SETTLING OEE
ee a : / r ee ANALYZED BY ee TELE TT litrtverta ess teaw | 1/16 mm. REMOVED BY 70. [wet stevines 2 ttt} tt H+ +] |} — a 50
ee eee ee ae =_= A - aa ot) ise Oo
es ae ea
WE ET PP
EDN: nee in hae
FIGURE |Ob
COMPARATIVE ANALYSES OF SAME SAMPLE WITH AND WITHOUT FINES
TOTAL SAMPLE ANALYZED Sy SELLING wwerys
SAMPLE ANALYZED BY SIEM EING U BiE
(MATERIAL LESS THAN ii6émm. REMOVED BY
WET SIEVING.) :
ea Lie a ee
eS
O78 26 BS sO alll SOs) O05 LOS) Oe: MEDIAN DIAMETER IN mm.
CUMULATIVE PERCENT