petrographic and engineering properties of loess€¦ · remolded and compacted as construction...
TRANSCRIPT
BNGINEERING MONOGRAPHS No. 28
United States Department o f the Interior BUREAU O F RECLAMATION
PETROGRAPHIC
AND ENGINEERING
PROPERTIES OF LOESS
by H. J. Gibbs and W. Y. Hol land
November 1960 $1.00
~
l' Cr~~
t If(/l
tl J\r
I 1;t,
""," ;;""I ,' ;'
,. '.'~'", r
'Jo' 0I .'.~!1~ j.
BUREAU OF RfCLAMA nON DENVER UBf!ARY
I'I'II~'IIIII~/II"~III'920251b3
United States Department of the Interior
FRED A. SEA TON, Secretary
Bureau of Reclamation
FLOYD E. DOMINY, COMMISSIONER
GRANT BLOODGOOD, Assistant Commissioner and Chief Engineer
Engineering Monograph
No. 28
PETROGRAPHIC AND ENGINEERING
PROPERTIES OF LOESS
byH. J. Gibbs, Head, Special Investigations and Research Section
Earth Laboratory Branchand
W. Y. Holland, Head, Petrographic Laboratory Section,Chemical Engineering Laboratory Branch
Commissioner's Office, Denver, Colorado
Technical Information BranchDenver Federal Center
Denver, Colorado
ENGINEERING MONOGRAPHS are published
in lirnit~d ~@ition~for till! tl!chnicA.l !t!.ff ofthe Bureau of Reclamation and interestedtechnical circles in Government and privateagencies. Their purpose is to record devel-opments, innovations, and progress in theengineering and scientific techniques andpractices that are employed in the planning,design, construction, and operation of Recla-mation structures and equipment. Copiesmay be obtained from the Bureau of Recla-mation' Denver Federal Center, Denver,Colorado, and Washington, D. C.
CONTENTS
INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . .
GENERAL DESCRIPTION OF LOESS. . . . . . . . . . . . . . . . . . . . .
ORIGIN OF LOESS. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DETAILED DESCRIPTION OF LOESS. . . . . . . . . . . . . . . . . . . .Type of Matrix. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Cementation. . . . . . . . . . . . . . . . . . . . . . . . . . . .
Particle Shape. . . . . . . . . . . . . . . . . . . . . . . . . . .
Granular Components. . . . . . . . . . . . . . . . . . . . . . . . . . .
Grain Coatings. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
OBSERVATION OF PHYSICAL PROPERTIES. . . . . . . . . . .
Location of Engineering Studies. . . . . . . . . . . . . . . . . . . . . . .
Gradation. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Specific Gravity. . . . . . . . . . . . . . . . . . . . . . . . . . .
Plasticity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Compaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Permeability. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Consolidation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Shear Resistance. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
RESEARCH FINDINGS FOR LOESS. . . . . . .'.'
. . . . . . . . . .
Results of Plate Load Tests on Loess. . . . . . . . . . . . . . . . . . . .
Results of Ponding Tests. . . . . . . . . . . . . . . . . . . . . . .
Example of Prewetting. . . . . . . . . . . . . . . . . . . . . . . . . . .
Behavior of Piles in Loess. . . . . . . . . . . . . . . . . . . . . .
Results of Pile Driving Studies. . . . . . . . . . . . . . . . . . . .
Results of Pile-load Tests. . . . . . . . . . . . . . . . . . . . . . . . .
Summary - Study of Piles in Loess. . . . . . . . . . . . . . . . . . . . .
DENSITY OF LOESS AS A CRITERION OF SETTLEMENT. . . . . . . . . . .
CONCLUSIONS. . . . . . : . . . . . . . . . . . . . . . . . . . . . . . .
LIST OF REFERENCES. . . . . . . . . . . . . . . . . . . . . . . . . . .
Page
1
1
4
678899
99
11111111111414
2020202222262828
32
35
37
Number
LIST OF FIGURES
1
2
Outline of major loess deposits in the United States. .
Page
2
Photographs of typical characteristics of loess deposits. . . . . . 3
3
4
Photomicrographs of loess structure. . . . . . . . . . . . . . . 5
Location of major loess deposits in the Missouri River Basin. . . . . . . 10
5 Trends of gradation curves for loess. . . . . . . . . . . . . . . . . . 12
6 Trends of plasticity characteristics of loess . . . . . . . . . . . 13
7 Trends of permeability for loess . . . . . . . . . . . . . . . . . . . 15
8 Consolidation characteristics of loess. . . . . . . . . . . . . . . . . 16
9 Generalized consolidation trends . . . . . . . . . . . . . . . . 17
10 Shear tests of representative loessial soils. . . . . . . . . . . . 18
11 Shear and volume change characteristics of dry and wet loess . . . . . . 19
12 Plate-load tests for a typical loess deposit for comparison of material atnatural moisture and prewetted conditions. . . . . . . . . . . . 21
13 Records of ponding experiments at Ashton pile test areas . . . . . . . . 23
14.
15
Prewetting of loess foundation by ponding at Medicine Creek Dam. . . . . 24
Medicine Creek Dam foundation settlement installations. . . . . . 25
16 Profiles and plans of test sites. . . . . . . . . . . . . . . . . 27
17 Gradation tests. . . . . . . . . . . . . . . . . . . . . . . . 27
18 Method of making pile load tests. . . . . . . . . . . . . . 27
19 Driving records. . . . . . . . . . . . . . . . . . . . . . . . . . . 29
20 Comparison of pile load tests for friction piles in loess. . . . . . . . . 29
21 Requirements for adequate supporting capacity of piles . . . . . . 30
22 Requirements for practical placement of piles. . . . . . . . . . 31
23
24
Natural density versus liquid limit relationship for loessial soils. . . . . 34
An example of subsidence of low-density loess after saturation from canaloperation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
v
INTRODUCTION
This monograph summarizes knowledgeof the physical and physical-chemical prop-erties of loess gained from Bureau of Rec-lamation experiments, investigations, andobservations. Because of the unstable prop-erties of loess which may cause settlementof foundations of structures, such knowledgeof the limitations of loess for engineeringpurposes is important not only to the geolo-gist and soil mechanics engineer but alsoto design and construction engineers whoare required to build structures on loessialsoils.
The Bureau of .Reclamation has parti-cular interest in the engineering propertiesof loess because of its requirements forlarge hydraulic structures in loessial re-gions of the western United States. To de-sign and construct such structures as dams
GENERAL
and canals in these areas, the Bureau re-quired information on the characteristicproperties of loess, particularly those prop-erties which would be affected by moisturechange s resulting from the operation ofhydraulic structures. This monograph dis-cusses the origin of loess and presents de-tailed descriptions of the pertinent charac-teristics of loessial soils including matrix,cementation properties, and particle shape.It discusses physical properties includingpermeability, compaction, consolidationand shear resistance characteristics, andsummarizes the results of loading tests offootings and piles. A criterion for deter-mining the critical nature of loess in respectto the probability of settlement is given onthe basis of observations and tests in theNebraska-Kansas area.
DESCRIPTION LOESS
Loess is an earth material which coversvast areas of the central part of severalcontinents of the world. The major loessialdeposits in the United States are shown onthe map in Figure 1. The most extensiveloessial area is in Nebraska, Kansas, Iowa,Wisconsin, illinois, Tennessee, and Missis-sippi. Other loessial deposits occur insouthern Idaho and Washington. The depositsof major concern to the Bureau of Reclama-tion are in Nebraska, Kansas, Idaho,'andWashington.
~cause the particle structure of naturalloess is often very loose and contains appre-ciable voids, problems related to large con-solidation under certain moisture and loadcombinations, poor stability, seepage, anderosion frequently arise. On the other hand,loessiql soils which have been reworked bywater action are usually in a more densestate and exhibit better engineering proper-ties. Likewise, loessial soils, which areremolded and compacted as constructionmaterials, also exhibit better engineeringproperties. Naturally dry loess exhibitsconsiderable stability permitting it to standon steep cut slopes, as shown in Figure 2a,provided the natural moisture contents arenot increased. Natural drainage troughswhere higher moisture contents accumu-late display slumping and subsidence, asshown in Figure 2b, which illustrates a cri-tical condition that can occur in hydraulicstructures that eventually saturate portionsof their foundations. Wetted loessial soils
OF
are also weakly resistant to moisture move-ments as displayed by the erosion patternsof natural drainage streams, as shown inFigure 2c.
Olr studies lead us to support the theorythat the finely divided rock-powder silt ofwhich loess is primarily composed was pro-duced by interaction of sever~l geologi~alprocesses, but mainly by abrasIOn resultl?gfrom movements of continental glaciers.With retreat of the continental glaciers, thesilt was deposited along flood plains of riv-ers. This silt was then transported, sorted,and redeposited by wind action. Loess iscomposed primarily of fine, angular grainsof silt, fine sand, and calcite admixed withrelatively small amounts of clay and claycoatings on the larger grains. Some organicmatter occurs in the form of roots and roothairs. The particle arrangement in the soilmass resulting from the wind-blown methodof deposition is frequently very loose. There-fore, loess is susceptible to high degreesof consolidation. However, the dry loesswhich exists in the arid regions of the mid-western states has high strength owing tothe binding characteristics of the clay andis often capable of supporting relativelylarge loads without excessive settlement.It is easily collapsed by heavy loading andwetting, and the bond between particles canbe reduced to practically no strength bywetting.
Loessial soils have been analyzed in
1
~.~?
2
'f
(/)
w~~(/)
aw~z::::>
WJ:~Z(/)~(i)
-0
wa.
a::W
=>a
~IJ..(/)(/)w0..J~0,«~L
L0WZ..J~::::>0
STEEP SLOPES CONSTRUCTED I N NATURALLY DRY LOESS SHOWING CONSIDERABLE STABILITY
((1)
- - -. , P X -
A NATURAL DRAINAGE CHANNEL WHERE HIGH MOISTURE CONTENTS ACCUMULATE, CAUSING SLUMPING AND SUBSIDENCE
TYPICAL EROSION PATTERN OF NATURAL DRAINAGE STREAMS IN L O E S S
(c)
PHOTOGRAPHS OF TYPICAL CHARACTERISTICS OF LOESS DEPOSITS
FIGURE 2 3
considerable detail by the Earth and Petro-graphic Laboratories of the Bureau of Rec-lamation.. The initial sununary of consolida-tion and related properties of loess observedby the Bureau of Reclamation* was reportedby Holtz and Gibbs in a 1951 paper 1..1.
Theundisturbed structure and the miner-alogical characteristics of loessial soils havebeen observed. The internal structure ofloess is shown in Figure 3. Figure 3a is aphotomicrograph of undisturbed loess soiltaken with a magnification of 12X. Thisphotograph shows the loose arrangementof silt particles with typical numerous voidsand rootlike channels. The structure istypical of natural wind-blown deposits ofloess which have not been reworked. Figure3b is a photomicrograph of a thin section
*Numbers in superscript refer to publica-tions in the List of References.
ORIGIN
Certain conditions must be present forthe development of loessial deposits. Theseconditions are:
(1) A source of intermixed silt and clayadequate enough to account for the knowndeposits
(2) A period of strong winds prevailingfrom one direction
(3) A place of deposition
(4) Arid to semi -arid conditions at leastfollowing the issuance of silt from the gla-ciers and at all times following deposition
The origin of loess has been a subject ofmuch debate for many years. It is now gen-erally agreed that loess was wind transportedand deposited, and that the sediments wereof glacio-fluvial origin. The finely dividedrock powder of which loess is composed wasproduced by the interaction of several geo-logical processes, but mainly by abrasionresulting from movement of continental orlarge mountain glaciers. With the retreatof the glaciers, the suspended load of theglaciers was deposited in outwash plainsand subjected to erosion both by wind andby water. The material carried off by thestreams was later deposited as silt alongthe flood plains of rivers and then redepos-ited by wind action. As noted by Hobbs 2/,anticyclonic winds blowing from the retreat-ing glaciers could have acted upon the melt-
which was cut from a similar undisturbedsample of loess. This photomicrographhas a magnification of 30X, and shows therelative size of the void spaces. The prep-aration of this thin section and other sec-tions was accomplished by impregnating acarefully cut specimen with a thermoplasticresin which acted as a binder during thegrinding process.
The majority of the silt grains in loessare coated with very thin films of clay. Thisclay is generally montmorillonite and formsintergranular supports or braces within thestructure. Calcite usually occurs in distinctsilt-sized grains throughout the loess in afinely dispersed state rather than as a ce-menting material. The thin clay coatingsand, to a lesser extent, the calcite, ap-parently bond the particles together. Uponwetting, this bond is easily loosened caus-ing great loss in strength.
OF LOESS
water silt issuing from glaciers after thesilt had dried upon outwash plains, and re-deposited this silt in regions bordering theglaciers. Other sources probably contri-buted to the formation of loess but theircontribution was small.
In the Missouri River Basin and UpperMississippi River Valley the source of thewind-blown silts is generally agreed uponto be of glacio-fluvial origin. At variousinterglacial periods and at the close of thePleistocene ice epochs, large areas alongthe flood plains were left devoid of plantcover and exposed to strong winds. Theflood plains of these rivers were composedmainly of sand, silt, and clay derived fromthe abrading action of the glaciers. Thesilt and clay particles with intermixed finesand were sorted from the coarser sand,transported, resorted, and redeposited bywind, the coarser sized particles being de-posited close to the source and the finerparticles deposited at greater distancesfrom the source. Smith 3/ in his investi-gation of the Illinois loess, and Swinefordand Frye~..I in their investigation of Kansasloess showed that the grain size of loessdecreases logarithmically away from thesource. Swineford and Frye traced thesource of most of the Kansas loess by parti-cle size distribution to the Republican andPlatte Rivers. These rivers carried muchsilt which had been derived by the abradingaction of glaciers in the Rocky Mountainregion.
4
( A ) PHOTOMICROGRAPH SHOWING THE TYPICAL LOOSE OPEN STRUCTURE OF SILTY LOESS SOILS. THE MICROBOTRYOIDAL TYPE OF STRUCTURE WHICH OCCURS FREQUENTLY THROUGHOUT THE LOESS IS ALSO SHOWN IN THE LOWER CENTRAL PORTION OF THE PICTURE. MAGNIFICATION 12X
( 8 ) PHOTOMICROGRAPH OF A THIN SECTION OF LOESS TAKEN TO SHOW THE MICROSTRUCTURE OF AM AREA SIMILAR TO THAT SHOWN I N FIGURE 4. NOTE THE EXTREME POROSITY (LIGHT AREAS) OCCURRING IN SUCH AREAS. MAGNIFICATION 3 0 X
PHOTOMICROGRAPHS OF LOESS STRUCTURE FIGURE 3
Vegetation, as pointed out in some the-ories, may have contributed to holding thesoil together after its deposition. Fineclayey silt materials, however, are notgenerally disturbed by wind and are pickedup only if there is intermixed silt or sand.Loess then could have been deposited with-out the sustaining benefit cover of grasses,and the advent of grasses after the deposi-tion of the wind-blown silt would provide anexcellent growing material for larger plants.The development of approximately verticalroot systems might give the loess its verti-cal cleavage and would contribute in formingin loess its vertical porosity and permea-bility.
.
Wind-blown, intermixed silt and claythat eventually became loess must have beendeposited directly to the ground surface fromaerial transportation. Silts transported bywind, but deposited under subaqueous con-ditions do not have the typicalloessial fabric.Adequate sources for loess may exist, butif no suitable terrestrial areas for deposi-tion exist, no deposits of loess will form.For example, the Sahara Desert probably
was and continues to be a source of inter-mixed silt and some clay but the wind-blownmaterial is deposited either in the AtlanticOcean, or as in the Arabian Desert, thewind-blown material is deposited into theRed Sea or the Persian Gulf.
Intermixed silt and clay may be pickedup by the wind from its source only whendry climatic conditions exist following theirdeposition from either a fluvial or glacio-fluvial environment. Winds are unable topick up intermixed wet silt and clay. Cli-matic conditions, favorable to deposition,therefore, do not necessarily have to bearid or semi-arid but may be Arctic. De-posits of loess have been noted in the Ma-tanuska Valley, Alaska 5/, and Greenland 6/.Photographs of silt risfrig from the dry floodplains of the Tanana River, Alaska, il-lustrate that the material must be dry be-fore it can be transported. After deposi-tion and the development of a cohesive fabric,the climatic conditions must remain aridor semi-arid for loess to retain its openfabric.
DETAILED DESCRIPTION OF LOESS
Loess is a quartzose, somewhat feld-spathic clastic sediment composed of a UBi-form]y sorted mixture of silt, fine sand, andclay particles arranged in an open, cohesivefabric frequently resulting in a natural drydensity of 70-90 pounds per cubic foot. De-posits may reach considerable thickness.
Materials which are not cohesive andwhich are composed predominately of eithergranular silt, and fine sand particles, orclayey silt and fine sand size aggregations,are not considered as loess. Dust depositswhich may be composed of particles finerthan or equal to predominant grain sizesfound in loess are not considered as loess.The dust deposits may be potentially capableof forming loess, but until the coarse wind-blown material develops cohesion, the termloess cannot be applied. Similarly, clayeysilt and fine sand size aggregations may ac-quire cohesive properties and an open fabricsimilar to loess, but these materials cannotbe classified as loess if the wind-blown clayconsolidates and the mineralogicalcomposi-tion is different from loess. Deposits whichconsist of cohesionless silts, fine sands, orclayey silt and fine sand size aggregationstransported and deposited by wind should bereferred to as aeolian or wind-depositedsilts or fine sands or clays.
Many materials transported by eitherwater or glaciers alone or a combinationthereof yield silts which resemble loess ingra.in size, distribution, and mineralogicalcomposition. In the majority of water-laidsediments, however, the constituent grainsare fairly closely packed after the deposithas been subjected to several cycles of alter-nate wetting and drying and burial. It ispossible that freshly deposited water-laidsilts may develop an open fabric similarto loess and may be cohesive to some ex-tent. However. with subsequent wettingand the deposition of another layer of siltthe open fabric is lost and consequently theunit weight or density of the water-laid siltincreases over the freshly deposited silt.
Since wind-blown silt, which is depositedunder subaqueous conditions, is not consid-ered as loess and will have a different fab-ric from the wind-deposited silt, it is notimpossible to have wind-blown silts de-posited under lacustrine or playa conditionsto be mixed with or graded into loess. Smalllakes most likely existed after Pleistoceneglaciation and gradually became filled withwind-blown silt, fluvial silts, or sands andcould have been covered over with loess.
6
Silts transported and deposited by windvary considerably in their dynamic physicalproperties depending upon whether the siltwas deposited under arid, semi-arid, orhumid conditions. The wind-blown siltsdeposited directly to the ground surfaceunder arid and semi-arid conditions developan open cohesive fabric and retain this fabricas long as the climatic conditions do notchange. Local variations in the density orcloseness of packing in the massive loessprobably occur, as local heavy rains wouldcollapse some of the open fabric. Loessdeposited under semi -arid conditions wouldprobably develop a protective rather thanimpervious surface layer, as the effect ofrainfall over a long period of time would tendto be distributed over a large area ratherthan being localized as in an arid region.
Wind-blown silt deposited under.humidconditions or under arid conditions and latersubjected to humid conditions develops aclose packed fabric by a collapse of inter-granular supports or a gradual bringing to-gether of the granular components by wettingand drying. The densities of these wind-blown silts are frequently more than 100pounds per cubic foot. Samples of materialobtained from Natchez and Vicksburg, Mis-sissippi,. have densities ranging from 102 to112 pounds per cubic foot.
The upper few feet of loess frequentlyshow a collapsed or a denser fabric than theunderlying loess. This collapse in fabricmay be produced by a variety of factors suchas the activities of man, a progressivechange to a more humid climate with in-creased weathering, or by sheet erosion.The upper few feet of loess form top-soil.The primary or open fabric of the top-soilis destroyed by man during plowing and farm-ing operations and possibly irrigation. Theseoperations tend to disperse the clay fractionmore evenly throughout the granular compo-nents and produce a soil in which the granu-lar components tend to be in contact. Ifloess is subjected to humid climatic condi-tions, it gradually consolidates and the wholedeposit attains a dense fabric. However, theupper few feet in this case are subjected toweathering and sheet erosion, and the com-ponents begin to alter. Also the clayey com-ponents are more evenly distributed through-out the granular components.
The upper few feet of loess or the wholemassive loess deposit may be affected byslope wash. If loess is deposited on a slope,gravitational forces act upon the body andthe mass slowly moves down the slope. Inthis process the open fabric is destroyedand a closely packed fabric results. Gravi-tational forces become more active when
there is an increase in moisture content ofthe loess body and the moisture content maybecome so large that soil flow results.
From an engineering standpoint, loessis a potentially unstable clastic sediment.Loess is in equilibrium with its environ-ment, but as the moisture conditions changeor the overburden load becomes too great,loess tends to consolidate and assume amore stable arrangement of its grains.Consolidated loess assumes propertiessimilar to those of any other water-laidsilt having a similar granular and miner-alogical composition.
The properties of loess, whether con-sidered from a physical or physical-chemicalviewpoint, are dependent upon two factors,namely, fabric and mineralogical composi-tion. If two loessial materials with the samefabric and mineralogical composition aresubjected to physical tests, they should anddo act similarly. However, if anyone of thefactors constituting fabric changes or thecomposition of the clayey fraction changes,the properties are changed materially. Achange in the granular components, such assubstituting feldspar for quartz, does notaffect the physical properties materially,but with the introduction of volcanic glasswith its shard-like form as a granular com-ponent, the properties of the loess wouldcertainly change.
Type of Matrix
The type of matrix in loessial materialsis commonly clayey and less commonly cal-careous. Calcite is present in many loessialmaterials as grains and crystals but may besecondary. In the open fabric types of loes-sial materials the larger silt grains are con-nected by what might be called inter granularsupports which are composed mainly of amontmorillonite-type clay mineral and pos-sibly intermixed. with small amounts of illite.
In other types of loess, the clayey com-ponents are distributed interstitially but withthe silt grains not in contact. Voids may bedistributed throughout the matrix but areusually not distinct. This type of fabricseems to be the more common type. Thedensity of loess with this type of fabric isgenerally less than or equal to the loess withintergranuJar supports. In a loess from Ken-newick, Washington, the granular compo-nents are so fine grained that the arrange-ment of the grains in respect to the matrixis difficult to see. The silt grains are evi-dently dispersed throughout the clay, whichis mainly illite, and a petrographic thin sec-tion reveals only a continuous fabric.
7
In wind-blown silts which have been de-posited under subaqueous conditions or humidconditions the matrix remains clayey, butin the se materials the clay evidently isslightly more compacted and the grains arein contact which gives stability to the mass.
Calcite may occur asa secondary matrixmaterial in loess owing to leaching fromabove or by the evaporation of capillarywater from ground water below. However,zones containing calcite are generally notof great thickness under semi-arid condi-tions.
Cementation
Many investigators have stated that theprimary cementing material in loess is cal-cite and/or dolomite, mainly because thesample effervesces if hydrochloric acid isadded. In samples of loess from Kansas,Nebraska, Iowa, Washington, and Israelwhich were analyzed by the Bureau of Rec-lamation laboratories, calcite and dolomitewere found by microscopic investigation tooccur as anhedral detrital silt-size grainsand as euhedral crystals, whereas the clayminerals occurred as the main cementingmaterials. In samples of loess from Natchez
. and Vicksburg, Mississippi, calcite anddolomite also occur as crystals and grains.Calcite and dolomite may help slightly to bindthe soil particles together, but the majorcementing materials appear to be the clayminerals.
If the main cementing minerals werecarbonates, the loessial materials shouldnot slake or soften appreciably in water. Allof the samples of loess examined slake read-ily in water, either immediately or withinone or two minutes. Further evidence thatthe carbonates are not the cementing mate-rial in loess is shown in consolidation andtriaxial shear tests. If the carbonates werethe main cementing material, the test speci-mens should not show (as they do) immediateconsolidation when water is added to the soilhaving low natural moisture content.
Zones of loess may contain carbonatesas the principal cementing material wherethe carbonates have been leached from theupper loessial material and redeposited atlower depths. The introduction of carbo-nates may be accomplished also by capil-lary action. If the ground water table issufficiently close to the loessial body, watercarrying Ca++ in solution as the bicarbonatemay be drawn up into the loess, and with achange in pressure or temperature,- calciumcarbonate would be precipitated in the porespaces. In the large pore spaces. the evapo-ration should be greater and calcium carbo-
nate would tend to concentrate in these poreswith formation of concretions. With re-peated dissolution and precipitation the car-bonates would tend to develop better crystal-line forms as are observed in many loessialmasses.
Particle Shape
The shape of the individual granularcomponents of loess ranges from irregularto platy, fibrous, or prismatic. The de-gree of rounding ranges from subround toangular. Quartz and the feldspars are com-monly irregular in shape, whereas the micaminerals are platy or shred-like. Horn-blende, zircon, pyroxenes, and apatite arecommonly in prismatic form and sillimanitein fibroUs form.
Calcite is present as irregularly shapedgrains, rhombohedral crystals, and aspatches composed of very fine-grained parti-cles. Dolomite is usually present as rhom-bohedral crystals.
Acidic volcanic g]ass in the shape of vol-canic shards is often observed in loess fromthe western United States. These shardshave a variety of shapes and are fresh ormoderate]y altered to montmorillonite. Dia-toms and globigerina tend to Occur rarelyin loess. In some loessial materials, verysmall tubes composed of opal are present.The tubes are thin walled and have a veryirregular surface. These tubes are thoughtto have formed around the fine root hairs ofplants.
The clay minerals, montmorilloniteand illite, occur as very fine platy crystalswhich are commonly cemented into aggre-gl;ltes between the granular components.Montmorillonite commonly occurs as thinhulls around the grains. Illite, however,has a tendency to occur as individual shred-like crystals. Platy crystals of kaoliniteand octohedral crystals of cristobalite inloess deposits in Kansas have been observedby Swineford and Frye 4/ with the electronmicroscope. -
Organic matter, which is visible underthe petrographic microscope, ranges fromfine capillary root hairs to plant stems.These plant stems may be partially decom-posed or totally carbonized.
The majority of the granular componentsare subangular in shape, the grains becom-ing more angular as the grain size decreasesand mare rounded as the grain size increase stowards the fine sand size.
8
Granular Components
The granular components, as deter-mined by X-ray diffraction and petrographicanalysis of the loessial materials fromSherman Dam site and Trenton Dam site,Nebraska, range in value as follows:
Quartz 25 to 37 percent
Feldspar, includingorthoclase (whichis predominant),plagioclase, andmicro cline 10 to 22 percent
5 to 10 percentAcidic volcanic glass
Biotite and musco-vite 4 to 6 percent
Green and brownhornblende andchalcedony O. 5 to 1 percent
Minor accessories include clinozoi-site, epidote, chlorite, zoisite, sillimanite,augite, sphene, garnet, rutile, zircon,apatite, magnetite, tourmaline, hematite,and a glauconite-like mineral. Small amountsof gypsum, 0.1 percent, are occasionallyassociated with loess. Similar minera-logical compositions of loess from variousparts of Kansas ha;ve been reported by Swine-ford and Frye 4/ except that the samplesexamined from-the various dam sites arelower in quartz and higher in clay contentthan other samples from Kansas examined.
OBSERVATIONS OF
Location of Engineering Studies
Numerous B.1reau of Reclamation struc-tures are proposed, are under construction,or have been constructed on the MissouriRiver Basin Project. Investigation for sev-eral structures, as well as special researchstudies, have permitted the gathering of testdata from several locations throughout themajor loessial regions of Nebraska andKansas. These particular locations areshown on the map of the Missouri RiverBasin area in Figure 4. The portion of this
Grain Coatings
Most of the ~ranular components ofloess from Kansas, Nebraska, and Iowaare coated with a very thin hull or shell ofa montmorillonite type of clay mineral.These clay hulls are firmly attached to thegrains and cannot be removed even afterlong dispersion in a dispersion machineand subsequent treatment with 1: 1 hydro-chloric acid. Swineford and Frye 4/ alsohave observed these clay coatings~on thesilt grains inloessialmaterialfromKansas.
In some instances, the clay coatings.completely envelop the grains, whereas ina few instances the grains are only partiallyenclosed. In petrographic thin sectionstabular or prismatic grains which have acomplete clay coating have been observedin which the coatings on opposite sides ofthe grain extinguish together as the micro-scope stage is rotated.
It has been observed that in the maJorityof instances water-laid silts that are derivedfrom the decomposition of pre-existing rocksdo not carry clay hulls, whereas the grains.in wind-blown silts commonly do. Clay hullsmay be formed on the grains of some presentsoils because of the downward concentrationof colloidal clay. These clay hulls probablyaid in bonding the granular components to-gether and in bonding the grains to the clayin the matrix. The presence of these clayhulls would also aid in retaining a certainminimum moisture content, and the pres-ence of moisture in these hulls would exerta surface tension effect thereby binding thegrains more tightly together with evapora-tion or rise in temperature. Hence, mois-ture films or moisture associated with theclay rims and matrix are important in thebinding action of loess.
PHYSICAL PROPERTIES
map bounded by the heavy line shows thelocation of the large loessial area which isof major concern to the Bureau of Reclama-tion and from which most of the test datafor this monograph were obtained.
The physical properties of loess, when,summarized for loesses from locationsspread over wide areas in Nebraska andKansas, indicate basic characteristicswhich are useful for engineering purposes.Based on numerous tests of soil samplesfrom these areas, it was found that the basic
9 "~,,~,' ,
BUREAU OF R£CLAi'vIAliON
....0 EXPLANATION. Dam sites
investigated. Canals investigated
LOCATION OF MAJOR LOESS DEPOSITSIN THE MISSOURI RIVER BASIN
FIGURE 4
properties of the loess are quite similarand, predominantly, the loess has siltycharacteristics. However, some of theloesses are more clayey and others aresomewhat sandy. Other investigators,Davidson and Sheeler 7/, have shown thatthe changes in properties of loess are re-lated to the distance of the deposits fromtheir probable source. The coarser or moresandy loess is deposited nearest to thesource and the predominance of the clayeyand finer loess increases farther from thesource. Although this trend is apparent,it is believed that properties of the loess,especiall¥ those of the Nebraska and Kansasarea, are sufficiently similar to establishcertain important generalized findings forresolving soil mechanics and foundationproblems.
Gradation
The graph in Figure 5 shows that gra-dations of loess are of uniform particle sizeand are consistently similar for differentsamples. From 148 samples which weregeologically described as loess, 76 percenthad gradation curves which were in the siltyloess range shown in the figure, 18 percentwere in the clayey loess range, and 6 per-cent were in the sandy loess range.
Specific Gravity
The specific gravities of these sampleswere found to range between narrow limitsof 2.57 and 2.69; the average was 2.65.
Pla sti city
The plasticity characteristics were de-termined by consistency (Atterberg limits)tests on many samples from 11 locations.The results are plotted in Figure 6 for thepurpose of showing the areas on a graph inwhich values are concentrated; a standardclassification chart was used. It is evidentthat the major concentration of points occursfor plasticity indices from 5 to 12 combinedwith liquid limits from 25 to 35 percent.When these data were studied in conjunctionwith the gradation curves, it was observed
, that this group is generally characteristicof the material which was previously identi-fied as silty loess. Points of higher plas-ticity indices and liquid limits representedmore clayey loess, and points of lowerplasticity indices and liquid limits repre-sented the sandy-loess soils.
Davidson and Sheeler 7/, in theirstudies of loess in western IOwa, showedthat the loess becomes more clayey whenthe distance of the wind-blown deposits isfarther from their probable sources. The
loess in western Iowa w~ very similar tothe predominating silty loess of the Ne-braska-Kansas area; loess farther east hadincreased plasticity and cation exchangecapacity,-and was thus more clayey incharacterlstics. Some clayey loess wasalso found in the Bureau of Reclamationinve?tigahons~ -;Although systematic trav-ersesof.:locations were not studied, it isprobabIE(that'distances of deposits fromsour-ce.s'tnay also be an explanation forthese ~Hte,r;ences in properties., .
Compacti<jHi
This' monograph primarily concernsundisturbed loess since it is this conditionwhich involves the special properties ofIDess. Recompacted loess is not considereda special type of material but can be saidto have typical properties of a fine-grainedsoil of silt and silty clay characteristics.Pecause of the relatively uniform propertiesof loess, the compaction characteristics donot vary appreciably.
Compaction characteristics of loesswere not analyzed in detail. However,Figure 7b, which is discussed below in re-gard to permeability, contains data on soilsrecompacted at maximum Proctor densitywhichrange from 100 to 112 pounds percubic foot.
Permeability
The permeability of natural loess isprincipally related to the density when itis considered that basic properties of gra-dation and plasticity, as previously shown,are relatively uniform. However, thereare other features which influence thistrencl to some degree. Natural loess isknown to have rootlike voids which wouldcause considerable variations in permea-bility; however, to some extent these root-like voids are also reflected in the low den-s ity. In very dense and reworked orrecompacted loess, these natural rootlikevoids are not present, and permeabilityvalues are lower. Although loess is con-sidered relatiyely uniform in basis proper-ties, the loessial soils which fall in themore clayey or more sandy groups shouldbe expected to have somewhat lower orhigher permeability values, respectively.
11
In Figure 7a, all of the permeabilitytest results are for undisturbed naturalloess with tests made in the vertical direc-'tion using the one-dimensional consolida-tion apparatus. By enclosing the concentra-tion of points with lines a trend is shown.In reviewing the gradation and plasticity ofeach sample it was found that t~ "E~
'tecllniCe. l\ecla:tne.tiO-o.of .e.C-O
~uree.u coloI-pen<Je1: .
90
80(!)
z: 70(/)
~600...
I- 50z:
~40
~300...
20......t-.:)
10
a
/ - -;7 I~I
/ /' ". I
4 I l~" I
/1t ~,' II
'/'"~,. I
'" ",'I
" '" ""II
,,",,(;::j -....J -.....I,1,""'~~,,",~~<c;, J..... c:>,' I
,,~ 1",(;::j ::::! ~, I
c.../ '" ~,' :I
V ,."I
~. :~--- I
I I -Iii II I I II II I I I I I" I I I I II I I II III
CLAY (PLASTIC) TO SAN D GRAVELCOBBLE
SILT (NON PLASTIC) FI NEI
MEDIUMI
COARSE FINEI
COARSE
,HYDROMETER ANALYSIS
I
SIEVE ANALYSISU.S.STANDARD SERIES ICLE~,R SqUAR~, OPENINGS
#200 #100 #50 #30 #16 #8 =t*4 tilt 3" '5'6"8"100 a
10
200w
30 z«
40 tJa::
50 I-z:
60 LiJu
70 a::w
8 a0...
..
90
100.002 .005.009 .019 .037 .074 .149 .297 .590 1.19 2.38 4.76 9.52 19.1 38.1 76.2127152
DIAMETER OF PARTICLE IN MILLIMETERS
S
Gradation data was obtained on 148samples from projects in theMissouri River Basin area
The curves generally take the directionshown by the boundary lines
TRENDS OF GRADATION CURVES FOR LOESSFIGURE5
For all samples tested76% were in the silty (oess zone18 % were in the clayey loess zone6% were in the sandy loess zone
X20
UI0Z
>-!::0
l-I/)eX..JQ.
......~
30
10
CL-ML--- -
00
CLAYEY LOESS"';..-
..... e./
/ e.I
I~.
\ e.+ .
SI L TY LOESS-, \ , *", A .~ ......
}:' ~ \ --/ lie.
/ ty.~ 0 J
CL / e. ~ye.e. fI
y~n 1
:~~ I ,/,f'8, */. e.L ~ '0-.-:'-
SANDY LOESS / E /1 ML- end OL
10 2~0cO 0
* 30
LIQUID LIMIT
CL
CH
.
MH end OH
40 50 60
Consistency Ii m its data were obtained for samples from t he following structures:
+ Cushing Dam0 Milburn Dam. .Amherst Dam
. Courtland Canol
. Cambridge Conal
TRENDS OF PLASTICITY CHARACTERISTICSFIGURE S
8 Trenton Dam and Railroad RelocationA Bonney Do m
* Davis Creek Do me. Ashton pi Ie test area
0 Rockvi lie Do mY Medicine Creek Dam
OF LOESS
falling within these lines were generallyfor the silty loess. Points falling abovethese lines and toward values of higherpermeability were for sandy loess, andpoints falling below these lines and towardvalues of lower permeability were for clayeyloess and possibly reworked natural loess.
Using these trends in Figure 7a as astandard for comparison, the results ofrecompacted and reworked loess are shownin Figure 7b. Most of these tests weremade on specimens which were recompactedfor the purpose of embankment investiga-tions. Therefore, it will be noted that most.of the specimens have a density range from102 to 112 pounds per cubic foot. For thesetests, the permeability range is rather low,from 0.01 to 0.5 foot per year. Such valuescorrespond to values observed for densenatural loess. Compacted samples havinglower densities were not available.
Consolidation
Only general trends are evident in thesettlement characteristics shown in Figure8, because considerable variations occurin density and moisture. Slight variationsin clay content and differences in moistureconditions influence the bonding character-istics from particle to particle.
The prominent effect on the amount ofsettlement is found to be initial density.Under a loading of 100 psi and at wettedconditions, natural loess, which varied ininitial density from 73 to 92 pounds percubic foot, consolidated to densities from95 to 103 pounds per cubic foot, or a~ muchas 30 percent or as little as 5 percent, asshown in Figure 8a. However, as shownin Figure 8b, under relatively dry naturalconditions, loess consolidated little untilloads are in excess of 100 psi. This indi-cates that loading alone does not break downthe particle-to-particle bonding character-istics until relatively large loadings arereached, whereas increased moisture appre-ciably affects settlement.
Figure 8c is another generalized picture,published previously by Holtz and Gibbs1/, of the consolidation characteristicsPlotted to linear scales. The trends forloess which is thoroughly wetted (shadedarea) show that it consolidates quite easilyunder small loads. Three examples ofconsolidation tests -show the contrast oflow settlement under dry conditions andhigh settlement under wet conditions. Whenwater is added during the test the curvesdrop to values near those obtained by com-panion prewetted tests. These curvesdemonstrate settlement characteristics
which may develop In structures placedon wetted loess, on dry loess, or on dryloess which is wetted after construction.
Using the data of the actual tests inFigure 8 as guides for establishing consol-idation trends, the curves of Figure 9 weredrawn. These curves demonstrate distinctcontrast in settlement between loess of wetand dry conditions for loads of 10 to 100psi, which is the general range of loadingfor most foundations. It is evident that themost pronounced settlements occur forloess at densities less than 80 pounds percubic foot, and relatively smaller settle-ments occur for loess at densities morethan 90 pounds per cubic foot.
Shear Resistance
The variations in shear resistance maybe attributed to the same general factorswhich cause variations in consolidation.There is evidence of definite influences ofinitial density, initial moisture content,and clayeyness of the loess.
In Figure 10, the solid lines representthe shearing resistance curves (Mohr en-velopes of triaxial shear tests) of natural un-disturbed loess and the dashed curves repre-sent loesses which were prewetted. Althoughthe tests include both sealed and drainedspecimens, the results have been adjustedfor pore pressure, and the shear resistancecurves represent strengths for effective(grain-to-grain) stresses.
The curves are generally of similarslope indicating similar frictional resist-ance. This would be expected becauseloess has such similar basic physical prop-erties. However, there is considerabledifference in the position of the resistancecurves; This is caused by ttle variationsin cohesion which are relate1d to the claycontent, moisture content, and density.The several curves which are bracketedand located high on the graph are thoseshowing high cohesive characteristics. Itmay be noted in the table in the figure thatthese soils are either clayey, of very highdensity, or of very low moisture content,or are soils having combinations of theseproper tie s.
The curves bracketed at the lower levelson the graphs are those of typical silty loessgenerally of relatively low density. Themajority of the curves are contained in thisgrouping. This lower group of curves con-tains test results for two test methods; thesolid lines represent tests on loess at natural(dry) moisture content, and the dashed curvesrepresent tests on natural loess which has
14
t---- II111.. UNDISTURBED LOESS
i~II<. ... y
I'-,
*y
y-'---:>/4'* c
'. /"ojc ,
b"'~'"- -, . .y .~/c,f)~A 0
"~~~'"'' "I I"'. y
I..
"I'
yy y
b6-c,C?
~~,c.c,a/
"""y ,,q,-.J
/
",°q,'t:,/ #y
' '-"9..' ~/"-
t)~~t) I
0/Cj
<.'l.// " 1"\, 0~yy
,.Y[' .. .'\y }j
0
""
{'f ,.
~~~y,
"-;yyy
I
*"*
I
~-- II1111~-6 *y" y" ..
"REMOLDED LOESS
".f- '
.........
!*" ty O.y
* """"'
"*
I. . ,,,,"/ 09..
,v
-..:"< '
b'"
",'"CO
"""""' .........
"I'\.
* ,
().c::.~c,c,I ' ,q,c,c,,,oq,/
",°'l.'b/"\ '"9..' ~/
,o""q,...oI
""\c..;
<.//
""
'\.
85
110
105
'"'...
::> 10000:
'"a.
ai 95...J
>-....
~ 90
'"0
>-0:0
80
750.1 1.0 10
PERMEABILITY - FT. PER YR.(0)
100 1000
85
110
105
,.:...
::>1000
0:
'"a.
iii 95...J
I
>-....
~ 90
'"0
>-0:0
80
750.1 1.0 10
PERMEABILITY -FT. PER YR.(b)
100 1000
EXPLANATION
. Trenton Dam and Railroad Relocation
.. Bonny Dam
*Davis Creek Dam
"Ashton pile test area
c Rockville Damy
Medicine Creek Dam+ Cushi ng Dam
0 Milburn Dam. Amherst Dam
. Court land Conal0 Cambridge Conal
*Enders Dam
N Erickson Dam
TRENDS OF PERMEABILITY FOR LOESSFIGURE 7
15
1.3
~75 1--1.2
~1'--14Z-524~
I
80 14Z-52b,
a '... .,0 ~U>
"-;::>- '": 1-
"-...:
.9I---in'"0 z <i(5
'"<f>0
> 90 >-~'" <t
.a0
7
100
.6
5_IIn
. 0.1
.7<
I
I "85 I
II- --- r- I -t+- --... - - - - W., -
-~~~fu U> --- --..
'"- ---
-~~>- -I-
"--- -- ,\
-in '"-j- - '
\Z ci. -
lU--- --
If4:\
'"<f> - --- -- \0
90 rr ~- II\
--- -0 <t -- -==== '1"+ '- --- - .-=-
F-i-t --- -~~--- -,--
I'100i I
i I,
!IIn
I I INOTE
The 5hoded area represent5 the consolidationcharacteristics shown by samples whichwere te5ted in a pre wetted condition (wetting
-Low denSity loess from by ponding). The results of all tests with
: Ashton Pile Test Area initial densities varying from 74 to 90 lb.....
,;Iesteddryj"--_okper cu. ft. fell within this area and the
>. ..... curves hod the general directional patternI -.~-Tested wet": of the area.
><
" """1-.....-',,
"Med. density loess-- .- from Trenton Dam
.-Tested dry', .:Tested~)I...--
zt-f --- ;;Woteroddedt---__-
I - -- ~-Woter added,
m /'/ /7., ,, ,, .
CTesteddry-, CWoteradded
I-Testedwet-- --, //
I Highdensity loessfrom Trenton Dam
I
I.J
1.2
1\\
~i\~~r---:
1.0
0I-..
'".9
~0>
'" r\
"'~ ~
~~
.a
.7
1\,.6
10LOAD -PSI
100 .50.1 10010LOAD ~ PSI
(0) SUMMARY OF CONSOLIDATION CHARACTERISTICSOF NATURAL UNDISTURBED LOESS WHICH HAS
BEEN WETTED PRIOR TO TESTING TO MOISTURECONTENT ABOVE 20 PERCENT
(b) SUMMARY OF CONSOLIDATION CHARACTERISTICSOF NATURAL UNDISTURBED LOESS TESTED ATNATURAL MOISTURE CONTENT OF LESS THAN
10 PERCENT
70
.,U>N...0>-I- 75;;..
'"'"u..
8g; 80
'"'": 1.0
'"0
> ;::« 85 «z
'"..z00
~ 90..~:;';:; 95u::;..~ roo>-I-in~ 1050
110
1.3
1.2
eO.90>
o.a
0.7
0.6
0.50 reo 220 240 260120 140lOAD-PSI
16060 ao 100 20020 40
(c) SUMMARY OF CONSOLIDATION CHARACTERISTICS OF DRY AND WETTED lOESSOF VARYING INITIAL DENSITIES
CONSOLIDATION CHARACTERISTICS OF LOESSFIGURE-8
16
75 1.2
in<D(\J
II 1.1>-I-- 80
'>«a:<.!)u 1.0r;:U Ow I--a.. «(/) a:w
@.9~a: >w>S 90r.:... .8::jUa:wa..
ai .7...JI
>- 100I--enZw .6a>-a:a
110
""'"
l'\- --- 1-. -- ~---
~1\c;;;;...
=-=- - 1-- - ~--- ~- ~\-. \- ...::.:.:. -- -1-
......... \ \ I'\
i-... --- ~'\ \~I-~ '\ \
~1\ 1'.\ \
-~ --- f-- f- \ '... \r--- r-r-. " '
-.- !-oj.. ~" ....
--- "-- -, ~1\-- -1'-... -
~~......j\""'",
~i'
1.3
,
.50.1 10
LOAD- P.S.L
100 1000
Wetted Loess (above 20%moisture content)
- Dry Loess (below 10% moisture content)
GENERALIZED CONSOLIDATION TRENDS/ FIGURE 9
been wetted prior to testing. It can be seenthat the low density loess which has beenthoroughly wetted is of very low shear re-sistance. Under low normal stress, thisloess shows a flat shear resistance curve,and on some occasions it has virtually noshear strength. Under higher normal stress,this strength increases because frictionalresistance was developed after consolidationtook place.
As a further explanation of the differ-ence between wet and dry conditions, thecurves in Figure 11 are shown. Thesecurves are for a typical silty loess of rel-atively low density. The solid curves are
for the undisturbed loess at natural mois-ture conditions, and the dashed curves arefor the same material which has been wettedprior to testing. The shear curves in Figurella show distinct differences in shear re-sistance between wet and dry conditions.The larger circles for dry loess indicategreater shearing resistance under appliedpressures. The higher level of the envelopeline indicates pronounced cohesion. Thesmaller circles for prewetted loess indicatelower shear resistance for applied stressessimilar to those used in the dry tests. Inthe wet tests, appreciable pore pressuredeveloped causing lower effective normalstress values.
17
LAB. TESTTYPE DENSITY NDIST.
SAMPLE OF AT TINE AT TINE REMARKSNO. CONDITIONS LOESS OF TEST OF TE ST
120-202 NATURAL CLAYEY 100.2 10.1 ~I12F-44 SILTY 94.1 8.4 I
I120-200 CLAYEY 84.9 8.6 I
I14!-478 SILTY 90.2 7.4
~ExtremelY clayey120-183 SILTY 84.2 4.5 I Loess and Loess
12F-143 SILTY 103.5 18.0 I of high densityI and low mois-
14Z-526 CLAYEY 83.1 13.3 Iture content
120-181 SILTY 79.1 8.2 ~I
14Z-527 SILTY 89.5 6.9 II
14Z-476 SILTY 83.8 8.6 III
13S-106 SILTY 89.0 6.0 I
13S-111 81.7 6.5I
SILTY II
13S-109 SILTY 83.8 7.3 II
13S-112 SILTY 81.3 9.5 II
14Z-524 SILTY 82.0 10.6 I
120-191 CLAYEY 78.6 20.3r
Typical si Ity
I Loess of lowI densityII
13S-105 PRE- SILTY 81.1 29.3 IWETTEO I
13S-111 SILTY 82.5 30.4 II
14Z-476 SILTY 81.6 34.8 II
14Z-521 SILTY 78.2 34.9 II
14Z-524 SILTY 78.0 33.9 II
13S-108 SILTY 78.4 28.9 I
EXPLANATION
- Tests at natural moisture condition
--- Tests at wetted cond ition
mCLen 100mWa::I-ma::~w:J:m
.....CO
150
/1/' -120-202
/'
//
/ ./
/~/Extremely clayey Loess and Loess of /high density and low moisture contenty
/~ \, // /~20-200
V '\ ) 12F-44,>-///13S-la6,././f\\ /./~F-143 d~./7
14Z-527-,L /l,/\I ~ .Y ./~././', 135-111
~ ./~ ./' .//./
./V I /-?' I 'Y:;:~0-180V /./ V /' "j~;'/ /"/ -
./
14Z-476- /V~ ~~~;: ~:-109 AND135-112
141:~~~::~~~~~3S-111
14Z-526T__, [.7~V I~~ /...;~ 135-1015~V/ '~.;..;, I
f14Z-524 V~'~ ~;~./ '-Typical silty Loess of lowdensity\. ~ //'
'- .:;:::' r~ V \ Typicalsilty Loess at\~
'/:;:;r... '-120-191 natural moisture content
/~~ ~".. '~Typical silty Loess of~ ~- highly wetted condition
E::;}i<".,4Z-476,521 & 524
-~~~---t3S-108 I I00 50
50
100 150
EFFECTIVE NORMAL STRESS-PSI
FIGURE 10
SHEAR TESTS OF REPRESENTATIVE LOESSIAL SOILS
FAILURE (MAX. OJ/<T3)
I I
Sealed Tests
~./ V
~~~Vl~ / - ......
0 ~r--...... .......
/ '" '"./
? 1 l/ " \~/ r:>-- '\."./' /j ~...-::::: \ \ \J ..... '...
l ...j~'r' '\ '/f1 ~bCf', I
,...75
.- -- ~'-, ~',
80 r--.o~IF
\uu n t'-...._I>- \ ~,..I-
85 en ".
z ' 1\I.LI ~b0
1\ \
90
INITIAL CONDITIONS NATURAL PONDEDDRY DENSITY (pcfL 82.02 78.02MOISTURE CONTENT (%L- - 10.57_- -- -- -- __33.96DEGREE OF SA TU RATION (%) - - -- 27.50- --- - -- __80.23SPECI FIC GRAVITY 2.660
.....co
80
~ 601
(J)(J)
I.LI0:
~ 40
0:<tI.LI:r(J) 20
00 20 40 60 80 100
EFFECTIVE NORMAL STRESS-PSI
(a)
1.2
1.1
Q)
10
i=<a: 1.00:
0
0>
.9
.8
120 I 10
DEVIATOR STRESS-PSI
0.1 100
(b)
SHEAR AND VOLUME CHANGE CHARACTERISTICS OF DRY AND WET LOESS
FIGU RE II
Figure lIb shows the consoliaationwhich occurred in the shear specimens.The dashed curves showed that wet loessconsolidated under much lower stress con-ditions than the dry loess. The rise inshearing resistance under higher stresses
RESEARCH FINDINGS
Several research studies, both in thelaboratory and in the field, were conductedto clarify specific problems related to theuse of loess as a foundation and constructionmaterial.
Results of Plate Load Tests on Loess
To demonstrate the settlement and bear-ing capacity of spread footings on loess, aseries of plate load tests were conducted ina typical low density loess near Ashton,Nebraska 1/.
"-
The settlement characteristics of thissoil are shown by the laboratory consolida-tion curves marked 14Z-521 and 524 inFigure 8a. The field tests involved the load-ing of square plates of three sizes: (1) 1 by1 foot, (2) 3 by 3 feet, and (3) 5 by 5 feet.Each plate was placed at a depth of 5 feetin holes equal to the plate size so that thefull effect of surcharge existed.
The results of the tests are shown inFigure 12. Settlements are shown in rela-tion to the pressure. The results of eachloading test are indicated for two sets ofplates: for the loess at natural moisturecontent, and for loess which was prewetted.In addition, the theoretically computedsettlements, obtained from the consolida-tion tests in Figure 8, are shown for the3- by 3-foot and the 5- by 5-foot plates.(Computing settlement for the 1- by 1-footplate was not practical because of its smallsize. )
The results show that this loess at anatural moisture content of about 10 to 12percent is capable of supporting consider-able load, about 5 tons per square foot,without serious settlement. When thisloess was prewetted by ponding, consider-able settlement occurred after loading.The settlements of the larger plates weregreater because the distributions of pres-sure included greater depths of compressi-
, ble loess than the smaller plates. Thus,these tests clearly show the error that couldarise if the results of field load tests onsmall plates were used directly to estimatesettlements of large footings.
is related to the greater consolidation whichoccurs in these specimens. It is interestingto note that wetting has caused expansion inthe loe ss since the initial density of thewetter loess is considerably less than theinitial density of the dry loess.
FOR LOESS
The important observation is that thewetted material for footings of sizes 5 by5 feet or larger showed increased settle-ment under loads as low as 0.8 ton persquare foot. For safety reasons, this ma-terial would probably only permit loadingsof about 500 pounds per square foot. Incontrast, the dry (natural moisture content)material did not show excessive settlementunder loadings as high as 5 tons per squarefoot.
T he computed values of settlementchecked the actual measured settlementsfairly well for the large size plate. As theplates become smaller, computed estimatesbecome less accurate. It is of interest tonote that the curve computed from consolida-tion tests and the plate load test curve (forexample, the 3- by 3-foot plate) have similarshapes for the beginning part, indicating thatconsolidation is the primary cause of settle-ment of the plate in that range of loading.For greater loadings, the plate load-testcurve shows greater settlement and crossesthe computed curve, indicating progressivesettlement and greater effects of shearfailure.
Results of Ponding Tests in Loess
Ponding of loess may in some cases bedesirable to develop the weakest conditionin the loess prior to construction. This islikely to be most important in hydraulicstructures which will have their foundationsthoroughly wetted by normal operation. Suchponding would permit settlements to takeplace before or during construction. If thefoundations are of low density and suscepti-ble to large settlement, ponding would bevery desirable in preparing them for piledriving. This will be discussed furtherunder pile driving and pile-load testing inloess. Also, ponding of foundations forearth embankments may be desirable sothat settlement can take place as the em-bankments are constructed and thus reducethe possibility of the foundation slumpingas saturation by the reservoir occurs.
20
t\).....
.06
t-=I.L
II-~ .08
~I.LJ...J
1=I.LJ .10cJ)
0 ~It:.: r"-==.~..:.i- -~ "-'=':~ -::::--~F I--~ ~ 1" - -- :.-=;'Il-:..-:::"- -- ,-,'"
. "-IIt'- ""' /1,\
?"--.: ::::,
.-:; . ~: " "r~
-- -~ ,/ -~":"'-7~::: :~~=.:::; ::~~- -.02"" '\', / " ,,/ ---
-""""",,--? " \ l--r-::::- --..
"" ",/ r \ -.....
~ \ >-</,~'
"" \ ""- Not. MOis~~-.-', \ ./ A','"
. :.04I ---Not. Moist -'(",,', ~' A ",_- --Not. Moist- _'hn n_, \<m---Prewetted ~: \1\ .. '. , I \ I'
~n--prewetted-- /I~ ", 'lc<-m--'-- --Pr~wetted- ~ '\ I'xl' PLATE--)! - I .
- \ [\ ' '. - '. - ' ,i :Estimated by i /'" ',,-- \ fEstlmated by t i' ; consolidation r ~ --. ---'--~- --m 'n.--
~ consolidation ~-~ [\
r l test, prewetted: \ \ , t. test, prewetted): '.,
'I i I ~ \ \ 1 I I I "l_--5'x5' PLATE ,\ \ \ 3'x3' PLATE- n)
\\ \", \~ \'
\~
l\
\
\
\ ,,
"
.12
\\
\\\
\~\'\
E XPLANAT tern;?-Assumed failure load
1'\
.14
.16
\~\
\
2 3LOAD TONS PER SQ. FT.
4 5
PLATE - LOAD TESTS FOR A TYPICAL LOESS DEPOSIT FOR COMPARISONOF MATERIALATNATURAL MOISTURE AND PREWETTED CONDITIONS
FIGURE 12
Figure 13 shows the results of surfaceponding in typical loess of low density nearAshton, Nebraska. The two ponding experi-ments are identified in the figure as Site 1,where the loess was about 40 feet deepunderlain by a medium to coarse sand, andSite 2, where the loess was in excess of 60feet deep.
In Site 2, the soil was wetted to a depthof 60 to 70 feet. In Site 1, the loess waswetted to the top of the sand stratum at about40 feet. The wetting was accomplished bycontinuously ponding 1 foot of water overthe areas for a period of 6 to 9 weeks. Mter2 weeks of ponding, it was decided that opendrill holes smuld be made to expedite wet-ting so that it could be accomplished in theallotted time. These drill holes were madein the western one-half of Site 2 and aroundthe edge of Site 1. The locations of mois-ture observation holes are shown in the planviews in Figure 13, and the depths of mois-ture penetration for various numbers of daysof wetting are shown in the profiles. OnSections AA and BB, it may be seen that theopen drill holes definitely assisted wetting.The wetting of Site 1 was somewhat slowerthan Site 2 because of greater densities.After ponding was stopped, Test Holes No.DHP-23 and -25 in Site 2, and DHP-26 inSite 1 showed that wetting was accomplishedto the proposed depths. Forty days of pond-ing were required for Site 2 and 63 days forSite 1.
Example of Prewetting
The foundation of Medicine Creek Dam,Nebraska, was thoroughly wetted by pond-ing and sprinkling before embankment con-struction (described by Clevenger 8 n.Figure 14 is a photograph of the dikes -andthe ponds. Moisture content of the loesswas raised from about 12 to 28 percent infrom 25 days to 2 months during whichtime 33 million gallons of water were used.Settlement measuring points were establishedthroughout these ponded areas. No signifi-cant settlement occurred from saturationalone. Base plates for measuring the foun-dation settlement caused by fill constructionwere installed on top of the loess stratumin four locations, designated BP-l throughBP-4 in the plan shown in Figure 15.
Fill construction on the loess beganabout mid 1949 and was completed in late1949. Base plate settlement measurementswere made periodically from beginning offill construction until July 1954. Three ofthe plates had settled from 0.8 to 1. 0 foot,and the fourth settled a total of 2 feet by1954, as shown in Figure 15. After thereservoir had been full for about 3 years,
T
the settlement curves become flat, indicat-ing that settlements virtually ceased. Thisexample illustrates the use of prewetting topermit a large part of the settlement to takeplace during construction.
Behavior of Piles in Loess
Piles are sometimes used as a solutionto the problem of large settlements of rigidirrigation structures such as powerplants,pumping plants, gate structures, bridges,and appurtenant canal structures. In at-tempting to visualize the suitability of piles,these questions were asked:
1. How must piles be placed since theloess is quite firm and sometimes hardwhen dry?
2. What happens to the pile-bearingstrength when the loess becomes thor-oughly wetted?
3. Can the natural structure of loesssupport friction piles or must there alwaysbe firm bearing?
4. What kind of piles are desirable?
5. Will end bearing piles become over-loaded as consolidation of surroundingloess takes place?
A research study, near Ashton, Ne-braska, of full-scale pile tests was con-ducted to answer these questions. The testsites contain loess which is representativeof that found at many proposed irrigationprojects. Site 1 contained end-bearing pilesresting in firm sand underlying shallow loess.Site 2 contained friction piles in loess greaterthan 60 feet deep. The loess at both siteswas at a medium range of density, 80 to 90pounds per cubic foot. The study includeddriving 44 piles and performing 28 pile-loadtests. Two general types of piles werestudied: nondisplacement (steel-H) pilesand displacement (timber and concrete) piles.Each test site contained an area in whichpiles were placed in loess at a natural mois-ture content of about 11 percent and an areawhich was prewetted by surface ponding to adepth greater than the proposed pile length.
The natural moisture area of each sitecontained the following types of piles andplacement methods:
Group (1)--3 timber piles placed bydriving only or preboringif necessary
Group (2)--3 steel H-piles placed bydriving only
22
MO-9
...::JB
SITE NO.2PLAN OF PREWETTED AREAS
10 5 0 10 20 ~O 40,"
,SCALEOF FEET
N .MO-140DH-P-26
!. MO-7
f~L- 7r MO-BC C
MO-19
SITE NO. I
...::-,8MO-I.
=0-131
MQ-2
ODH-P-25
ODH-P-23
.MO-12 ~~":' . . .MO-IQ ~
""'--MO-II
A .MO-15
t..:>'"
MO-4 MO-S. . . .0-3MO-5--1
A
N
-0
I" I" T000:i ::t!!; :!ii G.S. Elev. 2219
,.,'"'"N, ,
"- "-, ,'" '"Q Q
'"T0
'"
'"T0
'"
rf) v It) U)I I I I
a 0 a 02 := ::!!: 2
'"I02 G.S. Elev. 2219
Elev. 2216.8Elev, 2216.8
14 Days
21 Do s20 Doys
6 Days9 Days
\.iLOEPTHS OF WETTING
OF WETTING
10' 0 40 DoysSECTION A-A SECTION B- B
... ~,
'"0
"--'"
,IG.S. Elev. 2170 5~
':'0
'"
NOTEThese profiles should only be considered
approximate as they are based on observationsot scattered time intervals. The observationswere obtained from rapidly mode drill holes.Penetration holes, DH-P-23,25 and 26. modeafter ponding was stopped, ore the best proofof wetting.
Elev. 2167
216017 00 S
t2150
~I'llIj,' i1,,/,
DEPTHS OFWETTING
2140
2130
I40' (a 63 Days
SECTION C-C RECORDS OF PONDING EXPERIMENTSAT ASHTON PILE TEST AREAS
FIGURE 13
N -!2,"0 00
'"....
2210
2200
t2190
~
2180
2170
2160
PREWETTING OF LOESS FOUNDATION BY PONDING AT MEDICINE CREEK DAM FIGURE 14
,'-- RIGHT ABUTMENT
CUTOFF TRENCH
Note: BP = Settlement measuringplates on the foundation.
200I
0 200 400I I I
SCALE OF FEET
0.00
0.20
0.40
a::
I- 0.60 ~2420
"'a::
~"'en
"'a::I- 0.80 0::2400Z 0
"'....
.. 0
"'Z..J <t
I- 1.00 2380I- ..J
"'..J
en u:
~
Z2 1.20I-<t0Z::>0
'"
'"0 2360
ZQ
I-<t
1.40 ~ 2340
..J
"'1.60 2320
1.80
2.00.
DAM CREST_,
\\\ OPERATION
---\ PER IOD ----
1950'-1952 1953
-BPI1954
..y
PONDED AREA
600I
BP4
BP3
BP2
MEDICINE CREEK DAMFOUNDATION SETTLEMENT INSTALLATIONS
FIGURE 15
1951
25
Group (3)--3 steel-shell, cast-in-placeconcrete piles placed bydriving only or preboring ifnecessary
Group (4)--3 timber piles placed injetted holes
The prewetted area of each site con-tained only displacement piles placed bydriving onJy. At the friction pile site (Site 2),in addition to the single separated piles, acluster of nine piles was driven to determinethe consolidation effect of group pile driving.
Representative piles of each type andplacement method were load tested. Pileswhich were placed in the natural moistureareas were load tested before and after wet-ting the areas.
One of the purposes for testing end-bearing piles was to study downward dragon the piles resulting from the upper loesssettling when wetted. However, this partof the study did not materialize because theloess at this site did not settle as a resultof wetting alone.
Soil profiles and plans of the two testsites are smwn in Figure 16. In Site I, Fig-ure 16a, the loess was from 35 to 40 feetdeep. The Grand Island sand formation,was considered the bearing material, was ,atabout 45-foot depth. Between the loess andthe Grand Island sand was a transition zone ofsand with silt. The loess varied in densityfrom 84 pounds per cubic foot near the sur-face to 90 pounds per cubic foot at lowerdepths. The Grand Island sand was quitedense, having a density of 112 pounds percubic foot.
In Site 2, Figure 16b, the loess wasof considerably greater depth than the pro-posed pile length of 55 feet. The loess from0 to 40 feet was of relatively low density, 80to 85 pounds per cubic foot. This was under-lain by a somewhat clayey, old soil horizon,1 to 2 feet in thickness. Below this old soilhorizon, to 62-foot depth, the loess wasslightly more sandy and denser, being at 85to 95 pounds per cubic foot.
The loess in this study was describedas "silty loess" and the physical propertieswere similar, in respect to gradation, plas-ticity, and mineral contents, to other loessin the Kansas-Nebraska area. Typical gra-dation curves are shown in Figure 17.
The equipment used for pile driving wasa single-acting steam hammer which wasparticularly suitable in a research study ofthis kind because the energy of the blows
"r
could easily be interpreted and maintainedconstant for comparing the driving of onepile to another. The 5, OOO-pound drivinghammer was dropped 36 inches and was op-erated at 55 to 60 blows per minute.
The preboring of dry holes was done bydri ving an open end pipe into the ground forabout 4 feet, pulling it out, and blowing thesoil plug from the pipe with steam. Theprocess was repeated until the desired depthwas reached.
The jetted holes were made by ejectinga straight stream of water at 100- to 120-psipressure from a 1-l/2-inch-diameter jetpipe into the ground using a long vertical pipeguided by the leads of the pile driver. Thejet pipe was worked up and down with shortstrokes and followed the hole down as it wasmade. The finished hole was somewhat ir-regular in shape, varying from 7 to 12 inchesin diameter.
lDad tests were made by stacking stand-ard railroad rails on a loading frame abovethe pile to be tested. A hydraulic jack be-tween the loading frame and the pile per-mitted the application of the load in controlledamounts. The method of load testing isshown in Figure 18.
Results of Pile Driving Studies
The principal results of driving studiesare presented here by means of three graphs,Figures 19a, b, and c. The8e graphs dis-play the record of a representative pile ofeach of the general conditions.
Figure 19a represents the driving rec-ords of the displacement piles placed by var-ious methods in deep loess at Site 2, thefriction pile site. The solid line curve isthe driving record of a timber pile placedby driving onJy in thoroughly prewetted loess.Such a pile drove rather easiJy and graduallydeveloped resistance of about 36 blows perfoot. It was apparent that the denser andmore sandy loess below the old soil horizoncaused resistance to increase more rapidly.The Engineering News formula indicated asafe load of 35 tons for this pile. A similarpile placed in dry loess, shown by a dash-dot line, indicates a distinct contrast indriving resistance. It was only possible todrive this pile 25 feet before refusal resist-ance was reached. The safe load of this pileafter the soil is wetted should be that indi-cated by driving a pile in prewetted loessto a similar depth, and, in this case, by theEngineering News formula, would be about9 tons. A safe load computed from the driv-ing record in the dry loess was 90 tons,which shows that driving in dry loess, which
26
Plies I " Not Plies in Morsi Are0 Prewetfed Area
8 OGALLALA k c e m e n t e d FORMATION Silly Sond ' *. ,Piles In preweitedoreo
J e i t e d . . ~ + O O C t ; ' . ~ ~ ~ ~ ~ ~ ~ ~ ~ A R E A
Timber (PrebaredJ--
'-Steel Shell (Preboredl
( 0 ) P R O F I L E O F S l T E NO. I (END-BEARING SITE,
I ,-OLD SOIL H O R I Z O N - - ;LOVELAND LOESS 'S ILT , SOME S A N D
t G
Plles in Not. Plies in Moisl Area Prewetfed Area - -
; G R A N D I S L A N D S A N D FORMATION
N A l MOIST ARE?, ,PREWETTED
' O G A L L A L A
Timber------- ,' ' I / / Stee l -H - - - - - ' \ , .-Tim be,
Steel shel l (Preboredl' \ ' \ ~ l m b e r Cluster
Jetfed'
20
( b ) P R O F I L E OF S l T E NO. 2 ( F R I C T I O N P I L E S )
+ C-sur1oc.e 1 ~ P E O R I A N LOESS !SILT. LOESS
P R O F I L E S A N D P L A N S O F T E S T S I T E S FIGURE 16
S I E V E S I Z E S
.005.009.019.037.074 DIAMETER OF !ARTICLES I N rnm. I
GRADATION T E S T S FIGURE 17
METHOD OF MAKING PILE LOAD TESTS FIGURE 18
would be subsequently wetted, would giveerroneous indications of strength.
It was evident that preexcavation wasnecessary to place a displacement pile indry loess to the proposed depth. Further,it was found that to place a pile in a dry pre-bored hole, the hole had to be practicallyas large as the pile. A representative pileplaced in a dry prebored hole, shown inFigure 19a by a dash-double dot line, droppedin the hole to 35 feet without resistance andrapidly picked up high resistance from thereto the proposed depth of 55 feet. A pile ina jetted hole, shown by the dashed line, hadgradually increasing resistance from a depthof 12 feet.
Of the three placement methods whichpermitted the piles to be placed to the pro-posed depth of 55 feet, the pile placed in theprewetted loess by driving only was con-sidered to give highest strength becausemaximum consolidation of surrounding soilwas obtained. Piles placed in jetted holeswere considered to be the next be st as thepiles drove easily and some consolidationof adjacent wetted loess was obtained. Apile in a dry prebored hole was consideredto give the least strength because practicallyno consolidation was obtained. Pile testsshown later support these opinions.
In Figure 19b, a similar comparisonis made for piles placed in Site 1, the end-bearing site, where the loess was relativelyshallow. In general, the driving resistancein the loess was similar to that of Figure 19a.The impor:tant consideration is that all pileswhich reached the firm Grand Island sandstratum increased in resistance almost im-mediately regardless of whether the pile wasin a jetted hole, a prebored hole, or in pre-wetted material. Therefore, it appears thatthis sand will adequately support piles placedby any of these methods, provided they aresnugly held in the upper loess for adequatelateral support. The loadings are discussedbelow.
Figure 19c shows the comparison of dif-ferent types of piles, all of which were placedby driving only. The timber pile in pre-wetted material drove easiest. In dry mate-rail, the nondisplacement steel H-pile droveeasiest; however, the effect of greater clay-eyness in the old soil horizon was parti-cularly noticeable. Displacement piles, bothtimber and steel-shell piles, drove withmuch difficulty in dry loess and reached re-fusal values at relatively shallow depths.
Results of Pile-load Tests
Figure 20a, b, c, and d show the re-
sults of load tests on various representa-tive piles in deep loess at Site 2, the friction-pile site. Figure 20a is a load test for adisplacement pile, driven in prewetted loessto a depth of 55 feet. This pile was testedin two series of loadings. The second serieswas carried to complete failure at 170 tons.However, the curve indicates beginning fail-ure at 120 tons. This curve is used as ameasure of comparison for the other piles.Figure 20b shows the test of a displacementpile in a jetted hole. The pile was testedbefore wetting the area and showed similarstrength to that of the pile in the prewettedarea. However,' after the area was wetted,greater movements may be seen and failureoccurred at a somewhat lower load.
Figure 20c shows the test of a displace-ment pile in a prebored hole. Although thestrength of this pile was fair under dry con-ditions with an indication of beginning failureat 60 tons, the pile was considerably weakerwhen the area was wetted. The test indi-cates great movement under very light loads,particularly when comfBI'ed to the pile drivenin the prewetted area. Figure 20d shows atest for a pile placed by hard .driving to onlya shallow depth in dry loess. This pileshowed high strength with only slight move-ment under dry conditions, but after wettingthe area, complete failure was obtainedunder 60 tons. In summary, these tests showthat great loss of strength can be anticipatedif friction piles are driven in dry loess orplaced in prebored holes in dry loess whichis subsequently wetted.
Load tests in Site 1, the end-bearingsite, were all of high strength, indicatingthat firm bearing on the sand was entirelysatisfactory for all placement methods. Non-displacement steel H-piles in this area werefound to be satisfactory, when embedded 5feet into the bearing sand. Since it was pos-sible to drive steel H-piles in dry loess andnot displacement piles, a nondisplacementpile was considered the only type which waspractical to place in dry loess by drivingonly.
Summary--Study of Piles in Loess
1. From this overall study, opinions ofgeneral requirements for piles and piledriving in loess have been summarized inFigures 21 and 22 on the basis of moisturecontent and density of the natural loess.
2. Friction piles in deep loess had con-siderable difference in load-carrying capa-city, depending on their placement method.
a. For displacement piles in loesswhich will be wetted in the future, pre-
28
~0 0cp -50 .01 .01
600 20 80 100 120 140 160 180 200 .02
BLOWS PER FOOT(b) Driving record of displacement piles in Site I} the f- f-
"' "'end- bearing site ~.03 ~.03, \ , '0 f- f-Z Z
~.04\
~.04 \10
"'\ "'
\-' -'f- \ f- \
20 ~.05 , ~.05t
<f) <f), \
~30 .06 .06I
'" \a40 I
.07 .07
50
.01
.02
180 200
f-lU .03IU
':-....~ .04::E
"'-'f-~ .05<f)
Beginning offailure ---50
600 20 80 \00 \20
BLOWS PER FOOT(0) Driving records of displacement piles in Site 2, the
friction pile site .06Complefe --/
(oi/tlre ---
.0710
.080 25 50 75 100 125LOAD-TONS
(a) Pile driven in pre wetted area.
150
(c) Comparison of piles of different ty pes placed by driVingonly in deep loess
.080 25 75 100LOAD - TONS
(c) Pile placed in dry pre bored hole.
600 150140 160 180 200
.02
f-W.03
"'"-,
f-Z
"'.04
:E ,
"',
-' 'f- ,~ .05<f)
,
'''''''...
.06 -""'.....
.07
175 .080 25 50 75 100LOAD-TONS
(b) Pile placed in jetted hole.
II\
150 175125
175.08
0 25 50 75 100 125LOAD-TONS
(d) Pile driven in dry natural material.
175150
DRIVING RECORDSFIGURE19
COMPARISON OF PILE LOAD TESTS FOR FRICTION PILES IN LOESSFIGURE20
,..
r: 70
"-:>uCI:...n.u) 80m
-'I)0-
l-II)Z
~ 90
-''"CI:::>I-'"Z
r: 70
"-:>uCI:...n.u) 80m
-'I)0-
l-II)Z
~ 90
-''"CI:::>I-'"Z
60
100
1100
60
100
1100
(I) Firm end -bearing is required for piles.(2) Piles are generally required for most rigid structures.(3) If end -bearing is not available, 20 to 40 feet of end-
embedment in loess of densities above 80 pcf isrequired.
Friction piles of adequate embedment are permissible.(30 to 60 feet total embedment is generally required)
(I) Ordinary structure need not have piles.(2) Critical structures may use short friction piles,
generally requiring less than 30 feet of totalembedment.
10 15
NATURAL MOISTURE
CAS E 1. When higher moistu re contents are anticipated.(As in the case of hydraulic structures)
(a)./:
f2) p. Is ft)1":'
COl}dl .I":
f'e lIes ref/fl' 1/"0 1'101}
/"ICf; '10 IId,o I&/" o/"e I"ed el}d, I(.10
ib01}
PIII}
It/'eef eO"liJ 00Sf Uel}e" 1&,,/).40"1
/'ef/". 60 ,..Ies ot c:Oess
oie0/ Is I} "IUld 0111'" Ile~ "u
I"ed) eefOf' ;'def/,,:;dl!IOI}
'J>..COI}0/~Q'0e:: .0VO/~:
'b/""c:::I"'ed
(I) 01'01 e e "< 101}./lIS /"e Ie 8S.t';?) O/"dl e00e 00e(j o/" f/"I,,~ 201'0
C/"I!'/}O
"1' Co d0el} 0el}f dPI;. ICOI Sf"" IIdl!; fls O"e
/; es / Sf/" Cf"/,, 01} 7>.. gel} Pe/"01'0/
. (Uel} "Cf" e I} "<I. e/"o/;: 01$s'e00 e/"o// /"e0 eed 'J' 10/e
8q, l' /e 01' 1}0f .0el}f) Ss fh "se hol/.
01}.1 Sho"f e
PI/e0... f'". s'eef 'IC;-
.of' 101}
20
MOISTURE CONTENT- PERCENT
CASE 2. When natural moisture contents are notex pected to increase.
REQUIREMENTS FOR ADEQUATE SUPPORTING CAPACITY OF PILESFIGURE 2/
30
t-ZWUII:W 200..
t-ZWt-Z0U 15WII:::>l-(/)
0::;;
10
-'00:II:::>t-oO:Z
30
25
I-ZWuII:W 200..,I-ZwI-z8 15
WII:::>l-(/)
0::;;
10
-'00:II:::>I-00:z
'11\\\\011\\~teO o\\\~
\\\~~\)e Q
of\~I\\~ t\\\~?\\eS . IIWI \)~
\\\Q\O
.\"It '011\"'1'0. \\~
"\ fe "'eofl\)es {f\\\ V o\\\~. 0
IIseO,Of
lIeS \0 Iof\~\\\~ oe~e\OQe
.
11\0\)e \\\e\\\
Q\\~
\)~f\\\~ IS. '11\\\
e\\\\\~ to\ Q\\eS. -OISQ\~t\ 0\" It~ e\\O-\)eOe\\e\fO\IO\\
'11?f:I~Q\~~e~fe:\\\~f'll~~~0\)\ ~f':~\e\\:o~o'/.\\\\\)\\\Q
'\':;,\ee\-
,Q\ote '\eS
\\\~~ fO\)~\)\ \ot\\\~,~ \\\o~
\)e\\\e\\'1
QIIIseo,
Q\\'0
\\\Q
,\sQ\~te O\\\~ IS
\. '0\feS\)
\\\'00,:'1 D\, o~\~I\\~O tltO,ee
0 'o~\)e
\~~ooeQ
\oQeo.,tlt1 \)e:'o 0 \)e \)se
\~\\\~0\\
. \'0oe~e
'0\\011\ \)~ o~ O~I\\~ \\~~e
e\'I\\\~ \\\ Q\\e\ Q\~teo ., e\\O-\)e \\ \\\~~,11?~e.'II \ote~e \\\o~
\). e'l\eO
I \\\e\\Oeo,OISQ
-'0QI\es
'\ ~\)e \ ~eto\\\
'\':;,\ee\ tltO,ee'l'''I\es
\\\~\'0
\\0'( ~~\\\~.,~
:.0 \0 \\\e\\\ ~ "'O~I\\~ \\0 -\)e
\~te Q~ev. \\\ e
,:'1D\sQ
II~\\ M~ 'o~,I~
,tlt1(:>.\\;~ss\\)\eIIse
060 70 90 10080
DRY DENSITY- LBS. PER CU. FT.
CASE 1. When higher moisture contents are anticipated.(As in the case of hydraulic structures)
30
25
. ,,\Ote\\\;~\\tll\'I~olS~ '( "I'\\0\\- ,\\\011
\\\ O~ 0 'III. \oc.e\\\e \)e
Qloc.e , ,\t\)I\~DIsQ \\\o~ \~,
'I 01',,1\eS ."0\\ ~eo,
~ M\~I\\" 11\ ~ '0\\ 0'\)~ 0 '11\\\\0 e\fO\1
Q\oc.e \eQe\\
\eo~ '\)e oeo.lIO oeo.lI~ '\)o~eo
'0~1\eS ~~, '\)\e
,of ~o~\Oeo
~\\\e
Q~e
':;,\ee\- \~\\\~ 0\\ oes\~O11'0eO
QfeQ\\\ 0' \\e 1\\
,11'\)~
o~o~
'\)e \\eS, ,'\)\~ '\)e0 '(\\e 0 . o\)\e
Q
~e'l\\\\~ ~e\\\e\\'1 Qo~ QOSSIO\)e'jo\\'0\
oesl\\ \\0\ '\)e
,~1 o\sQ\O '\)of\\\~\\\ o\)'(~\\\e
e \\\e\\\0
loess'III
('e
1:'1D~~ Qf~~o\\O\\
IS'11011\0~1~'II\\e\\\~:
s'If\)c.\1I ''Qe\\e e\\\\\~
'I'c.o'\)\e
0''(
\\o\e, \\Q~e'll
~QfOC.I fO\\O\\
(:>.\\\\O\)~ \)\~ \\0 0\ oQe,tlt1
''0QfO\)O
"'~\\of\\\
I \eO VJ'IIe\
060 70 90 10080
DRY DENSITY- LBS. PER cu. FT,
CASE 2. When natural moisture contents are notexpected to increase,
REQUIREMENTS FOR PRACTICAL PLACEMENT OF PILESFIGURE22
31
wetting the soil is recommended sincehighest strength will be obtained be-cause of maximum consolidation of sur-rounding material.
b. Placing displacement piles injetted holes may be possible in certaininstances since holes smaller in diam-eter than the piles may be made and somewetting of the adjacent loess is obtainedfrom the jet water.
c. Placing displacement piles inpre bored holes in loess at low naturalmoisture content is not recommendedsince virtually no consolidation of sur-rounding material is obtained on driv-ing, and in this case, after the founda-tion is wetted, a weak bond between thesoil and the pile can be anticipated. Inaddition, considerable weakening of theloess surrounding the pile can be ex-pected.
d. The use of displacement pilesplaced by driving alone in loess at lownatural moisture content is not recom-mended, since driving is extremely dif-ficult and adequate penetration cannotbe obtained. Also, the load-bearingcapacity after wetting the soil is low.
3. Nondisplacement steel H-piles canbe driven in loess at low natural moisturecontent, but adequate embedment in per-manently firm material is necessary sinceshearing resistance in loose wetted loessis quite low.
4. End-bearing piles resting on firm-bearing material such as dense sand werefound to be satisfactory regardless of theplacement method, provided they aresnugly held for adequate lateral support.
5. Both types of displacement piles,timber and steel shell, had similar drivingand loading characteristics. The strengthlimitations of the pile itself would mainlygovern which pile would be desirable.
6. For piles placed in loess and assum-ing material underlying a pile foundation iscompetent, the following rules are recom-mended:
a. When loess is at densities below80 pa.m.ds per cubic foot, piles must havefirm end-bearing or a substantial amoo.ntof embedment of the ends in dense mate-rial of permanently high shearing resist-ance.
b. When loess is at densities from80 to 90 pounds per cubic foot, frictionpiles will generally be satisfactorily sup-ported only by adequate embedment in theloess, based on wetted conditions.
c. When loessisatdensitie'sgreaterthan 90 pounds per cubic foot, moder-ately loaded structures may be sup-
ported without piles, and if piles arefound necessary for critical structures,relatively short lengths may be used.
AS A
DENSITY OF LOESSCRITERION OF SETTLEMENT
In the construction and operation of lb-reau of Reclamation dams, reservoirs, andcanals, the effects of saturation on the set-tlement of loess is of particular importance.A simple criterion for judging the suscepti-bility of loess to settlement is of value inpredicting problem areas and the require-ment for foundation treatment. Approxi-mate limits of density have been used for theloess of the Nebraska-Kansas area, sinceother basic properties are generally simi-lar. Extremely low-density loesses ap-parently subside under their own weightwhen they are subjected to saturation. Mod-erately low-density loess settles appreciablywhen subjected to loadings and saturation.These critical coriditions of loess frequentlyoccur because of low natural moisture con-tents and arid and semiarid climatic condi-tions.
Limiting values of density for varioustypes of soil, which indicate the suscepti-bility of a soil to subsidence on saturation,can be defined using the following basicprinciples:
(1) The liquid limit represents the mois-ture content, based on a standard labora-tory test, at which the soil is in a weak orapproaching liquid condition.
(2) If the natural density of the soil islow enough that the porosity is more thanthat required for the liquid limit moisturecontent to saturate the soil, it is logicalthat the soil could develop this weakenedcondition and could be susceptible to sub-sidence as added moisture approached thisamount.
32
(3) On the other hand, if the porosity ofthe natural soil is less than that requiredfor the liquid limit moisture content, suchas for a dense condition, it is logical thatthe soil could become saturated and stillbe in a plastic state and thus not subsideunless loadings are great enough to com-press it. (It may even expand if it con-tains expanding-type clay minerals. )
(4) Denisov, 9/, an investigator andauthor of publications on loessial soilsin Soviet Russia, has used an equationwhich may be applicable in correlatingdensity and liquid limit:
= Porosity at liquid limitKct Porosity at natural density
When the ratio, Kd, is less than1, the soil has possibilities of sub-siding on saturation.
Based on the above principles, a chartor graph can be drawn with dry density ver-sus liquid limit moisture content, as shownin Figure 23. The curves on this graph, asshown in the figure, represents the conditionat which the porosity for a particular drydmsity equals the porosity at the liquid limit.For practical purposes the curves are con-sidered as a single line, because the positionof this line varies only slightly according tothe specific gravity of the soil. Points abovethis line represent conditions easily suscep-tible to subsidence upon wetting, and thefurther the points are above the line, themore susceptible the soil is to subsidence.Soil with points below the line are not sus-ceptible to subsidence unless loadings areapplied to compress the soil.
In Figure 23, liquid limit data from Fig-ure 6 are plotted for samples having naturaldensity determinations. These data showthat most natural density values of loessialsoils fall above the dividing line. Also, asmentioned previously in regard to Figure 6,most liquid limit values of loessial soils fallbetween 25 and 35 percent.
On the basis ofr.
revious observationsby Holtz and Gibbs 1 , without regard to thedensity-liquid limit relationship, it was con-cluded that the probability of extreme settle-ments of loessial soils was not great for den-sities above 90 pcf. In viewing the plot inFigure 23 for values of liquid limit between25 and 35 percent, this conclusion is con-sidered reasonable. Later experiments inponding and field plate-load testing, Fig-
ures 13 and 12, respectively, indicated thatloessial soils having densities between 80and 90 pcf were weak and suceptible to ap-preciable settlement under loading, but theydid not subside by wetting alone. It shouldbe remembered, however, that this pondingmay not have fully developed saturationthat would occur in extended canal and res-ervoir operation. Such prewetting suffi-ciently weakens the soil to permit settle-ment to take place as relatively small struc-turalloads are applied, as shown in Figure12. Pile driving for pile foundations caneasily be accomplished under such conditions.It has further been observed that loess withdensities below 80 pcf is extremely suscep-tible to settlement upon saturation.
Subsidence has been observed on somecanals in loessial soils. Figure 24 is a pho-tograph of some subsidence that has occurredat a canal lateral on the Columbia BasinProject in Washington. The density and liq-uid-limit were determined on undisturbedsamples in nonsubsided soil next to the sub-sidence that has occurred. These data areplotted in Figure 23 by x-points. Also shownin Figure 23 are data from Meeker Canal inNebraska (asterisk points) where similarsubsidence has occurred. The data fall wellabove the dividing line; three points werenear the value of 80 pcf density, and allpoints were below 90 pcf. These soils arerelatively low in liquid limit, having valuesfrom 22.0 to 27.5 percent which cause adensity of 80 pcf, as shown in Figure 23 tobe critical in respect to the average loessialsoils.
It has been observed that the naturalloesses of the Nebraska-Kansas area fre-quently have moisture contents below 10 per-cent. Pile tests and plate-load tests haveshown that such loess resist very large loadswithout appreciable settlement. Pondingtests in the field and wetting of consolidationtest specimens in the laboratory indicatedthat when moisture contents were raised toabove 20 percent, appreciable settlementscould easily take place under loading. How-ever, as can be seen in Figure 23 about 35percent moisture content or greater is neededto saturate low density loess. Therefore,increased moisture content is an importantcharacteristic for permitting settlement totake place, but density is the primary featurewhich governs the susceptibility to settle-ment. The critical ranges of density valuesare governed by the type of soil which is in-dicated by the liquid limit.
33
60
IL.
0Q..
70I>-I-- X(/)
Z 80wa
>-Q:a 90
0-I<ta::::>
I00I-<tZ
110
10 20 30
LIQUID LlMIT-CIJo40 50 60 70 90
r(
120 \, \
" "Specific gravity = 2.70 i"
.t r- Lines of 100 "10 saturation--Specific gravi y = 2.60 )
. --- Trenton Dam and Railroad Relocation11---Ashton, Nebraska Pile Test Area0 ---Amherst Dam. ---Courtland Conal~ ---Cambridge Conal'* -- -Meeker Conalx -- -La tera I PE41.2
1 Col urn bia Basin Project
NATURAL DENSITY VS. LIQUID LIMIT RELATIONSHIPFOR LO E S S IA L SO I L S
FIGURE 23
34
AN EXAMPLE OF SUBSIDENCE OF LOW-DENSITY S O I L AFTER SATURATIOY\I FROM C A N A L
OPERATION
FIGURE 24
CONCLUSIONS Certain features which govern the soil
mechanics and foundation properties of loess have been established in this monograph.
to settlement on saturation. The higher the plotted points a r e above the lines shown the more susceptible the soils a re to subsidence on saturation. The critical values of density will vary according to the plasticity of the soil. However, silty loess is similar in i t s plasticity characteristics and the liquid limit values range from 25 to 35 percent. Because of this similarity in basic properties along with consolidation and strength studies, it i s concluded that initial (in place) density i s a good criterion for indicating the proba- bility of settlement for structures. The fol- lowing are recommendations of logical bound- ary values for density of silty loess. More sandy material will be somewhat more criti- cal.
The basic propert ies of loess in the Nebraska-Kansas a rea a r e similar. This similarity has been noted in other locations. However, some trends of changing charac- teristics have been observed in relation to the distances of deposits from their probable source and in relation to climatic environ- ment s which prevail. The loess of the Ne- braska-Kansas area is predominately silty (75 percent of a l l samples tested). There a r e some instances of more sandy loess and, on the other hand, some clayey loess. From a general soil mechanics viewpoint, loess appears to have rather uniform character- is t ics of gradation and plasticity.
a. Silty loess having densities of Figure 23 is a chart on which natural l e s s than 80 pounds per cubic foot is
density and liquid limit values can be plotted considered loose and highly suscep- to aid in evaluating the susceptibility of a soil tible to settlement upon wetting.
b. Silty loess having densitiesfrom 80 to 90 pounds per cubic footis moderately loose to medium denseand is moderately susceptible to settle-ment. particularly for critical or heavilyloaded structures. Extreme saturationmay result in subsidence.
c. Silty loess having densities ofabove 90 pounds per cubic foot is some-what dense and may be capable of sup-porting ordinary structures withoutserious settlement. Settlement will beof normal characteristics with loading.
d. A more general criterion hasbeen used for earth dams in which 85pounds per cubic foot is used as thedivision between low and high densityloess or the division where special foun-dation treatment is required for lowerdensities.
Because loess is predominately a siltysoil and has at least a small amount of clay,there is pronounced adjustment in particle-to-particle strength as a result of adjust-ment in moisture content. In the Nebraska-Kansas area, natural moisture contents arefrequently below 10 percent, and the naturalloess is frequently capable of supporting verylarge loads. However, when moisture con-tents are raised due to wetting, as in foun-dations of hydraulic structures, the mini-mum shear-strength conditions and full set-tlement with load should be expected. Thefollowing are recommendations of boundaryvalues for moisture content:
a. Moisture contents below 10per-cent are considered very dry and maxi-num dry strength and high resistance tosettlement should be expected.
b. Moisture contents from 10 to 15percent are considered somewhat dry,giving moderately high strength.
c. Moisture contents from 15 to 20percent are approaching moist conditions.
Acknowledgment
d. Moisture contents above 20 per-cent are considered somewhat wet tomoist, and will generally permit fullconsolidation to occur under load.
e. Experience has shown that mois-ture contents of near 25 to 28 percentcan easily be obtained by surface pond-ing, and about 35 percent moisture (de-pending on density) is required for com-plete saturation.
Certain shear characteristics have beenestablished:
a. Dry loess at natural moisturecontents of less than 10 percent gen-erally has inherent shear strengthsresembling cohesion which permit nat-ural loess to stand on very steep slopesfor considerable heights, although itmay be of low density. This cohesivestrength is apparently due to the smallamount of clay which gives the loess aparticle-to-particle bond. This co-hesive strength has been observed tobe as much as 15 psi. Generally, it isabout 5 to 10 psi for the Nebraska-Kan-sas loess. The slope of the shear re-sistance curve (tan ~) is about 0.60 to0.65. Such values of cohesion and tan 0permit loess to stand on steep (or nearlyvertical) slopes 50 to 80 feet high.
b. Wetted loess is greatly reducedin strength. Experience shows that co-hesion is generally reduced to less than1 psi when the loess is wetted and evenfor initially dense loess, it becomes lessthan 4 psi. An im}X>rtant finding in thesestudies is that high pore pressure de-velops in wetted loess as consolidationtakes place, and therefore the develop-ment of the additional shear resistanceresulting from friction may be restricteduntil drainage takes place. Drainagemaybe sufficiently slow to cause lateralmovement of embankment foundationsduring normal construction periods.
Much of the petrographic and geologicnarrative was written from original notesand work compiled by members of the Petro-graphic.Laboratory Section, in particularthe late M. E. King.
36
LIST OF REFERENCES
1. Holtz, W. G., and Gibbs, H. J., "Con-solidation and Related Properties of Lo-essial Soils, " Special TeChnical Publica-tion No. 126, American Societyfor Test-ing Materials, 1951.
2. Hobbs, W. H., "TheGlacialAnticyc1onesand the European Continental Glacier, "American Journal of Science, Volume241, No.5, May 1943, pages 333-336.
Smith, G. D. (1942), "Illinois Loess--Variations in its Properties and Dis-tribution, " University of Illinois Agri-cultural Experiment Station Bulletin490.
3.
4. Swineford, Ada, and Frye, J. C., "AMechanical Analysis of Wind-blown DustCompared with Analyses of Loess, "Symposium on Loess, 1944, AmericanJournal of Science, Volume 243, No.5,May 1945, pages 249-255.
5. Tuck, Ralph, "The Loe ss of the Mata-nuska Valley, Alaska, " Journal of Ge-ology, Volume 46, No.4, May-June1938.
37
6. Hobbs, W. H., "The Origin of the LoessAssociated with Continental GlaciationBased upon Studie s in Greenland, " Inter-national Geographic Congress, Paris1931, Compte rendu, 1933, tome 2,fasc. 1, pages 408-411.
7. Davidson, D. T., and Sheeler, J. B.,"Cation Exchange Capacity of Loessand Its Relation to Engineering Prop-erties, " Special Publication No. 142,American Society for Testing Materials,1953.
8. Clevenger, W. A., "Experiences withLoess as Foundation Material, " Trans-actions American Society of Civil Engi-neers' Volume 123, 1958, page 151.
9. Denisov, N. J., "Settlement Propertiesof Loessial Soils, " a book published bythe Government Publishing Office --Soviet Science, Moscow, 1946 (Trans-lated and abstracted by K. P. Karpoffand H. J. Gibbs, Bureau of Reclama-tion, Denver, Colorado, June 1951,page 14).
GPO 842939