m’tmmmm

102
g 434
'fWTHSF

 

 

.mnw» .-.. , a >.. . con'QA-me-‘a‘ .- “'W. W-

m ROLE Of 50“. IN EARTH-FILL DAMS

Thad: for the Dogma of l. 5.
MICHIGAN STATE COLLEGE
Arthur E. Biflol
1949

THES‘g

 

 

 

The Role of Soil in Earth-fill Dams

A Thesis Submitted to

The Faculty of
HICHIGAN S ATE COLLEGE
of
AGRICULTURE AND APPLIED SCIENCE

by

Arthur E. Bittel

Candidate for the Degree of

Bachelor of Science

June 1949

“1:515;

 

TABLE OF CONEEI‘TTS

iii

TAB LE OF C ONTI‘INTS
CHAPTER PAGE
I o lNTRODUC T I ON 0 o o o o o o o o o o o o o o 1

II. A.BR EF HISTORY AED REVIEW OF EARTH
DALI FLJLTU'RES o o o o o o o o o o o o o o o 3

III. DESIGN AND COHSTRUCTION. . . . . . . . . . 8
IV. LABORATORY EVESTIGATIONS OF SOIL. . . . . 19
Compressibility . . . . . . . . . . . . . 21

Shear . . . . . . . . . . . . . . . . . . 24
Permeability. . . . . . . . . . . . . . . 30
Optimum Uoisture Content. . . . . . . . . 55

V. CONSTRUCTION HETHODS . . . . . . . . . . . 39
Rolled Fill . . . . . . . . . . . . . . . 40
Hydraulic Fill. . . . . . . . . . . . . . 50

VI. Failures . . . . . . . . . . . . . . . . . 55
Hort Peck Dam . . . . . . . . . . . . . . 57
Alexander Dam . . . . . . . . . . . . . . 59
Clendening Dam. . . . . . . . . . . . . . 61
Horse Creek Dam . . . . . . . . . . . . . 65

Delhi Dam.. . . . . . . . . . . . . . . . 65

VII. Summary. . . . . . . . . . . . . . . . . . 66

DBLIOGRIXID}1YO 0 0 O O 0 O O O O O O O O O O O O 69

LIST OF FIGURES

FIGURE

1.
2.
3.
4.
5.
6.
7.

Dam profile with berms. . . .
Downsville Dam. . . . . . . .
Shear Resistance Test . . . .
Scrapers. . . . . . . . . . .
Bottom Dump Wagon . . . . . .
Sheep's-foot Roller . . . . .

Hydraulic Sluicing Operation.

PAGE
12
15
28
42
43
46
51

iv

CHAPTER I

INTRODUCTION

The particular purpose of this study was to invest-
igate the functions of soil in earth dams and the various
means used to attain these functions. It is further honed to
present at the end of this paper, suggestions, which might
guide later students towards particular lines of study.

The subject of earth dams is perhaps one of the most
hazy in the minds of students. And the average citizen of
the United States would hardly have a thought that soil
could be used as a building material other than in the form
of aggregates.

Additional publicity will not harm anything and will
perhaps save the citizen some money in the form of taxes.

There are two terms, consolidation and compaction,
which are used in construction wark and which are often
confused. They should be defined at this time to eliminate
possible future confusion. Consolidation refers to the
gradual increase in density of the soil due to static loads,
generally its own weight or the action of water. Compaction
refers to the increase in density of soil due to moving

loads such as rolling, tamping, etc.

CIL’IPTER II

A BRIEF HISTORY Psi-TD RWIEW

OF EARTH DA}: FEATURES

A dam has been defined as a barrier to contain or
keep back a body of water. During the days of early history,
mankind had no access to engineering literature; but, never-
theless, he was pushed into making various types of crude
structures to protect himself from the ravages of the
swirling water. The earliest recorded notes concerning the
restraint of the flow Of water come from the Nile Valley,
the Euphrates and Tigris River region, and parts of India.
By this early date these areas were either already showing
signs of being parched or had too much water in the wrong
places from.floods. It was the logical conclusion that
artificial means must be resorted to if the people concerned
were to stay on their paternal lands.

In India, the lower ends of natural depressions in
the ground were walled off with earth and water was
collected behind the walls to be used as needed. These
undertakings were in the nature of community or area enter-
prises. Such bodies of water and their impounding walls are
now called tanks in many lands, including the United States.
The Veranum Tank in India has a water area of thirty five
square miles restrained by a dam of silt-clay soil twelve

miles long. This tank has existed for unnumbered centuries.l

How did these structures to store water come into

being? Early writings show that man was building low

 

1 John C. Page, "Dams," Encyclopedia Americana, VIII, p.453.

 

embankments of loose earth, perhaps using a few rocks and
timbers that could be carried to the site on his back. His
methods included carrying dirt to the site in baskets and,
at a later stage, tamping the dirt with his feet to compact
it, thus slowing down seepage and reducing settlement.
Possibly even wooden or stone tampers were used. These last
two items definitely showed up in construction a few
centuries later. After these lowly beginnings came the
basketwork dam of moderate height, fifteen to thirty feet,
and now the great giants of today.

Single purpose dams have in most cases given way to
multi-purpose projects which call for larger and larger
amounts of stored water for power, recreation, flood
control, etc. The ever expanding industries of today demand
more and more power'which is most effectively used in the
form of electricity. more power demands in turn more feet
head of water stored at the dam.

The reasons that earth was used in ancient times were
controlled by economics and environment. The dirt was close
at hand, at the most only a basket-haul distant. All that
was generally needed was a little loosening. Ancient man was
no stranger to the uses and presence of dirt as he not only
lived on it, but beside it on his skin, and under it in the
form of a cave or a hut. Also, at the beginning, there were
no tools available with which to use other materials.

The reasons just stated still hold with the exception

of the last. The study of soil mechanics has resulted in
the advance of many theories as to why the earth reacts in
certain situations and has enabled us to use earth to a much
greater advantage than was ever thought possible. In the
last fifteen years, earth moving machinery has advanced in
great strides to the point where earth can be moved ten, and
more, cubic yards at a time and at a Speed of thirty miles
an hour: But the fact still remains that earth dams are se—
lected because the necessary fill is available nearby and
can be utilized more economicly than the expensive process
of hauling in cement and reinforcing steel. Then too, "No
type of structure that man builds to confine water can
compare in permanence with earth dams."2

This paper is concerned with the relatively large
dams used for flood control purposes, power, and irrigation.
Low embankments and levees will not be discussed although
many of the things stated here will apply to them as well.
In the past the public has seen news items about

levees during flood times, and when an earth dam.was brought
to their attention it was considered merely a matter of sand
bags and piles of earth. This misconception resulted in
several failures in early earth dams. These dams were not
designed by qualified engineers and competent supervisory

personnel were not placed in charge of the work. This

_ r
—

0
2 Burr Bassell, Earth Dams, p. 6.1

 

problem has been partially solved in the last few years, but
the education of both the public and the engineering pro-
fession must continue until the place of the earth dam is
understood by all. V
Besides the points already suggested for the selection
of an earth dam, there is one feature which lends its use
very well to the landscape engineer or the conservation
officer, in situations where looks are one of the guiding
factors. This feature is the ease with which the earth dam
can be blended in with its surroundings. If the country is
primarily rocky, both slopes of the earth dam can be covered
with rock riprap. If the country is predominantly grassy or
wooded, the downstream slope can be completely sodded and
the upstream slope sodded down to the predicted high water
level, with the rock riprap below that covered by the water.
Whereas, the concrete dam will always stand out with its
characteristic color and in many cases, a downstream face

which is very badly stained.

CID-‘LPTER III

ES IGN AI‘TD C ONSTRUCTION

Present dams are often a unit of a large overall
program, examples of which are the Tennessee Valley
Authority and the missouri.Basin Project. But whether for a
large organization or for a small private power company,
once it has been decided that a dam.is needed in a certain
area, the procedure then followed is essentially the same.

The first phase of the project is a series of pre-

liminary studies, often called -paper reconnaissance. The

 

best maps of the area are procured, usually from the United
States Geological Survey or from the united States Coast
and Geodetic Survey. from.these maps an outline of the
catchment area is prepared. This outline is the basis for
the gathering of all information regarding rainfall, floods,
and other data which would influence the characteristics of
the flow of water in this particular area. If the rainfall
and rate of flow information cannot be obtained from
reliable sources, this is the time to establish rainfall
gaging stations and flow stations throughout the catchment
area. A few months to several years may elapse between these
preliminary studies and the final design of the complete dam
structure; and any accurate information obtained during this
interval will be welcomed by the designing engineers.

The reservoir basin is selected with land economy and
water storage the controlling factors. Land economy is not a

dominate factor in the western states but it has appeared in

the Tennessee Valley Authority and the missouri Basin Project.

10

in the eastern part of the country, land values are high and
the area to be inundated must be carefully calculated. All
land not a part of the public domain must be purchased from
the private owners, sometimes at a very high cost, even
without consideration of the legal expense involved.

In addition, the total water surface to be exposed to
the air must be considered. In some parts of the country, an
evaporation rate of ten to twenty inches of water a year can
be expected with no detrimental effects. Swinging toward the
other extreme in the Southwest, an evaporation rate of
eig ty to one hundred inches a year is not uncommon. In this
area water is more of a concern to the people than the cold
war now raging in other parts of the world. A loss of this
proportion can be enough to reduce the water behind the dam
to a point where the committments cannot be met.

A consideration of the preceeding information will
narrow the possible dam sites to a minimum..Aerial photo-
graphs can be used very effectively at this point to clear
up any questions which might arise concerning any particular
point.

With the larger types of dams, the xpense is not the
greatest factor in influencing the final decision as to the
final site. The amount of water that must be stored to
prevent disastrous floods or to provide a sufficient amount
for irrigation are primary, then the other factors will

follow along.

11

At this point preliminary, on the snot, investi-
gations will be initiated. These consist of the preparation
of topographical maps of secondary order of the possible dam
sites, random borings, and test pits in likely areas; and
then, back in the office, tests of materials as found at the
various sites. finally fough estimates of the cost of the
various types of dams at each location are figured to aid in
the final decision.

The selection of the site can be influenced by the
results of the borings and test pits; but these two items
will have more effect on the shape and size of the dam. The
decision to employ an earth dam can be influenced by many
things other than have already been mentioned. The borings
might show that bed rock lies one hundred fifty to two
hundred feet beneath the layer of river fill. This would
require an enormous amount of excavation and concrete to
construct a safe masonry dam. Yet an earth dam can be built
which will have a good safety factor.

This can be done by spreading the base of the earth
fill out until the resultant soil pressure has been reduced
to the point where settlement is of minor importance. In the
Spreading out process, the earth dam.is exaosed to another
danger, that of erosion from heavy rains. One way to reduce
the erosive action of the rains is to establish berms on the

slopes of the dam as shown in fig. 1.

12

Dam Profile With.Berms

     

---
---‘------- ¢ —"

Original Ground Line

Figure 1.

The berms are pitched towards the center of the dam
to collect the rainwater and conduct it by means of ditches
to a masonry gutter which leads the water safely to the
bottom of the slope. As can be seen, the berms are used to
Spread the base of the dam without flattening the lepes.

If a diversion channel is to be dug, a large amount
of earth must be moved. The most economical method would be
to use it right in the dam. There are always exceptions to
the rule. The Anderson Ranch Dam has a Spillway channel
constructed in solid rock. Of course, if the bedrock is
overlain by only a couple feet of tepsoil, the selection of
an earth dam.is dependent on the finding of good barrow pits
at a reasonable hauling distance. In this case the cost of
hauling cement, aggregate, and steel reinforcement from the
railhead must be balanced against the cost of hauling the
fill dirt from.the distant barrow pits.

With the site selected, a detailed topographic map,

based on the preliminary map, will be built up with

13

additional field work. This map, or COpies, will be kept up
to date as the construction progresses, showing the true
picture of the area.

Systematic borings and test pits will be made at the
dam site, the diversion channel site, and at the prouosed
barrow pits. The results obtained will influence the design
and the methods of construction. A coordinated system of
borings is more likely to reveal faults and slippage planes.
These planes will be hard to find in the foundation. The
extent of the possible fill material upstream from the dam
can be obtained from the borings. It is a good thing to
deepen the channel for a ways upstream.from the dam.as this
will provide a place for the silt to settle out. It will be
dependent upon whether or not excavation in front of the dam
will weaken the foundation.

All of the preceeding information is now correlated
and placed in the hands of the designers. Detailed plans
are made for each part of the structure, including the dam
itself, the Spillway, intake, forebay, outlet, powerhouse,
access roads, bridges, temporary or permanent, and other
details peculiar to each dam. from,these plans the engineers
prepare the engineer's estimate. This estimate is used to
compare with the bids entered on the construction of the
dam.and also to form a basis for the financial maneuvering
which must now begin if not already in progress. In many

cases the original apprOpriations will not be large enough

l4

and more money must be procured.

With the letting of contracts, the construction
starts. Access roads, bridges, and living quarters are
started initially. As soon as the working force can be built
up, the preliminary operations are started at the dam site.
One of the first problems is to move the river out of its
course so that the foundation area may be prepared. This can
be done by means of a diversion channel, a tunnel, or by
cofferdams. The method of construction of an earth dam.is
such that the use of cofferdams is limited, with the except-
ion in the case of a cofferdam thrown across the breadth of
the river upstream of the main dam to divert the water from
the construction area. Then during construction, the coffer-
dam.can be made a part of the dam at the upstream toe.

The damsite is cleared of all vegetation and, if the
inundated area is not too large, it would pay to remove the
trees and large brush in the reservoir basin. A.benefit from
this activity would be the use of the lumber and timbers
that could be sawn at the site. The clearing Operations
would lessen the possibility of stoppage of the intake or
damage to the turbines by driftwood.

Depending upon the design, the foundation area may
have all soil stripped down to bedrock, or trenches may be

sunk<hywn to the rock or to the desired depth for the

erection of cutoff walls. figure 2 shows some details of a

foundation trench with the construction of a cutoff wall in

 

DoWi'SVIllEfiML

Figure 2.

 

15

l6

progress. in the distance can be seen wingwalls on the sides
of the embankments. These are actually cutoff walls extend-
ing as shown to prevent percolation through the ends of the
dam. As can be seen in the figure, the cutoff walls are con-
crete with steel sheet piling as forms. Sometimes the cutoff
walls are made of steel sheet piling only.

Another form of cutoff wall is the clay puddle. The
term puddle does not necessarily mean mixing a gooey mass
of clay and water. One type of clay puddle consists of a
mixture of clay, sand, and gravel mixed with water to a
stiff consistency and then tamped into place. This will form
a watertight membrane. The function of a cutoff wall sunk
either to bedrock or to a predetermined depth below the dam,
is to reduce seepage through the foundation and to lower the
line of saturation.

Very careful preparations are made to assure a bond
between the first layer of earth fill and the foundation. In
the case of bedrock, compressed air is used to clean the
rock; and, at this point, the plans might call for grouting
all cracks and crevices in the rock. This will depend upon
the condition of the bedrock. The concrete cutoff walls are
usually extended a few feet into the main body of the dam or
almost to the top to aid in keeping seepage down and to
provide stability.

With earth covered foundation, the top layer of 8011

is scarified so that the first layer of fill will be able to

17

mix with it.

If the Spillway is a part of the dam, its construct-
ion should be accelerated so that the concrete walls will
always be higher than the fill surrounding them during
compaction. This will enable special small rollers and hand
tempers to compact the soil right up to the wall. Hand
tamping with hand operated vibratory hammers is an expensive
process. To cut the hand work to a minimum, a small sheeps-
foot roller has been developed which has no frame on the
outside except for the hitch to the prime mover. This
enables the roller to be moved right up next to the wall.

The fill for the impervious section can be moved into
place by carryall, truck, conveyor, or combinations of these.
The movement of carryalls and trucks should be routed care-
fully on the fill to prevent too much travel in one section.
The large rubber tires of the vehicles tend to form a crust
on the fill with little compaction underneath the surface.
many specifications call for the crusted section of the fill
to be disked or broken in some way immediately before the
succeeding layer is laid to assure maximum.bond. A mois-
tening Operation migit also be carried out before the new
layer is laid. The layers should be sleped away from the
center of the dam to shed rain water readily. A well
compacted surface leped in the correct manner will shed a
night's rain and be ready to receive more fill by the

beginning of the next day's operations.

18

Again to assure bond and stability between sections,
the semi-impervious and pervious sections are carried un
along with the center core. The semi-impervious sections
usually consist of moistened rock screenings and the
pervious sections are of rock.

The dam is comnleted by laying riprap on the unstream
slope for wave and ice nrotection and either sodding or

laying rinrap on the downstream sloue.

1

‘he operation of the dam consists of maintaining the

DrOper flow Of water through the outlets and snillways and

-

L

taking all measures to prevent the water from overtOpping
the dam as this has been disastrous in all cases. The
construction is such that large masses of water moving over
the top will erode masses of fill away and collanse others.
haintenance consists of keeping the outer surfaces of
the dam in repair and a schedule of inapections to assure
the continued safety of the dam. One of the things to look
for is seepage from.the downstream toe or moist areas on the

downstream face. This is an indication of undesired high

percolation and must be taken care of.

CHAPTER IV

LABORATORY INVESTIGATIONS OF SOIL

As stated earlier, soil mechanics has played an
important part in the construction of earth dams in the past
few years. The latest findings have enabled us to build the
dams bigger and higher and in locations where, at the turn
of the century, an engineer would have been called crazy to
even consider the possibility of constructing a dam there.
An example is the Garrison Dam on the Missouri River. A rock
formation was found at a considerable distance below the
surface under deposits of well consolidated silt and clay.
This material is subject to much deformation under heavy
loads, but the dam.was designed as an earth dam and allow-
ances made for two-way settlement. Today even the textbooks
on soil mechanics admit that their subject is not an exact
science, but enough information can be obtained from the
application of soil mechanics to help us measureably in the
design of earth dams.

when the soil samples are procured from the site

investigations, four major laboratory tests are performed

upon the samples: compressibility, shear, permeability, and
optimum moisture content. Several other tests are performed
to identify the various soil types encountered. They are:
void ratio, Specific gravity, Specific weight, porosity,
water content, degree of saturation, effective size, and the
Atterberg limits. The four major tests mentioned first, how-
ever, are the ones that actively affect the design of a dam

by making possible a prediction of the behavior of the soil.

21

PART I
COMPRESSIBILITY

The first test, compressibility, can be defined as
the measure of the volume change produced by an increase of
pressure acting on the soil. The volume change can be
accomplished by a decrease in the volume of the voids, or by
a reduction of the contained water through squeezing the
water out of the soil. The soil itself is not being comp-
ressed but rather is shifting and deforming under the
pressure to meet the changed conditions. The water must flow
'through the soil to escape, thus compressibility is contin-
gent upon the permeability of the soil.

Pressure can be applied to the soil in three
dimensions, but the science of soil mechanics is limited at
the present to one dimension. Consolidation was defined at
the beginning in order to definitely limit the word in its
meaning. The reason for this is apparent when one looks

through several soil mechanics books. Taylor in Fundamentals

 

of Soil Mechanics says "This one-dimensional compression
involving a slow escape of water, gradual compression, and a
gradual pressure adjustment is called consolidation."3 Here
compression and consolidation are linked together. In other
writings compression and compaction are used interchangeably
with no distinction made between the two. lt would be best

to use compression as defined first in sonnection with soils

 

3. D.W. Taylor, fundamentals of Soil Mechanics, p.212w

’30
N

to eliminate confusion.

There are several different types of equipment which
have been designed to determine a coefficient of compress-
ibility. hey can be divided into two groups. it is necesary
in all cases to duplicate the moisture content and density
which will be found in the field. In the first group are the
tests in which a sample of soil is placed into a cylinder
and then confined at each end in such a manner that only
water can escape. Pressure is applied and the deflection of
the moving end recorded. The information is plotted on a
load deflection diagram and a coefficient determined for
the particular pressure to which the soil will be subjected.
The coefficient is defined as the volume change per unit

volume divided by the pressure. C‘s—Av , in which A is the

hd
measured deflection in inches, d is the thickness of sample
in inches, and n is the applied pressure in pounds per square
inch. A steel cylinder is used with one movable end. This
method keeps two dimensions constant and allows only one to
change.

In the second group are the unconfined compression
tests. An example of which are the tests performed during
the site investigation for the ban Jacinto Dam near houston,
Texas. These tests are performed in the field immediately
after the samples are removed from the core sampler. The
speed used is to cut down evaporation losses which would
affect the test if not retarded. The same principles apply

as in the first group except the sample is not rigidly con-

25

tained. As long as standard conditions are maintained, a
load-deflection diagram.may be drawn up and the desired
information ex racted.

The equipment needed is a simple bench with a pulley
arrangement for applying the load and a means of reading
deflections. A sample of sandy or gravel soil would offer
no resistance on this machine and have little or no values.

The advantages of this method is its simplicity and light

P.

t

H.

weight of the testing frame making deal for field wor .

The effect of compression on two different types of
soil illustrates the reasons why the compression test is a
part of the standard testing procedure. Uhen a load is
applied to a sample of sand, 95% of the ultimate deflection
occurs within one minute of the application. on the other
hand, a sample of clay will take consid“rable time to reach
the ultimate deflection; there will probably be an apprec-
iable amount of deflection between thirty minutes of

application and one day of application.4

 

4o Ibido, p. 215‘9.

24

In addition to the vertical forces producing
compression in the foundation, there will be horizontal
forces acting on the dam tending to slide the dam or
portions of the dam downstream. These horizontal forces
produce a condition or a result called shear. This shear can
be accomplished in two ways. In one, part of the mass
remains stationary and the rest of the mass moves relative
to the first. in the other, one part of the mass moves in a
direction opposite to the direction of movement of the other
portion. This movement tends to take place along a plane or
a series of closely related planes.

in general, vertical shear planes are curved due to
the internal friction of the soil. This internal friction is
thought to be one of the elements of the resistance to shear
along with the surface tension of the capillary water con-
tained within the soil. This is illustrated by the fact that
a dry soil has less shearing resistance than when it is wet.
Sand which usually is considered as having no shearing
resistance, will develop a small shear resistance when
moistened. Shear is a phenomenon such that when it occurs,
failure has taken place or is impending.

The possible rupture or shear plane may be in one of
three locations. First in the dam itself. This is generally

manifested in the form of a lateral movement of a portion

25

of the dam. it is possible to produce this condition in all
soils by increasing the height and slope of the earth fill.
flattened slopes are a means to stabilize the fill in an
earth dam. This location will require tests of the fill
material after it has been compacted to determine the
resistance to shear which has been develoaed through
compaction.

Secondly, shear may occur between the base of the dam
and the foundation. his would be very difficult to test.
The safest way is to be guided by the eXperience of others,
and do every thing possible to assure the maximum bond at
this juncture.

Thirdly, the failure may occur within the foundation.
When the foundation consists of rock, a series of well
chosen samples should be taken to determine the condition of
the rock in the foundation. The presence of small or large
fault lines will have an effect on the design of the dam. A
study of the angles of inclination of the fault lines will
give the experienced observer a clue as to the possible
points of shear. Good design is the best corrective measure.

A foundation material of soil also has to be well
sampled. Shear tests conducted on the samples can lead to
predictions as to future slides and also to possible
corrective measures which might be taken, such as wooden or
steel pile walls, or excavation of the offending material.
The latter might be done in the case of a layer of fat clay

which lies a short distance below the surface of the original

26

soil. This condition is more than likely to result in a
slide of a portion of the dam due to shear taking place
within the layer of low shear resisting clay.

The determination of shear resistance is as approx-
imate as the other tests used in the study of soil
mechanics. Taylor in Fundamentals of Soil Mechanics5, lists
seventeen different factors affecting shear strength. They
are of such diverse nature that the engineer in the field
can only set up a shear testing procedure and then make sure
the testing standards are kept the same at all times.
Included in these standards should be one which requires
that the structure of the samples to be tested should be the
same. This will require undisturbed samples or samples
compacted to the same density. Another inportant item is the
fact that confined water will carry part of the applied load.
Thus, the water content must remain the same for each sample
during the test, in order for comparable results to be
obtained.

There are two shear tests which are perhaps more
adapted for use in the field than other more complicated
ones. In one the sample is an undisturbed core which is
tested in the same cylinder that originally received the
soil. In this way disturbance of the soil structure is
reduced to a minimum. The cylinder is eSpecially made for

this test. It consists of two cylinders about three inches

 

5. Ibid., p. 364-5.

27

long separated by a short cylinder seven sights of an inch
long. The three inch cylinders are clamped in a jig and a
deformation dial is mounted to bear on the top of the short
cylinder. At the bottom of the short cylinder is attached a
loading tray. After the setup is complete, the dial is
zeroed and increments of loads are added until failure. Each
increment is left on the tray for a period of ten minutes.6

The information gained is plotted on a graph as shown
in figure 3, and a value is obtained for the ultimate
shearing resistance in pounds per square inch. This test is
for cohesive clay soils only and would not give satisfactory
results when used with sand. No attempt is made to keep the
contained moisture constant other than using the sample
immediately after the paraffin is removed. In this test the
clay soils retain the water while the more granular soils
would let the water escape.

The second shear test is similar to the first except
that there is single shear and a compressive load can be
applied at the same time the shearing force is applied. The

application of the applied forces is as shown:

COMPPCSJIOP’

  

J b an r
Shear Line.

 

Shcar "—'——fi‘ __._-_,__'

 

 

 

6. William.S. Housel, Applied Soil Mechanics, p. 61-6.

 

 

 

 

 

 

28

v.
Q
\
>{
3 77078 In Mina/e:
. / 6’ 7 9 IO
\t I T I f I
8 l
20 '
5" Shear . 2.80;;é/sq. in.
a“
2%-
\
v
t
\l
:-
.t
l.
l i 1 1 ‘ n 1
o 5 /o 15‘ 20 35 50 35'

01%r8/76’3 , 3 m”). from /0 mm. read/n3.

SHEAR RESISTANCE TEST

FIGURE 3 .

29

This test is representative of the forces that will be
applied to the foundation and fill material , and perhaps
is more used in connection with earth dams than the first
test. With the compression load constant for each run, the
plotting of the results could be accomplished in the same
manner as in the first test. Several runs could be taken
varying the compressive load in each run.

The distinctive feature of this test is that
disturbed samples may be used. Each sample may be compacted
to the desired density using the proper moisture content.
Provisions can be made for the escape of the water through-
out the test, depending upon what is desired of the test.
The test can be used to examine samples of sand. Too much
value should not be attached to the test in this case. Sand
has very little cohesion and therefore little shear
strength. This is one reason why earth dams are not generaly
built on a foundation of sand. There are excewtions, but if
a strata of sand is encountered in the foundation, a careful
study should be made. Perhaps it can be left there or
possibly removed. It is a point of weakness and will affect

the design.

FEET II I
PERIEABILITY

The third major test is m,asuring the ease with which
water travels through the voids in the soil. This prOperty
is called permeability, and, in the case of earth dams, is
one of the most important tests to be made, both for the
foundation and the dam itself. r“he purpose of the dam
determines the amount of seepare that is to be allowed
through the dam. A‘watertiaht dam could be used for
irrigation or power where the water is costly and must not be
wasted.'Whereas, a flood control dam is used to lower the
flood crest and seepage is not too important as long as the
dam is not endangered. All earth dams are saturated to some
extent, but if the water finds a path by which it can move
through the dam and come out on the downstream face of the
dam with any velocity at all, trouble will start and, unless
preventative measures are taken, a failure will occur. In
some cases the water moves through the foundation with
enough velocity to carry small particles of soil along with
it. The flow path is enlarged and nipina results. This will
show up at the downstream toe in the form of sand boils. The
usual cure for this is the dumping of rock at the point of
boiling until no further movement is detected.

A machine used to determine the permeability of
samples of soil is called a permeameter. These machines are

divided into two groups, the constant head and the variable

50

31

head designs.

The constant head permeameter is one constructed so
that the head or water pressure is kept cons ant at both the
point of influx and efflux. It requires a more elaborate
setup to perform the test than the variable head permeameter
and is not too suitable for tests of impervious soils as it
takes such a long time to complete the test.7

The variable head permeameter in its simplest form
has a graduated glass pipe fixed in position over the
apparatus containing the sample to be tested. A valve is
turned and a record is kept of the time required for a
certain amount of water to pass through the sample by the
action of gravity only. This test can be simply set up for
use in the field and will give accurate results.

Because of the stratification of soils, samples should
be taken both vertically and horizontally as the seepaae will
take both paths to pass obstacles. The horizontal samples
are the ones which will interest us the most; but in some
cases, a layer of soil will run vertically for a distance
and then run horizontally. If this happens, the tests are
run on the samples of each type to give an idea of what to

expect. Test pits are of great help here.

 

7. Fred L. Plummer, etc., Soil Mechanics & Foundations,
p. 89.

52

Items affecting permeability are: grain size,
properties of the liquid, The void ratio, the soil
structure, amount of gas present, and the temperature.8
For earth dams the liquid will always be water and the
properties will not be changing. The amount of gas present
would not be too hard to determine in the field, but
keeping the amount constant during a series of tests would be
rather difficult in this situation.

The other items, grain size, void ratio, soil
structure, and temperature are of major concern and a
reasonably accurate account must be kept of them,
eSpecially the temperature. The others are controlled during
the compaction prior to the test if disturbed samples are
used. Accuracy of the highest degree need not be strived for
as it is common knowledge that widely varying coefficients
of permeability will be obtained in the same soil from
samples taken just a few feet apart.

It is best to use undisturbed samples for the tests
but, if they are not available, disturbed samples may be
effectively used as will have to be done in the case of the
compacted fill. The method used is to compact the samples to
varying densities to represent the possible combinations,
obtain the permeability coefficients, and plot them on a
graph to the void ratios. The resulting curve will provide

information as to what to expect.

 

8. D. W. Taylor, Fundamentals of Soil Mechanics; p. 111-2

33

Darcy's Law, k :.£13 , can be rearranged so that the
quanity of flow through aqsoil may be predicted. k, as the
coefficient of permeability, can be expressed in feet per
minute, with A as the area in square feet, 1 as the
hydraulic gradient, and t the time in seconds? The
difference in head between iIflux and efflux can be desig-
nated h and the length of travel of the water as L.

" Then i : H . The Darcy formula furnishes very reliable
results, if the value of the hydraulic gradient i is small'l10

The knowledge of permeability can be used to construct
a flow net for a dam. A flow net consists of a cross section
of a dam and foundation, both upstream and downstream.from
the dam. The net consists of flow lines and lines repre-
senting equal pressure heads through and under the dam.

The flow lines are always perpendicular to the equal
pressure lines. A lot of eXperience is needed to construct

a reasonably accurate flow net; but, when once accomplished,
the information gained can be used to great advantage in the
design of a dam.

The upper boundary line of flow is called the line
of saturation and represents the most critical flow line to

be obtained.

 

9. Dimitri P. Krynine, Soil Mechanics, p. 52-3.

10. lbid., p. 53.

54

If this line of saturation comes out of the dam above the
downstream toe, danger is eminent, as explained earlier,

and safety measures must be taken. Dense blankets of earth
on upstream slope and floor, dense core walls, cutoff walls,
interior drains, toe drains, and toe blankets are all
measures to keep the line of saturation below the downstream

toe.

35

PART IV
OPTIEUM MOISTURE COITEIT
As areas for excavation are determined, the soil
therein is tested to determine its suitability for use in
the dam. lf suitable, it is stockpiled for future use. in
most cases it will be necessary to find borrow pits in order
to obtain enough fill. The test for optimum moisture content
will be run on the prOSpective fill material to determine
some of the compaction characteristics of the soil mixtures.
The procedure was originally intended to duplicate the
degree of compaction possible to be obtained with standard
rolling equipment under field conditions. The present
rolling equipment is now constructed so that the pressure
per square inch obtainable can be changed at will. Again,
experience is the best teacher in performing the test and

determining the specifications for the field compaction.

3
f

The test stems from the rel aion that there is one

C

U)
[—1.

Q:

water content at which maximum en tv is reached by a

particular degree of compaction. ”he first step of the test
is to take a disturbed sample, pulverize it throughly, and
then dry it.

The equipment is a standard cylinder of two parts
and a tamper which is sometimes used with a guide. The
cylinder is assembled and filled one third full with the
dried earth. it is then tamped, using a certain number of

prescribed blows. Two more layers are compacted in the same

manner, after which the top part of the cylinder is removed.

36

The excess soil is struck off and the cylinder is wei hed
to find the net bulk weight of the soil, from which the

density is computed. A small sample is then removed for

DJ

eterminine the water content. Th is procedure is followed

“'_)

0

until the density of the compacted soil b _ins to decrease.

After the first weifihine and the sample is aken, water is

('0
0'

added to the soil sanpl y incre eme ents of a Specified per-
centage of the bulk dry ht o. This water mus be throuchly

mixed into the sample before the next compaction is begun.

1
{i

As with the other tests described, there are many
variables present. The number of blows per layer or the

1

height of hammer fa ll can he clia.n§ed and a e;fferent optimum

moisture content trill be oota ined. The type of support for
the cylinder during tamping will have an effect on the
results. The National Perk Service prescribes a sand box

support to duplicate the conditions in the field.11

I

The weight per cubic foot of the sa mple ootained
with each comgaction in the test is converted to compacted
dry weight per cubic foot. The dry weight densities are
used so that all tests can be referred to a standard.

is then plotted on a graph with the

F"
:3

The informat o

‘

unit reim_nt p

L-J

.otted against their reSpective moisture per-
centases. The graph will show two curves, one for the unit

\_4

dry weights and one for the unit vet weights.

 

11. National Resources Committee, Low Dams, p. 298

57

The high points on the curves will represent the
optimum moisture content. it has been established in the
field that it is best to stay one or two percentage points
on the dry side of the ontimum.moisture content for best
results during construction. Some clays when too wet and
compacted tend to form small shear planes. If these are
distributed throughout a whole layer of fill, a danaerous
condition exists and will probably cause a failure. Yith
a dryer soil, it is easier and takes less time to add water
than it does to dry the soil in case it is too wet. For this
reason many borrow pits have to be abandoned because they
cannot be drained properly. This possibility should be con-
sidered when investigations of the borrow pits are
proceeding. Sometimes an adequate drainage system.will
salvage the pit and save expensive hauling of the fill, or
it may be possible to loosen the fill dirt and let it lay
for a few days to dry out before using it.

The Anderson Ranch Dam used a belt feeder to move
the fill from the borrow pit to the site. The fill in most
cases had the proper moisture content, but at times it was
too dry. It was then necessary to Spray water onto the fill
at the discharge end of the conveyor system. An ineXpensive
method compared to that used at other sites. The impervious
material consisted of a sandy clay, of which 95% passed a
one quarter inch screen and 55% passed a two hundred mesh.
The density obtained under standard compaction was 119

pounds percubic foot with an Optimum moisture content of 12.5

38

The pervious material used was an alluvial sand and gravel.
Not so much care was taken with this material, and the
material from slides was also used in the pervious zones.
It was compacted by sluicing with no effort made to control
the amount of moisture present.12

Up to now no comment has been made as to the cost and
size of the laboratory unit. This can vary from an outfit
costing a couple hundred dollars to a setup costing twenty
or thirty thousand dollars. The size will in almost all
cases vary in proportion with the size of the project.
Minimum.equipment will include an accurate balance, seives,
containers, tools, and some Special small equipment for the
classification of soils. In addition there will be the

special items of equipment already mentioned.

 

12. D. S. Walter, "Reclamation's Biggest Earthfill Dam,"
Engineering_mews-Record, V. 139, 18 Sept 1947, p. 396-402.

40

ART I
ROLLED FILL

There are, at present, two methods of consolidating
or compacting the earth fill during the construction of
earth dams. They are the rolled fill method and the
hydraulic fill method. In addition, there are combinations
and variations of these two methods, but none are used
extensively and will not be discussed here.

Thirty and forty years ago, the hydraulic fill method
was much used. This was because in most cases the fill
could be moved cheaply and faster by water, and the consoli-
dation was more complete than was possible with the smooth
rollers and other equipment used at that time. The
inventions and innovations introduced in the design of the
earth moving equipment and in the compaction equipment has
reduced the advantage of the hydraulic fill to the point
where the rolled fill method is used almost entirely on
small dams and to a greater extent on the large dams.

The rolled fill method can be more effectively and
easily controlled by the laboratory. Whereas, the hydraulic
fill method always has the possibility that fingers of
pervious material will be extended through the core during
the sluicing operations. The limits of the hydraulicked
dense core cannot be as well defined as in the rolled fill

method.

41

The first step in the construction of an earth dam by
the rolled fill method is to determine the dry weight den-
sity which is considered desirable for the dense core of the
dam. The weight which is to be above a point in the fill
should be considered in the determination of the dry weight
density. Thus, the desired dry weight density would vary
from a heavier weight at the bottom to a.1ighter weight at
the top of the dense core as the fill progresses upwards.
The procedure for the dete mination of the dry weight
density has already been discussed. This density will be the
goal towards which all the compaction operations will be
aimed.

As mentioned earlier, the foundation should be care-
fully cleaned before the first layer of fill is spread. All
loose and weathered rock, trees and bushes, and roots are
removed. The fill dirt is brought to the site by trucks,
carryalls, or by conveyor belts. The trucks and carryalls
can partially Spread the dirt as they unload. The conveyor
belt will need some sort of supplementary transeortation in
order to move the fill dirt over the whole length of the
site. A large capacity rubber tired wagon called the
Mississippi Dump Wagon or Euclid Bottom Dump Wagon is used.
This vehicle can move very large loads over short and medium
distances. Fig; 4 shows the carryall or scraper in Operation.

Big. 5 is a view of a bottom dump wagon.

42

 

Part of R. A. Heintr’l "Cater-pillar” Zoned Equipment fleet, building highway
near Moscow, Idaho. In the background a D8 pulls 1 "Caterpillar” No. 80
Sat-per on a Minn-length haul. The lcrnper in the foreground in being push-
looded by a D8. In a moment the "Coterpilln'" Dielel DWIO Tractor will
roll ad with it to a distant till.

Scrapers

Figure 4.

 

Bottom Dump Wagon

Figure 5 .

44

The fill is usually spread to a depth of ten to
twelve inches and is then compacted to a death of six to
eight inches. Dozers and road graders are used to Spread
the fill before compaction. The first layer of fill is very
carefully compacted. This is a necessity as one of the most
probable routes for seepage through the dam is at the
juncture between the foundation and the fill. This is
especially true for a rock foundation. An earth foundation
can be plowed up or disked before the fill is Spread to
assure a bond.

Succeeding layers are spread and compacted in the
same manner as the first layer. if the vehicle traffic over
the fill tends to form a surface crust, it may be necessary
to disk these areas before the next layer is Spread, again
to obtain a bond. One means of avoiding this particular
trouble is to route the traffic over the surface of the fill
in such a manner as to minimize the traffic over any one
section.

Another point for possible failure in the dam is at
the line between the concrete structures and the fill. One
attempt to avoid this trouble is to construct concrete wing
walls extending out into the fill from the main body of
concrete to act in the same manner as cutoff walls. But no
matter what is put up in the line of concrete work, the fill

should be carefully compacted in these areas.

45

One means of Speeding up thisoperation has been to have on
hand a few sheep's-foot rollers which are smaller versions
of the larger types used on the main fill. These miniatures
are constructed so that there is no framework on the outside
of the roller except the hitch to the prime mover. With the
use of these mi gets, it is possible to compact the dirt by
roller within a very few inches of the concrete. The final
few inches left are finished off with the aid of pneumatic
air tempers operated by hand. The special cuts down the
amount of time needed by the hand tempers, thus cutting

down on some of the costs. Another way to assure a tight fit
between the concrete wall and the fill is to build a slight
batter in the wall, so that as the fill is compacted, it is
wedged outward from the wall, thus eliminating any tendency
for the wall and the fill to separate.

The sheep's—foot roller mentioned above is one of the
reasons for the increased popularity of earth dams. figure 6
shows a view of a sheep's-foot roller in action. The roller
consists of a circular drum on which is attached a number of
projections which appear similar to small legs with a slight
elongation on the front edge which look like small feet. As
the roller is turned, the feet and then the legs move down
through the loose soil to compact a small portion of the
soil at the bottom of the loose layer. The projections are so
arranged that they will not enter the same spot previously

touched by the roller ahead.

This International TD-18 Diesel Cr;

and matched 10-yard scraper is improv
2-mile stretch of farm-to-market road f7

owner—Wayne County, Nebraska, Ct
Highway Commission. The TD-18 is di
ditches, sloping banks and leveling tom

 

Sheep's-foot Roller

Figure 6.

46

Also the arrangement is such that only one row of
feet are in complete contact with the fill at any one time.
This assures maximum.load per foot. These rollers act in the
same manner as the hooves of animals, whence the name. In
fact, before the advent of modern machinery, bands of horses
and herds of sheep were driven over new fill in order to
achieve compaction at some dam sites. The only improvement
has been the mechanization to gain more control over the
Operation. Further refinements are accomplished by removing
projections, adding them, or by adding sand or water to the
interior of the drums, in order to increase or decrease the
pressure per square inch out at the ends of the feet.

The other earth moving equipment used on the jobs are
power shovels, bulldozers, road graders, draglines, and
mechanical loaders. These machines are the same as used in
other types of construction and will not be discussed here.

As the various Operations on the fill are carried out,

Ho

it s necessary to maintain constant check on the effective-
ness of the compaction. This is done by taking samples from
the compacted portions of the embankment and performing the
test for the dry weight density.

The procedure followed is to use a cylindrical
sampler which can be forced into the earth and then with-
drawn with the sample inside. To obtain the exact volume

of the sample, sand is placed into the hole from a

previously measured quanity.

48

The amount of sand used to fill the hole to the original
level can be used to compute the volume of the sample. From
this point on, the test is identical with those mentioned
earlier.

if the density has varied from that desired, it may
be necessary to add or reduce the amount of water in the
fill, or to increase or decrease the compaction produced by
the equipment. The moisture content should be carefully
controlled by the laboratory, using a value slightly lower
than the optimum moisture content. To vary the compaction
mechanically, it will be necessary to either vary the number
of passes of the compacting equipment, or to vary the
pressures obtainable from them. Too great a variation would
be cause for plowing or dishing and recompaction.

There is a danger from over-compaction that should be
recognized; this is especially true with impervious material
high in clay content. Over compaction will compress the clay
to the point where shear planes will be formed, and, if this
condition prevails all through a layer of fill, there will
be a danger of slides as the embankment is carried higher.
The laboratory should be on the watch for this condition,
and if found to exist, it should be taken care of before
any more material is deposited on the fill. The bad layer
can be plowed up and thoroughly pulverized before being
compacted again. If the moisture content is at fault, the
material will have to be Spread out for drying before it

can be used again.

Once the embankment has been built up away from the
foundation, it will be found desirable to slepe the too of
the fill away from the center of the dam. The compacted
earth will shed water very readily, and, if given the slope
of two or three percent, a very heavy and sustained rain can
be drained off so that construction can proceed within a
very short time after its cessation. If the too had been
slowed towards the middle, the water would have collected
there and in a short time, traffic would have made an
impassable quagmire at the center.

The preceeding material has concerned the impervious
portion of the earth dam. On each side of the impervious
core will be placed a semi-impervious section and then a
pervious section. Their purpose being to contain the
impervious core and to protect it from outside dangers.

The latter two sections are free draining. The pervious
section will consist of graded rock and the semi-impervious
section will usually be of the rock screenings plus other
graded material if needed.

because of the worous nature of these fill materials,

their compaction is more simple. it matters little if the

Ho

semi-impervious fill is placed too wet, so it 8 usually
thoroughly Sprayed with water before compaction. The rock
fill can be dumped in place from trucks and then smoothed

up to grade with a dragline.

HYDRAULIC FILL

The hydraulic fill method of constructing earth dams
J . .

50

is still used at times and probably never will be completely

abandoned. its distinctive method of placing fill will
always be the most economical at some locations. Actually
there is a method of fill called the semi-hyraulic method;
but this method is very crude in comparison with the two
methods analyzed here and should not be used if at all
possible to use a better method. it consists of dumpina the
fill dirt into the pool of water from trucks or shovels,

1.

rom tee coarser
n "il

.0
.L

with no attempt made to separate the fines
materials. It is true that some consolidatio. w 1 take
place, but no separate dense core will be built up and
percolation will not he hindered. This method could be used
on the construction of a stall water tank on a form, but it
should not be used on anything laraer.

The most efficient use of the hydraulic fill method

1

will take place when a good supply of water is on hand, and
when the borrow pit is located at an elevation hieher then
that of the dam site. The fill will be loosened hydraulicly
at the pit by means of a stream of water from a pressure

nozzle. A view of this operation is shown in Figure 7. The

mud is collected in a pool and is sluiced down through pipes

to the dam site.

 

L .
l

r. . comma
smart Paonucr
Lil-QWER. CO5L..- .-

Veteran G-E Motors Drive the Pl:

{ Still on the iob after

   
   
   

Hydraulic Sluicing l

One 14,000,000-yard sand dune will be -
nozzle to make way for the construction
rate activated sludge plant. This unique I
depends for its pumping power on six 8
tion motors built back in 1934 for cons
After serving successfully on that proj‘
nected for a different voltage to go back
six days a week.

Electric drive, which offers the maxin
imum maintenance, helps contractors bu
Whether you buy or build construction t
tric representative to tell you about the in]
drive. Apparatus Dept., General Electric C

\G _V -

Hydraulic Sluicing Operation

Figure '7 .

51

52

Before actual sluicing operations begin, it will be
necessary to build both ups ream and downstream banks or

dikes of the fill dirt to contain the stilling pool. This

Ho

nitial Operation is similar to that described in the rolled
fill method, including the care taken to secure a bond with
the foundation.

A pipe line will be laid on the inner side of each of
these banks near the top. These pipes will have oweninas in
the sides at intervals which can be onened and closed. The
flow of the hydraulicked material can be directed to any
point desired with the use of these gates.

As the fill laden water leaves the pipe, it flows
over a beach for a short distance and enters the stilling
pool, where it tends to move towards the center of the pool.
The instant the water leaves the pine, its velocity is
retarded, until at the center of the pool, there is no move-
ment. As a result of the falling velocity, the laraer
particles tend to settle out first in the area next to the
pipes and the finer materials are carried towards the center
of the pool. These finer materials will be in susnension in
the peel in the middle of the depression. As these materials
settle out, a very dense mass will be formed.

The limits of the dense core will be determined by
the actual area covered by the still water. This is a very
difficult Operation to control and all dams with a

hydraulic core tend to be indefinite in regards to the

53

dividing line between the impervious core and the semi-
impervious layer next to the core. Shoddy construction
methods can easily lead to trouble because of this feet.

As the stilling pool nears the tOp of the outside
shells, the tops of the shells are built up and the sluicing
pipes are raised to the new elevation. This procedure is
repeated until the desired elevation is reached for the core.
The dam is usually tepped off by the rolled fill method.

As has been explained, a very dense core can be
obtained. The necessary items needed to make a hydraulic
fill dam one of the best is plenty of competent supervision
and inspection. The progress of the stilling pool must be
watched carefully, and if it is found that the water is -'
carrying too many fines that won't settle out, a remedy
should be found at once. A change in borrow pits might be
necessary. Usually some of the still water in the middle of
the pool is pumped out at intervals to prevent a semi-liquid
core being formed due to the particles which won't settle
out.

If the water and borrow pit cannot be found at the
desired elevations, other means can be used to fill hydrau-
licly. The fill can be sluiced out and pumped up to the
dam site. The fill can be trucked, shoveled, or run on a
conveyor belt to a sluice box, from which it can also be
pumped to the site. This method is liable to have a

bottleneck develop at the sluice box.

54

What happens is that the water currents in the box do not
cover the whole area of the box interior; consequently some
of the material settles out and in time forms a series of
baffle plates within the box, slowing down the currents
some more. The remedy is to test models of sluice boxes
until an efficient shape is found, and then to make sure,
arrange for several nozzles to be attached to the box. If
any dead areas then develop, water under high pressure is
introduced to prevent settling.

The fill material must be of the right sizes. Because
of the method of placing used in the hydraulic fill dam, the
fill dirt must be of non-uniform sizes so that there will be
the larger particles to form.the beaches on which the finer
materials can be washed down to the center. Too much of the
large sizes will leave a porous core; and too much of the

smaller sizes will lead to an unstable dam section.

CHAPTER VI

FAILURES

56

There is one method of learning that is successful
where others fail. That is the method of learning from the
observation and study of the mistakes and failures of
others. This is a means of gaining experience second hand;
but this method can be very successfully used when a study
of earth dams is being made.

Today's dams are more massive than yesterday's; as a
consequence, they receive much more public attention when a
failure occurs. Then too, an earth dam does not seem as safe
to the layman as the concrete dam. It can be proved on paper
that one is as safe as the other, and through careful design
and construction, they can be made that way.

A few earth dam.failures will be presented on the
following pages. An effort was made to confine these
examples to ones in which the fill was the cause or indirect
cause of the failure. Many earth dams have failed because of
faulty design, inadequate Spillways, drain pipes not collared
right, and other items of design. ho analysis will be made

of these last.

57

trim Foylrfv .. 32C::1 #1
The fort Peck Dam on the Missouri River is an example

of a hydraulic fill dam'which had a partial failure during

construction. Its finished height is two hundred forty two
feet, the length of the main dam is ten thousand six hund-
red feet, and the volume is 124, 000, 000 cubic yards.

"On 22 September 1958, when the hydraulic fill
was w'ithin t'enty feet of the top of the dam, a slide
occured in the upstream portion of the dam near the
right abutment.

The total length of time consumed in the slide
was about ten minutes. Ap1roximately 5 ,000 ,000 cubic
yards, or less than 53 of the material in the dam
moved. There were a‘oout 180 men in the slide area at
the time it started and eight of them lost their
lives.

Then portions of the u1stream shell neares
the pool be~an to slide into the sinking core pool.
ln a lesser degree, similar cracks and sliding and
slumping were taking place on the upstream portion
of the dowml tream bee ch. Simultaneously with these
developments, the main mass of the upstream shell,
almost inte .ct, wa1s moving out into tlle reservoir, in
a swing simgler to that of a gate hir~ed at tz1e rivht
abutment.

Immediate investigations were ordered on the matter

of the slide and also in view of continuing with the con-
struct1on of the dam. Many drill holes were taken and some
test holes were dug.

Kany opinions have been advanced as to the cause of
this slide. The investigating board decided th1t a formation
of shale in the foundation was the 1u1e€1.te cause. This
shale had a low shear resis ance, the effect of which has

been discus ed previous_y.

 

15. Engineer Corps, U.S. Army,Report_of the slide of a
Portion of the Upstream Face of Fort Tech Dam, July 3§,p. 4

58

Once the shell begins to fail, the liquified portion
of he core would exert a large hydrostatic pressure at the
point of failure and would tend to increase the failure.
This example illustrates the ever present danger during the
construction of a hydraulic fill dam.

The presence of this shale had been discovered during
the site investigations. The design of the dam.was su1posed
to be such that this weakness had been covered. Curiously
enough, one of the recommendations of the board was that the
upstream slepe be flattened to a 1:8 slope or flatter.
Thirty foot berms were constructed in some places. The
original plans called for an upstream slope twice as steep
as the downstream slope.

One possible conclusion is that the low shea did
not show up in the laboratory tests and that a large enough

factor of safety was not provided.

59

TIES 1111.12.11 ER DAM

The failure of the Alexander Dam on the Island of
Kauai in the Hawaiian group, 26 harch 1950, is an excellent
example in the hydraulic fill group. Extensive tests of the
fill material were made before construction started. in
overabundance of fines was noticed and steps were taken to
reduce the danger from it. They included the addition of
chemicals at certain stages of construction to aid in the
settlement of the fines.

"After the accident it was learned that men at
work on the dam had noticed a cessation of drainage
through the bank in the vicinity of the initial
break at least ten days before failure. Up to this
time, the drainage had appeared free and well
distributed."14
The failure consisted of a break in the downstream

shell, through which practically all the core pool moved
downstream, washing away much of the shell.

The core of a hydraulic fill dam consolidates as the
particles settle out and the water either drains out through
the banks or is pumped out. in this case drainage in a
certain area was retarded for some unknown reason. A liquid
pocket was built up within the shell, and, when sufficient

ressure was developed broke out. The core pool moved into
n. ’ J.

this void and on out carrying the downstream banks with it}5

 

14. Joel B. Cox, "Hydraulic-Fill Dam.of Fine Volcanic Ash
Fails Disastrously," Engineering news-Record, 22 May 1930,
p. 871.

15. Ibid., p. 869-71.

60

Two points should be stressed here. The workers had
not been well enough informed as to cause and effect. A good
information program to acquaint the workers with all phases
of the project will pay dividends. Secondly, the inSpections
of the completed portions of the dam were not complete and
thorough.

This dam was completed a few months later success-
fully. AS the dam was built up, drainage systems were
installed in the shell, to forestall the forming of liquid
pockets.

An interesting note is that rolled fill dams
generally fail after construction, whereas a hydraulic fill
dam usually fails during construction. The latter is due to
the high liquid pressure within the two constricting dikes

forming the outer portions of the dam.

61

The Clendening Dam

The Clendening earth dam on the Luskingum.River in
Ohio was approximately completed in 1937 when a partial
failure of the dam took place. The visual results of the
failure consisted of a sagging of the top at one point of
about five feet and a bulging of the upstream face at the
same point of about five feet. The toe line of both the
upstream and downstream faces showed no signs of movement.

An investigation showed the following facts. The
foundation had a tOp layer of forty feet consisting of a
fat clay and silt. The bedrock was one hundred feet below
the surface. The borrow pit material was found to be the
same material as the top layer of the foundation. The
moisture content was above average in both the pits and the
dam. The dry density of the dam.was found to be ninety six
pounds per cubic foot, a low value.

The semi-impervious shell on the outside of the core
was a random fill, a large part of which was shale. This was
supposed to be broken down by the rollers to form a dense
mass. This was not done, and the saturated shale became
very slippery and formed slickenslids, a name for slippage
planes. This unstable mass was no help in containing the
core.

The fact that the moisture was too high was realized
at the time of construction, but no other material was

available except at exorbitant hauling distances. nothing

62

was done to change the design of the dam because of this
undesirable feature.

As a result of the investigations, as additional rock
fill was added to the upstream face extending from within
seventeen feet of the top down past the toe for a consider-
able distance. The new rock fill was about fifteen feet
thick horizontally. This additional weight was sufficient to

stop the movement and to stabilize the upstream part of the

An interesting result of the analysis of the test
holes which were dug was that the fat clay in the foundation
of the dam had very little effect on the failure. it was
found that there was just the eXpected amount of settlement
in the foundation. This is a proof of the statement made
earlier that earth dams can be built on other than rock

formation successfully.

16. Creager, Justin,and Hinds, Engineering_Eor Dams,
p. 772-5.

63

The Horse Creek Dam

The Horse Creek Dam near Denver, Colorado is an
example of the construction methods sometimes used in the
early part of this century and of how not to build an earth
dam. The one correct thing done, according to present day
standards, was that substantial concrete collars were built
in both the upstream and downstream portions of the dam
around an outlet pipe laid through the bottom of the dam.

A description of the construction follows:

" The material was deposited in layers as thick
as four feet, ....... . As the layers were three to
four feet thick, the Fresnos could hardly have been
used very much for pulling down and leveling up the
earth as it was dumped from.the wagons.

The bottom two feet of the dam are said to
have been placed moist. The remainder of the material
was very dry, and there was no effort made to moisten
it to aid in compacting. Aside from the packing
effect of teams and wagons, no effort at comwaction
was made. ...the structure settled seven feet the
first season and 34,000 cubic yards of additional
material had to be placed on the top to bring it to
grade again.

The dar was not provided with a puddled core
or core wall of any kind, dependence apuarently being
placed on the concrete lining on the upstream face."
Preliminary investigations showed that the bedrock

came very close to the surface at intervals. it was proved
later that the initial break occured at a point where the
bedrock was within four feet of the surface. it was decided
that this shallow rock forced the line of saturation up

above the toe of the dam.

4.-

17. Joel D. Justin, Earth Dam Projects, p. 27-9.

 

64

This meant that the exposed earth was saturated, pining

.'

started, and as can be seen from the above description,

there was nothing to retard the buildup of the flow through
the dam and failure results—(2.

Adequate compaction and a core wall might have
combined to bring the line of saturation down to a safe
point. There are innumerable ways to shape the core; and all
types of cut-off wa-ls, diaphragms, and innvrvious blankets
are used to reduce per elation to the point where no harm
can be done. The impervious blankets are constructed in the

same manner as the dense core except they are usually not

over six feet thick.

65

Delhi Dam

The failure of a small rolled fill dam at Delhi,
Lichigan, 8 April 1926, illust ates the results of poor com-
paction, insufficient slope to the upstream face, and insuf-
ficient spillway capacity. The dam was just completed prior
to the coming of winter. Flood conditions from a thaw the
following Spring brought the water to within one foot of the
top of the dam.

The only compaction effort had been the passage of
tractors over the fill. This resulted in a very incomplete
compaction. The rising water in the reservoir soaked the
outer surface of the dam and froze. This formed a crust. As
the fill within consolidated, it moved away from the crust
creating voids. The unanticipated height of water was suf-
ficient to break the crust at Spots. The water rushed in and
soon made a path through the entire dam, carrying large
portions of the fill with it.

It was later recoru.nded that the suillway be incre-
ased, and the upstream face protected by selected stone to

take care of the 1:2 slope.18

 

18. Engineering hews-Record, V. 96, p. 942.

CIL'QTLIR VII

S'LEIM‘LRY

67

The requirements of an earth dam are thorough
investigation, safe design, careful construction, and
vigilant maintenance. These items have been illustrated in
the preceeding pages.

There was not enough time available to examine each
section of this study and present the details as regards to
soil. As stated in the introduction, the use of soil in
earth dams was to be scrutinized in all its aSpects, which
has been done in a general nature.

The study of soil seems to lead one into blind paths,
yet when the soil is used carefully and intelligently,
structures can be built which will be lasting. The constantly
changing characteristics of soil must never be forgotten as
one moves from one spot to another.

Further study is needed in all phases of earth dams.
The laboratory test would be an esoecially fruitful subject.
yarticularly the determination of the optimum.moisture
content. This test could be set up by using the same sample
and the same range of moisture contents. The variable would
be the number of blows of the tamper. The object would be
to determine the effect of increase or decrease of the com-
paction effort.

A further study could be made to correlate the number
of blows of the tamper to the pressure obtainable from the

roller.

68

The earth sample could be varied with the other
factors kept constant. An indication would then be found in
regards the effect of the same compaction on different types
of soil.

The construction process of both the rolled fill and
the hydraulic fill can be analyzed more thoroughly in an
effort to uncover any inefficiencies or new methods not yet

used.

B IB LI OGR’EPhY

70

13 n3 LI OGZL"; m
BOOKS

Encyclopedia Americana, 30 vols., Americana
Corporation, 1947, hew'York.

 

Bassell, Burr, Earth Dams, New York, The Engineering
News Publishing Company, 68pp.

Taylor, Donald W., Fundamentals of Soil Fechanics,
JeW'York, John Wiley and Sons, Inc., 1948, 700fipp.

nousel, William 8., Applied Soil mechanics, University
of michigan, 1948, 189 pp.

 

Plummer, Fred L. and Dore, Stanley M., Soil Mechanics
and Foundations, New York, Pitman PuinShing Corp.,
1940, 473 pp.

 

 

Krynine, Dimitri P., Soil Mechanics, New York,
MoGraw-nill Book Company, Inc.,gI941, 451 pp.

 

National Resources Committee., Low Dams,
Washington, D.C., U.S. Government Printing Office,
1939, 431 pp.

Creager, Justin, and Hinds, Engineering for Dams,
New fork, John Wiley and Sons, Inc., 3 vols.

Justin, Joel D., Earth Dam Projects, NeW'York,
John Wiley and Sons, 345 pp.

nanna and Kennedy, Design of Dams, New York,
McGrawaHill Book Company, Inc., 478 pp.

Terzaghi, karl, Theoretical Soil Mechanics, NGW'YOTK,
John diley and Sons, Inc., '

71

B. PER ODICAL ARTICLES
Walter, D.S., "Reclamation's Biggest Earthfill Dam "
Engineering News-Record, CXIL (18 September 19475,
396-402.

Engineering_Hews-Record, IVC (10 June 1926), 942.

 

Cox, Joel B., "Fill Dam of Fine Volcanic Ash Fails
Disastrously " Engineering Mews-Record, CIV
(21 May 1930 869-71.

Leonard and mattern, "Planning the Watauga Project,"
Civil Engineering, XVIII (March 1948)

 

Leonard and Mattern, "Constructing the Natauga
Project," Civil Engineering, XVIII ( May 1948) 21-6.

 

Rowan and Graham, "PrOper Compaction Eliminates Curing
Period in Cinstructing Fills," Civil Engineering,
XVIII (July 1948) 50-1.

 

Armstrong, R.W., "Delaware Aqueduct to Increase New
York's Water Supply Fifty Percent," Civil Engineering
IXX (February 1949) 17-21.

"Davis Dam Completes Storage Regulation of Colorado
River BeIOW'Boulder," Civil Engineering, XVII
( January 1947) ‘

 

"Garrison Dam on the Missouri River to Hake Rolled Fill
history," Engineering News-Record, CXKXVI 920-2.

 

kramer, Hans, "Deep Cutoff Trench of Puddled Clay for
Earth Dam and Levee Protection," Engineering
News-Record, CXXXVI 986-90

 

Thompson, J.N., "Soil Studies for San Jacinto Dam,"
Civil Engineering, XVI 57-9

 

C. UNPUBLISHLD F TLRIAL

Corps of Engineers, U.S. Army, "Report of the Slide of
a Portion of the Upstream.Face of Fort Peck Dam,"
July 1939, 4.

N01 ’54 ROOM USE 0NW

NO 15 "3‘
No 29 '5‘

De i0 ‘3“

 

 

"' “TI?“ ‘33}?7‘

'“T‘RT

- w '{l—I‘V‘Iv—u .1..."

»' wwv rum-

v w‘mvwn

u-mp.

'~~)‘1Ilv' r‘fl'vu .. t‘unpngrv‘ h,

MILH
1:3Ii311)‘l)9)3)

1‘
H“

330

- - 4: "151'

H1
\

NIH““HMHH‘HMH‘I

=1 Ann-y ., pup». .