I
‘l
I

,l ‘
i I
’I
i i
‘ I
I
I
I II

 

—I_p_\
Irv-b I
CD\I4>- x"

GRADATION AND CONSOLIDATION
OF POROUS BACKFILL MATERIALS

Thesis for tho but» of
Civil Engineor
MICHIGAN STATE COLLEGE
Gail Charles Blomquisf
I948

 

 

 

 

 

 

I
This is to certify that the .- v
thesis entitled .
GRADAI'IUN AND CuI‘iSJLIuAI‘lQN OF PURUUS
BAChFILL ulA‘l‘LZ-CLALS
presented by

I h I
GAIL CHAhLbS BLCMQUIST

has been accepted towards fulfillment })&l‘"“¥t l

of the requirements for

I (I: {1“ A's" I t . l

3 o c o a g l y' \ .3‘ . L. , ‘.

Pr chonaldegree in C1v11 Englneerlng rain“; , (.5; r}: H; .
i»‘\ I all.“ Ix“ l.‘\."

 

i ‘-« .
Major professor ll ; I" 'w.‘

‘11:“. ."'
A: I .' I .3. I K;
\ ‘_ . \A .

I\ ’I y
I: {.nyl:. fél,‘ \N‘.‘ " 9, . "
rifle}? L‘ . ‘. 7- ‘l
n I t e. .. - 1 a
n ‘ .. - I. l'm . ‘

. mi;

N

DateW_

IIIIII

 

 

InfiA‘

 

 

GRADATION AND CONSOLIDM‘ION or POROUS

BnCKFILL MATERIALS

By
GAIL CHARIES BEQIQEUI see

A THESIS
Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements

for the degree of

CIVIL ENGINEER

Department of Civil Engineering

1948

TH 5515

 

FOREWORD

One of the most important problems today for any structural or high-
way engineer is that of settlement in foundations or of substructural
materials. While this may be a relatively simple problem as compared to
other prbblems of soil mechanics it is a very important one to the high—
way engineer. Generally, even the most elaborate subsoil investigations
fail to give a complete picture of all conditions, hence the application
of any such investigations must be tempered with judgment and experience.

The deformation of soils is much greater than that of steel or con—
crete. The usual construction materials are considered rigid and to
follow the principles of elasticity within their elastic limits. Soils,
on the other hand, must be considered as plastic solids under some cir—
cumstances and as viscous liquids under others. The deformation of many
types of soils is a progressive action and will not reach its maximum
for some period of time. The deformation of a soil mass results from a
change in both shape and volume as compared to other structural mate-
rials which only change in shape. In the study of settlements the vol—
ume change of the subsoil layers is more important than deformation due
to change in shape.

Settlements of structural approaches, fill settlements and failure
of pavement surfaces have brought the point to light that the selection
of materials which will eliminate excessive volume changes after place-
ment is very important to the highway engineer and also to the field of
highway construction. The use of better materials and closer control
has decreased the settlements on these structures considerably, but it

has not proven to be the panacea for the entire problem since settlements

3318710

 

 

are still apparent in structural approaches and irregularities are found

in the surfaces of our modern highways.

With an aim of finding methods for further limiting the above men-
tioned objectionable settlements and subsequent failures of structures,

a study has been carried out on the gradation and placement of embankment

and structural backfill materials. This report is a presentation of the

results and tests made and also the results of studies on gradation and

consolidation of granular backfill materials.

GRADATION STUDIES

One of the functions of a granular backfill is that of promoting
drainage away from the structure of any water seeping into the backfill
material from adjacent soils or any water that might gain entrance from
the surface. The gradation of a good backfill material must be such
that it will allow water to pass through it without being trapped or
ponded on any part of the subgrade or on the structure proper. However,
any water passing into and through a backfill material in all probabil-
ity will carry with it the finer particles of the adjacent soils from
which it is entering, or it also may carry away the fines of the back-
fill material as it drains. In either event it will lead to subsequent
settlement, by the formation of voids, in either the backfill material
or in the adjacent soil allowing the backfill material to change shape
or be decreased in volume. If the fines from the adjacent soil are car—
ried into the backfill material it will become choked and pending will
occur on the surface or in adjacent soil masses making them viscous,
and deformation will occur causing settlements in the backfill material.

A backfill material should be of such texture that it can be placed
during any weather condition and will not require special handling; for
example, drying before being placed. The ideal material would meet all
of these requirements and also require a minimum of effort to consoli-
date it to such density that no further detrimental settlement would
occur.

It has been definitely established that any type of backfill mate-
rial when placed loose will undergo rearrangement of particles due to

the vibrations set up by traffic. The amount of settlements will vary

-5-

with.types of backfill material and the size and proportions of the
backfill in place. From the foregoing facts it is evident that the ma-
terial should be well consolidated before placing the road surface over
it.

The United States Corps of Engineers made a comprehensive study of
the above requirements of soils in the Waterway Experiment Station at

Vicksburg, Mississippi. From their studies the following formulae were

developed:

1. To insure the prevention of detrimental movement of adjacent
protected soil (subgrade soil or filter layer) -

l§% Size of filter_material Not reater than 5
85% Size of adjacent protected soil g “ ‘
2. To prevent movement -

85% Size of filter material
Size of perforation or slot opening

- Greater than 2

5. To insure satisfactory permeability -
l§% Size filter material _ Greater than 5
15% Size of adjacent protected soil

The above formulae to be applied following laboratory investiga-
tions, to each type backfill material, each change in subgrade soil, and
each type of outlet having varying size openings.

The scope of investigations covered herein includes a study made
with the view in mind of establishing a gradation range for all granular
backfills. The gradation range established included the majority of the
natural gravels and sand deposits of Michigan. (These studies were car-

ried on in the Michigan State Highway Laboratory.)

The studies included standard Mechanical Analysis tests and filter
tests on varying kinds of natural materials. Tabulated records of these
studies can be found in Appendix A.

The filter tests were carried out in a glass faced rectangular tank
50 inches in length, 15 inches high, and 4 inches wide. Imbedded in the
material and an integral part of the tank were two cylindrical metal
screens 1-1/4 inches in diameter having 1/52 square inch openings spaced
1/32 inch apart. These two screens were placed six inches apart and
5—5/4 inches above the bottom of the tank, being symmetrically located
on the face of the tank. These screens provided an outlet, or overflow,
for the water filtering through the material in the tank after water was
introduced from the top of the tank.

The first filter test was run with a coarse aggregate having the

following gradation:

Size of Per Cent
Sieve _§a§ggyg_
l" 70
1/2" 45
# 4 20
# 8 14
# 10 10
# 20 7
# 4O 5
#50 4
#100 2
#200 O

-5-

The material was thoroughly mixed and was then placed in the tank.
Water was introduced into the tank through a sloping trough that had
openings provided in its bottom to allow the water to fall on the sur-
face of the material uniformly.

The water filtered down through the material and seeped out through
the two screen openings into overflow cans where all water and sediment
were collected. Following a twenty-four hour period, during which the
water was being introduced continually, the sediment in the overflow
cans was collected and dried. This material totaled 8 grams, dry weight.

Figure l is a view of the face of the tank following the twenty-four

hour test.

 

 

Figure No. 1. Coarse aggregate filter test following twenty-four hour
test.

-6-

Following the twenty-four hour test, 50 pounds of sand was spread
in a uniform layer over the top surface of the coarse aggregate. The

sand used had the following gradation:

Size of ‘ Per Cent
Sieve ' Passing
# 40 100
# 60 95
#140 18
#200 9
#270 5

Before water was introduced on the surface the screen Openings were
closed by means of valves provided to maintain a head on the filter tank.
Water was introduced as before and allowed to fill the tank to about 2
inches covering the sand layer. When this level was reached the valves
on the screen openings were gradually opened, allowing water to drain
off. At the same time, additional water was introduced to maintain a
1-1/2 inch head on the sand layer. As before the sediment was collected
from the overflow. The valves were Opened gradually and were completely
open at the end of a three-hour period. The amount of water added to
maintain the head decreased gradually for several hours. After running
the test for five hours no movement of material was evident within the
tank. Following the test the sediment in the overflow cans was dried and
weighed. The dry weight of the sediment was 52 grams.

Figure 2 shows the material in the tank at the conclusion of the

test.

 

'\.*""
;\
6‘.

. 0’ -
p a...

 

Figure No. 2. Coarse aggregate filter test after sand was added to the
surface and water allowed to pass through.

The gradation of the material in the tank following the two fore-

going tests was as follows:

Size of Per Cent
Sieve Passing
l" 74

1/2" 50
# 4 26
# 8 16
# 10 15
# 20 13
# 40 12
# 50 10
#100 2
#200 O

The second filter test was carried out with a finer material as a

medium. The gradation was as follows:

Size of Per Cent
._§ieza_ 2222iaa.
# 4 - 70
# 8 57
# 10 50
# 20 40
# 40 50
# 50 25
#100 20
#200 5

The material was thoroughly mixed and placed in the tank. Water
was introduced and allowed to rise to a head of four inches over the

material. This is shown in F1: e 5.

   
 

     

1r

.. . :
V«‘ r" ‘.

' \
{0‘ .(0 _.~‘ ‘ " s_ . ’ - 1.;

Figure No. 5. View showing material at start of second filter test.

 

-9-

The valves were opened as before and water introduced to maintain a
constant head or as near as possible. The water drained off slowly and
diminished its rate gradually. This operation was continued for
eighteen hours. During the test period water was filtering through at
the rate of 2.08 gallons per hour. Following the eighteen-hour test the
tank was allowed to drain and the sediment in the overflow cans was

dried and weighed. During the entire test only 1.8 grams of material

was filtered out. Figure 4 shows the tank following the test.

I‘v-Ji‘-‘ .. , , .’
V , f‘:‘~):‘."r _- \’, '1‘! "I' t,'- {i
. «; ~13 {.1
"$3- flafi$4#~ifif; Jr;
V “ x I?" {I w
. ' ,. .--.- .
‘:,: 6. ’.‘.:( fl, .‘.1 .#. .‘I .. .
. v ‘ ‘ _ ' f .

;._ ,.
31.9.,

 

Figure No. 4 Fine gravel and sand following eighteen-hour ponding test.

-10-

The ponding was continued for an additional 2 hours at which time a
uniform layer of 50 pounds of silty loam material was spread in the
water on top of the fine gravel and sand filter. The gradation of the
silty loam was:

Size of Per Cent

Sieve Passing

# 40 100

# 60 99

#140 94

#200 79

#270 65
* .005 mm

* Mechanical Analysis
With the addition of the silty 10am the amount of water passing was
reduced to 1.52 gallon per hour. The test was continued to 56 hours at
which time only 0.88 gallon per hour was passing through with the same
head as at the start of the test. During the last few hours of the test
there was no apparent decrease in the flow below 0.88 gallon per hour.
The sediment cans contained a total of 0.5 grams of material following
the addition of the silty loam filler. Figure 5 is a view of the tank
at the completion of the test.
A six-inch layer of material was removed from immediately over the

screens and a gradation study made on it as follows:

Size of Per Cent
Sieve Passing
# 4 69.1
# 8 55.4
# 10 50.7

-11..

(continued) Size of Per Cent

Sieve Passing
# 20 40.1
# 40 24.4
# 50 ' 19.0
#100 4.6
#200 1.1

The above gradation when compared to the original gradation will

show very little of the silty loam soil was carried into the material,

or through it, from the sediment weight following the test.

  
    

   

 
  

     
 

     
 

“f4 _ . 1 '
' ,3 .. 3 ‘ -
‘. Q45" . ’ ‘ .' , \' . :_
'..~.-’.. 0 . 8‘ 7 : ‘
. > i . ::)f_- 5.“ a ‘ '. s ' . \-
. -.-‘-,‘. w-w _ .‘ ff]; -.';
. . -' rag -\.~~._‘;:§~ ,' .
u - . . > h I - I‘ ~ '0 . ~ 4 ‘. ~ . O
.L~ 9"“ It? \0 £7 \' 1,
' 6‘ ’3‘ 1““. d _ J”? _. -'
~ g‘r. -. - 2-.- w ' .3
.° t )0. .d -
’5'}
«,1
.. q. - L
- ' 'Q
. ( ‘
rg- . \-v ~
I ~' ‘._

 

Figure No. 5. Fine gravel and sand material after 56-hour ponding.

The standard Highway 62A Specification gravel was used in the third

filter test. The gradation of the 6—A gravel is as follows:

Size of Per Cent
Sieve Passing
1—1/2" ‘ 100
l" 79
1/2" 55
# 4 0.5

Following a thorough.mixing, the gravel was placed in the tank and
water introduced as before. As compared to the foregoing tests, the
only limit to the amount of water flowing through the material was the

size of the outlet pipes. Figure 6 shows a view of the tank prior to

the test.

 

 

 

Figure No. 6. Standard 6-A gravel in tank before test.

-13-

After testing on the Sea gravel for a short time without any evi-
dent change in its structure or any sediment carried out by the water,
a layer of the 6-A gravel was removed and replaced by 27—1/2 pounds of

sand spread in a layer of uniform thickness. The gradation of the added

sand was:
Size of Per Cent
Sieve Passing
# 40 100
# 60 95
#140 18
#270 5

With only one of the outlets open to flow the test was continued
as before. There was no apparent reduction in the rate of flow as com-
pared to the Sea gravel test. A view of the tank under these condi-
tions can be noted in Figure 7. The sand was observed to move through
the 6-A gravel with little difficulty. Piping can be noted in the upper
left corner of the tank.

When the second valve was Opened the piping effect was accelerated
leaving a very distorted surface on the sand layer, as shown in Figure
8.

With the rapid flow of water through the filter medium the sediment
carried through was not recovered, consequently no data on per cent
passing through the tank was available. The rate of flow through the
tank was approximately four gallons per minute.

' A mixture of 6—A gravel and 2N8 sand was used in test NO. 4. The

proportions were 70 per cent and 50 per cent reSpectively.

-14..

 

U?)

Figure No. 7. 6-A aggregate test showing the movement of sand through
the backfill material.

The gradation of the 6-A gravel was as follows:

Size of Per Cent

Sieve Passing
1-1/2" 100
1" 75
1/2" 40
# 4 O

The 2NS sand had the following gradation:

-15-

1335‘

y’mym .‘.7."

.4 w

 

W") 6‘?)

Figure No. 8. 6—A aggregate test showing distortion of sand surface.

Size of Per Cent
Sieve Passing
# 4 99.7
# 8 86.7
# 16 61.9
# 50 58.1
# 50 9.6
#100 1.2
Loss on Washing 0.6

The two materials were mixed thoroughly and placed in the tank. As

-15-

noted in Figure 9, segregation occurred during the placing of the mate-
rial despite all precautions and care taken. Additional segregation and

consolidation of the entire filter was evident with the addition of

water.

 

 

Figure No. 9. 6-A aggregate and sand filter before test.

The water was introduced into the tank until a head of several
inches was established over the surface of the filter material. While
holding this head level, a uniform layer of 26 pounds of sand was spread
over the filter surface. The head was then maintained at one inch over
the surface of the sand. The gradation of the sand layer added was as

follows:

-17-

Size of Per Cent

Sieve Passing
# 40 100
# 60 95
#140 ‘ 18
#270 5

Following the addition of the sand layer, water was flowing through
the outlets at the rate Of 28.0 gallons per hour. Seven hours Of con-
tinuous running under the one-inch head of water, the rate of flow had
decreased to 15.5 gallons per hour.

The surface was distorted by piping action and the sand had pene-
trated to a depth of several inches in many portions of the filter
material.

The test was shut down and allowed to drain for several hours.
Again the water was started with a resulting discharge Of 14 gallons per
hour. Five hours following the second portion of the test, the flow
through the discharge pipes had dropped to eight gallons per hour.
Figure 10 shows a view of the test following forty-eight hours of
testing.

A layer of the filter material immediately above the screens was
removed and tests carried out on it. The resulting gradation of the

material was as follows:

Size of Per Cent
Sieve nggigg_
1—1/2" 100
1" 85.8
1/2" 56.6
# 4 50.9

-18..

(continued) Size of Per Cent

Sieve Passing
# 10 25.4
# 20 17.0
# 40 - 8.8
# 60 5.2
#100 0.5
#200 0.2

 

Figure No. 10. Note penetration of fines into the filter material after
48-hour test.

Following the 48-hour test the sediment collected was dried and

weighed. The amount was found to be only 0.5 grams. The small amount

-19..

Of sediment collected following the extensive test indicated the mate-

rial at the screens was functioning as a very efficient filter.

The last test was carried out on a commercial filter material com—

posed of standard 6-A slag with the following gradation:

Size of Per Cent
Sieve Passing
2" 100
1-1/2" 98
l" 75
1/2" 51
# 4 1.8
Loss on Washing 0.8

The slag was placed in the tank with all precautions to prevent

any segregation before the introduction of the water. Water was intro—

duced and allowed to accumulate to a head Of three inches over the slag

filter. While maintaining this head a layer of sand of the following

gradation was spread over the slag:

Size of Per Cent
Sieve Passing
# 10 100

# 20 99.7

# 40 99.2

# 60 79.9

#140 5.6

Loss on Washing 5.0

As in the test of the filter of standard 6—A gravel, only one of
the valves was Opened at the start of the test with the result that the

water moved through the filter very rapidly and carried with it much of

-20—

the sand, depositing it in the slag. The water also carried a consider-

able amount of the sand out through the outlets into the sedimentation

cans.
The valve was then closed and the second valve opened. The water

drained out very slowly following the opening of the second valve indi-
cating a choking up of the drain by the sand carried down during the
first part of the test. The first valve was again opened with a

resulting discharge as great as before which would further indicate the

impervious layer over the second drain. A view of the slag filter at

the close of the described test is shown in Figure 11. The distortion

of the surface of the sand layer can also be seen in Figure 11.

 

 

A 6-A slag filter showing infiltration of sand filter.

Figure No. 11.

-21-

After a careful analysis and study of the foregoing tests and their
results, a specification for granular backfills was decided upon. The
specification, which is quoted in the following paragraph, is believed
to cover most requirements of a granular backfill and the materials are
readily available and obtainable in practically all of the sand and
gravel deposits of Michigan.

The specification of the over—all requirements of a granular back—

fill is as follows:

Size of Per Cent
Sieve Passing
2" 100
l" 70 - 100
1/2" 45 - 100
# 4 20 - 70
#10 10 - 50
#40 5 - 50
Loss on Washing 0 - 5

The problem of a backfill material to meet the requirement where
the adjacent soil is largely silt or very fine sand (.005 to 0.105 mm)
could not be covered in an over-all specification that would function
favorably on most soil types encountered. Therefore, the following

specification is recommended for use in backfills of this class:

Size of Per Cent
Sieve Passing
l" 100
1/2" 75 - 100
# 4 50 — 70
#10 55 — 50

-22-

(continued) Size of Per Cent

Sieve Passing
#40 20 - 50
Loss on Washing 0 ~ 5

Backfill materials recommended in the first specification can be
found in natural deposits in Michigan and are available to the majority
of construction projects.

The material recommended in the second specification may be found
in natural deposits in some localities and in others it would require
processing to proper gradation. This material is such that there will

be no infiltration of the adjacent material.

The tests described bring out concisively the need for proper gra-
dation for granular backfill materials. Of equal importance with the

gradation is the method of placing these materials.
CONSOLIDATION STUDY

Following the tests to determine the specification for the grada-
tion of granular backfill materials the study was centered on methods of
placing materials in backfills and their manipulation to achieve the
desired densities. This portion of the study was carried out on scale
models in the laboratory and later moved into the field on a larger out-
door model. On the outdoor model heavy construction equipment was
employed while in the laboratory only simulated and comparative tests
were conducted. The next and final step was to move the study into the
field on actual construction projects in various portions of Michigan.

A controlled density study of backfills under construction was the pur4

pose of this phase of the study. As part of this final study, a density

-23-

investigation was carried out on backfills in place that varied in age
from one to several years.

The first test on methods of placing and consolidation was carried
out on dune sand and pea gravel. These two materials were chosen for
their uniformity of gradation. Methods of consolidation were by ponding,
tamping and vibration. The tabulated results of these tests can be
found in Appendix A. A brief discussion of the tests will be covered in
the following paragraphs.

Approximately 95 tests were conducted on the dune sand. Of this
number 50 were on the material placed in its loosest possible state; 21
tests were carried out to study consolidation by flooding or ponding;

9 tests were carried out employing vibration as a means of consolidation,
and 16 tests with tamping.

The tests carried out on the loose material were to study the
effects of bulking, voids, and to determine a loose volume density. All
of these were determined for varying moisture contents which were used
for comparison in later tests.

The density of dune sand with less than one per cent moisture was
found to be 95.5 pounds per cubic foot. The average density of ten
samples with moisture contents varying from 1.15 per cent to 17.6 per
cent was found to be 75.65 pounds per cubic foot. The average moisture
content of all the tests in the loose state was 8.21 per cent with a
moisture content of 1.15 per cent yielding the high density of 78.0
pounds per cubic foot and a low density of 75.54 pounds per cubic foot
with a moisture content of 15.6 per cent.

The average per cent voids for the ten samples, with varying mois-

ture contents, was 54.4

-24-

From the above tests it would be a good assumption that any loose
deposit of dune sand, or a sand of comparable gradation and Specific
gravity, would approximate a density of 76 pounds per cubic foot with a
total per cent voids of about 54.

Following the determination of the above standards for dune sand,
tests were carried out to eXperiment with varied methods Of consolida-
tion with the view of reducing the voids ratio to a minimum with the
minimum effort and high efficienqy. The moisture content was varied in
each test to also study its relationship to consolidation and efficienqy.

Inundation Tests: The first method of consolidation tried was
inundation with water. Before the tests were carried out on the material
in this series it was found to have moisture contents below one per cent
and a loose density of less than 95 pounds per cubic foot. The voids
totaled 45.4 per cent for all of the material prior to flooding. The
average density of all tests following the inundation period was 97
pounds per cubic foot. The moisture content immediately following the
draining off of excess water from the material was found to be 21.5 per
cent. A total Of five tests were carried out to arrive at the above aver-
age data.

Similar tests were carried out on six samples of dune sand with
moisture contents ranging from one to two per cent. Average density and
per cent voids prior to flooding were 76 pounds per cubic foot and 54.2
per cent reapectively. The average moisture content was 1.57 per cent
for all six trials. Following the flooding the density averaged 88
pounds per cubic foot and an average per cent voids 0f 47. The moisture

content taken, as before, was found to be 25.2 per cent.

-25..

The above tests tend to indicate that the per cent of moisture in
the dune sand prior to test has a marked effect on its consolidation
range by the inundation method. The second series of tests indicate a
final density which was lower than the loose density when the moisture
content was less than one per cent.

The next series of tests, four in number, were also run on dune
sand but varied in the sense that the water was introduced from the bot—
tom of the sample instead of ponding it on the surface. This series of
tests were also carried out on the material in its loosest state with
moisture content varied as before. The first series of four tests had
an average density, prior to inundation, of 95.9 pounds per cubic foot,
voids totaling 45.4 per cent, and an average moisture content of 0.156
per cent. Following the test with indirect inundation the density had
increased to an average of 95.8 pounds per cubic foot, and the voids
decreased to 42.2 per cent with the moisture content averaging 27.5 per
cent before complete draining.

The indirect inundation tests on material having a moisture content
in excess of one per cent gave results of an average density of 89.6
pounds per cubic foot and total voids of 46 per cent. The average den—
sity before the test was found to be 74.7 pounds per cubic foot
resulting from a low density of 70 pounds per cubic foot and a high of
89.1 pounds per cubic foot with moisture contents of 10.6 per cent and
1.29 per cent respectively.

The conclusion from the first group of tests is further borne out
by the results of this group of tests. Low moisture contents before

inundation resulted in greater densities with lower voids.

—26-

Vibration Tests: In the next series of tests the dune sand was
consolidated by means of vibration. The vibration was accomplished on a
vibrometer table which is mounted on springs that can be either employed
or not as desired. The springs tend to dampen out the sudden shock of
vibrating motion when they are employed in the system. The vibration is
obtained by a conventional type vibrating motor attached to the under
side of the table. In these tests the term "mild vibration" will be used
to indicate the springs are in force and "extreme vibration" when they
are not.

Two tests were carried out with dry dune sand. The first of these
tests was under mild vibration for four minutes and it gave a resulting
density of 100.5 pounds per cubic foot, with 59.6 per cent voids. The
second test was under extreme vibration conditions for a period of four
minutes and it resulted in a density of 106.44 pounds per cubic foot
and a per cent voids of 55.9.

Five other tests were run on the same material but with varying
moisture contents. The first three of these tests were with mild vibra—
tion for four minute periods. The average density that resulted was
96.5 pounds per cubic foot and 42 per cent total voids. The average
moisture content for the three samples was 5.56 per cent. The average
density and per cent voids for the two tests with vibration were 104.55
pounds and 57.1 per cent respectively. The average moisture content
was 4.1 per cent.

The above results on vibration tend to indicate a slight lowering
of the attained density by increasing the moisture content, but there

are not sufficient test results to draw any definite conclusions. The

-27-

difference between mild vibration and extreme vibration are evident in
the results, but again not conclusive.

Cone Tests: The cone test is a method of consolidating granular
materials to compare attained densities of varying samples to a standard.
The method compares favorably with the results obtained from the density
tests by extreme vibration. The relative merits of this test over the
vibrometer table test is the simple and rapid means of checking density
requirements in either the field or the laboratory.

Appendix B contains a complete description of the test and equip-
ment along with data on 14 tests carried out on dry dune sand and 15
tests on dune sand with varying amounts of moisture. From the data
sheets it was found that the average density for the 14 dry samples was
105.25 pounds per cubic foot. With the above density the average total
voids was 57.5 per cent.

The average density from the 15 tests on the dune sand with varying
moisture contents was 104.1 pounds per cubic foot, with a total per cent
voids of 57.5. The moisture contents varied from 5.5 to 15.7 per cent.

A.__Ai S. H. 0. Test T-99: As a basis of comparison for the test

 

conducted to this point the standard test for the compaction and density
of soils as designated by the American Association of State Highway
Officials was carried out on 16 of the samples of dune sand with varying
moisture contents. The average density attained was 104.5 pounds per
cubic foot when compacted under this Specification. The total per cent
voids averaged 57.16 with the moisture content averaging 7.41 per cent.

The minimum density was 100.5 pounds per cubic foot with a moisture

.28-

content for 1.24 per cent and a maximum density of 105.9 pounds per
cubic feet at 15.02 per cent moisture.

0n the test results it can be seen that the most unifbrm results
were obtained with vibration, cone and tamping methods. They also gave
appreciably greater densities. From the results of the test methods it
is apparent that the method of vibration for consolidation is the one
most applicable and suitable to general construction practice to achieve
the maximum compactive effect with the least compactive effort. The
A. A. S. H. 0. test method gave comparable results with the vibratory
and cone density tests. Since this test was designed primarily for use
on soils of very fine gradation, it is not generally adaptable to the
deposits of native granular materials, therefore not generally suited
to field construction.
Consolidation of Pea Gravel

A series of tests were carried out on pea gravel having the fol-

lowing gradation:

Size of Per Cent
Sieve Passing
1/2" 100
5/8" 96
# 4 2O
#10 2

The average loose density of the pea gravel was 94.17 pounds per
cubic foot, with a total voids of 41.9 per cent and an average moisture

content of 1.4 per cent. A total of 55 tests were carried out to arrive

at the above data.

-29-

Inundation: The tests by ponding were carried out in the same
manner as those described previously in this report.

The average density following ponding of the pea gravel was 95.5
pounds per cubic foot, the total voids averaged 42.4 per cent. These
averages are not based on extensive data and are not a standards, but
it can be noted the evident decrease in density with saturation and a
resulting increase in total voids. Based on these six tests further
tests were abandoned as it was apparent that any method of consolida-
tion involving saturation or near saturation were ineffective and
impractical.

Cone Test: Pea gravel with the same gradation as that in the pre—
ceding tests was consolidated by the cone method. The average results
were as follows: Density, 107.1 pounds per cubic foot; total voids,
55.96 per cent. This is an increase of 15.7 per cent over the loose
density and a decrease in total voids of 18.9 per cent.

Vibration Tests: Vibration tests were carried out on the pea
gravel which resulted in an average density of 104.28 pounds per cubic
foot and a total voids average of 55.7 per cent. The results of the
vibration tests indicate an increase of 10.7 per cent in density and a
decrease of 14.61 per cent in total voids over the loose density
state.

The tests were run on the pea gravel to bring out the point of
increase in the consolidation of granular materials over their attained
densities when not properly compacted during placing. The resulting
consolidation is caused by the traffic vibrational effect and super-

imposed loads of structures. It should be remembered that in foregoing

-50-

tests of this study it was determined that pea gravel was not suitable
for use as a granular backfill material, since the gradation of the
material was of such nature that material placed over or adjacent to it
infiltrated into the pea gravel medium. This was not desirable since
it gradually shut off drainage through the backfill material.

The foregoing tests have proved conclusively that two factors to
be considered in the discussion of granular backfill materials are gra-
dation and consolidation. First, gradation was found to have great
importance in a majority of backfill projects. There are some cases,
it is true, where gradation would be of no paramount importance. An
example of this would be where granular backfill materials are adjacent
to native soils of clay or clay loam, where there would be little or no
danger of infiltration of water from the adjacent soils. With this
condition prevailing the backfill could be much coarser, and a possible
material would be pea gravel. Care must be exercised with these con-
ditions though to insure that there is a cut-off incorporated to pre-
vent the infiltration of the surface material into the backfill
material.

The second, and possibly the most important, factor is the
resulting consolidation of backfill materials when they are not properLy
compacted during placement. This consolidation has proved to be of
prime importance in bridge approaches with resulting settlements of the
pavement slabs to the point of causing hazardous irregularities in the
surfaces. This condition prevails in a majority of the existing bridge
approaches where no controlled density was employed during the placement

of the backfill material.

VOLUME CHANGE AND CONSOLIDATION METHODS

With the two factors mentioned above in mind, the next step in the
study was to experiment with various means of consolidation to study .
settlement in backfills of various selected granular backfill materials
and also the relative efficiency of each consolidation procedure.

These tests were carried out on small laboratory testing equipment
and were later substantiated by tests on a scale model of a bridge abut-
ment backfill.

The preliminary laboratory tests were carried out with dune sand.
First, the minimum voids were determined for dune sand when it was com-
pacted to its maximum density. This value was found to be approximateLy
thirty per cent. Assuming this value and comparing values obtained by
vibration and tamping on a tenefoot fill, which was assumed to be an
average height fill, when placed loose would have a possible settlement
of twenty-nine inches. This would be a settlement equal to approxi—
mately twenty—five per cent of the original height of the fill. Tests
by ponding on the ten—foot fill of dune sand would result in a possible
settlement of 14.4 inches or approximately twelve per cent. With
tamping and vibration employed as a means of consolidation, a possible
settlement of from O — 7.2 inches could be expected depending on the
effort exerted on these consolidation methods.

Following this series of tests to determine possible settlements
following various consolidation means the operations were transferred
to the bridge backfill model. This model had one side representing the

back of the abutment and the Opposite side was sloped to represent the

-52-

back lepe of the excavation. See Figure 12. The volume of the model
was 15.6 cubic feet and was so constructed to permit removal of the
front face, representing the abutment face, to facilitate inapection of
the material following compaction. The removable front also aided in
unloading and recharging of the model. The test bin was designed so
that a platform scales could be used in conjunction with the test to

a,have an absolute check on the density. Arrangements were such that the
model was jacked up from the scales and lowered to the platform of the
scales following the compacting stage. Care was taken to insure that no
tests were carried out with the scales supporting the model. This
arrangement provided a uniform and accurate method of checking density
on granular materials, which has proven to be a serious problem in all
such materials.

The purpose of the test with the scale model of a representative
backfill was to substantiate results of the foregoing tests and to
determine the following:

1. The compactive effort required to consolidate backfills

of dune sand, natural sand, back run gravels, pea
gravel, coarse aggregate and slag to the point where
there would be no detrimental settlements.

2. The effect of moisture content on compaction of granu-

lar materials by vibration.

5. The relative density requirements for granular mate-

rials when used for structural backfill.

4. The most efficient means of attaining specified densities.

 

 

VxN

 

 

 

 

 

 

m4. .oz 0.3»;

 

4<_mm._.<2

.jExoqm m0 mHmMH 2_
0mm: 1600.2 wqdom

 

 

 

 

To find the answer to these problems a number of tests were carried
out. A standard procedure was followed, an outline of which follows:
1. The material to be investigated was placed in the mold
loose and a weight per cubic foot was determined. Every
precaution was taken to try and simulate field conditions
in the way in which the soil would be placed in struc-
tural excavations. The material was allowed to slide down
the incline and fill up the bin finally being struck off
level with care not to compact any of the top materials.
Figure 15 shows the model filled before consolidation.

2. The material was then compacted with a bullet-nosed vibra—
tor. The vibrator was allowed to penetrate into the mate—
rial and also was operated on the surface in a horizontal
position. The vibrator was allowed to come in contact
with the sides and back-slope of the box at times during
the test, but was not allowed to operate in this manner
for extended periods of time. When the vibrator did come
in contact with the sides of the bin the entire model was i
set into vibration and would undoubtedly have a marked
effect upon the results. The importance of this point is
brought out later on the tests employing the larger out-
door test model. Following the vibrational period, the
volume change was measured as shown in Figure 15. A
number of measurements were taken on each sample from a
level across the top with an average settlement figure

finally determined. A weight was also taken with the

-54...

 

 

Figure No. 15. Model filled with coarse aggregate in loose state.

 

 

Figure No. 14. The bullet-nosed vibrator employed on the model tests.

scales to determine the density of the material following

the consolidation. This density could be calculated by

the known volume and the determined weight.

 

Figure No. 15. Checking settlement after consolidation.

3. The material was then removed from the model by removing
the face from the model. The material was again placed
in the bin with the vibrator operating continually
during the filling Operation. Care was exercised not
to charge the model too rapidly not allowing the vibra-
tor time to effectively consolidate the material

adjacent to it. Density measurements were again taken

-55..

but no settlement data could be taken with this pro-
cedure, or phase of the test.

4. The last step in the model study tests was to place
the material in the model in 4 to‘6 inch layers and
hand tamp each layer with the standard construction
tamper which weighed 25 pounds and had a surface
approximately eight inches square.

A record was kept on moisture contents and relative densities on
all tests. Following each vibration or tamping test the material was
flooded in the model for a period of 20 minutes to determine relative
drainability of each material and test. After a 24-hour drainage
period, moisture samples were taken and also check density measurements.
The materials were then emptied and allowed to air dry before being

used in another test.

Detailed reports on all scale model data can be found in Appendix
C of the report.
The conclusions arrived upon from the series of scale model tests
are as follows:
1. No material can be placed without some means of com—
pactive effort that will not result in detrimental
settlements.
2. Sand and bank run gravels will yield the greater den-
sities with the minimum of compactive effort. The
coarser the individual particles, the greater is the
required compactive effort.
5. Moisture content has no apparent effect on vibratory

consolidation, while with tamping a marked effect was

-56..

5.

noted.

This same observation was noted earlier in

the study for the inundation methods for consolidation.

Tamping can be used

effectively on any

granular material

with resulting maxi—

mum density, but
close control must
be maintained to
keep the layers
being consolidated
to six inches or
less.

Vibration will
yield results com-
parable, and in
some instances
better, than
tamping for con—
solidation of well
graded granular
materials or those
which have uniform
size and are well

rounded particles.

Figure No. 16.

   

Front face removed
showing stability after
compaction.

Vibration is advisable on all such soils since the same

-57..

final results as tamping will be gained with far less
effort and control.

6. When tamping is to be employed as the consolidation means,
the moisture must be maintained at or near the optimum as
determined by the standard A.A.S.H.O. test T-99.

All of the foregoing tests tend to confirm the above conclusions,
but there is too little information available to arrive at any definite
conclusions or recommendations as to the most efficient method of con-
solidating backfill materials on field projects under construction con-
ditions. It has, however, been determined and demonstrated that some
means of consolidation is needed to eliminate settlements in backfill
materials regardless of the type of material being used. The method of
hand tamping or even mechanical tamping would and has proven tedious in
field practice. The use of bullet-nosed vibrators would also prove to
be a slow and tedious means, but would be an advantage over the tamping.
With the results of the previous tests and with the above considerations
in mind, a means was sought to apply the vibrational consolidation to
field construction projects and practices.

The next series of tests carried out were at Scottville, Michigan,
where both field vibrational equipment and facilities were available
for tests of the scale planned. The objectives of these tests were:

1. To study various vibratory apparatus with view toward a means
of consolidating structural backfill materials to avoid
detrimental settlements.

2. To determine effective vibratory range and establish a maxi—
mum lift thickness for placing materials.

3. Study effects of frequency variations on consolidation for

various types of vibrators.

-53-

4. To study lateral pressures as related to consolidation by

vibratory means.

5. To study various types of vibrators with correlation of

effective vibrating time to degree of consolidation.

The granular material on which all of the tests were carried out
was a bank run gravel from near Scottville, Michigan. The physical
properties of this material are included in Appendix D.

The material was confined in an outdoor form 16 feet long, 10 feet
wide and approximately 6 feet deep. The material was employed in the bin
in layers as indicated in the following data. The box was constructed
of 5-inch planking, well braced to avoid any bulging of the side walls

so that approximations could be determined as to volume loss on various

 

 

Figure No. 17. Charging the field bin with mechanical equipment.

-59-

 

tests. The box had a natural ground floor to approximate actual condi-

tions as nearly as possible that would exist on any backfill project.
Three Goldbeck Pressure Cells were installed in the bin to measure
variance in lateral pressures caused by vibratory consolidation methods.

All of the cells were located on one face; one in the lower corner, one

on the lower center face, and the third at the intersection of the

diagonals of the wall.
The following is a detailed description of the tests run and equip—

ment emplqyed on the project. Tables included in Appendix D indicate

densities and moisture contents as recorded on all tests.

The power for the various types of electrical vibrators and mechanical

..

tampers which were used was furnished by a standard mobile power unit

. a. _ .

~ .~-m_fi

Portable power unit for electrical vibrators.

    

 

 

__J_

Figure No. 18.

-40-

shown in Figure 18. The unit is portable, being mounted on a welded

steel wheelbarrow base equipped with a pneumatic wheel. The generator

is rated at 2 K.V.A., 110 volts, 5 phase, 60 cycle AC, 1920 R.P.M. at 60

The generator has a safe Operating frequency range from 50 to
Maxi-

cycles.

80 cycles, with a synchronous motor speed of 5,000 to 4,800 rpm.

mum 5 phase load of 10.5 amps. It is powered by a self-excited and V-

belt driven 5 to 6 HP air-cooled engine.
After the experimental gravel bin had been filled level full with

the drag line, the first experiments tried were with a standard PT—21A
heavy duty paving tube motor to which 6' - 2-1/2" - 1/8" angle irons

were clamped at one end to the motor brackets, the other end being

pointed or spear shaped. Figures 19 and 20.

 

. I
i-'

r*‘ ' '7

 

 

 

 

 

 

 

 

in

 

.~ . x _ “uh-1 .- .5” 1- «a. a" 4w-..w' gig. Wflv”; 5:. ‘°"r y‘fi‘wff, . ,;._ "b 1 '-
Figure 19. Penetration type Figure 20. Spear type vibrator.
vibrators.

-4l-

The paving type vibrator motor is a 110 volt, 5 phase motor which oper-

ates at approximately 5,600 vpm at 60 cycles.

The drag line crane lifted the vibrators of a vertical position over
the bin by means of a loop strapped to the motor bracket. As the machine
was started, the drag line slacked off allowing the spears of angle iron
to penetrate into the material. No trouble was experienced in getting
the spears to penetrate to their full depth as they were lowered. Vibra—
tions could be felt to a slight degree at least three feet from the
°vibrator, but there was no visible settlement of the material surrounding
the vibrator. The vibratory equipment was allowed to run in this
lowered position for two or three minutes and then slowly withdrawn with
no visible settlement of the adjacent materials. Densities which were
taken in the area where the machine Operated showed very slight increase
over the loose density for the material.

The next test was carried out with a flexible shaft concrete vibra-
tor, similar to the type used on the indoor bin tests. Figure 14. This
machine was powered with a 4 HP air-cooled engine. The vibrating head
used was 2-5/8 inches in diameter and 18~5/8 inches long. The head is
capable of operating up to 7,500 me. The vibrator was inserted in the
material at various locations. No trouble was experienced in effecting
the tube to work down in the material. While considerable vibrations
were felt throughout the area adjacent to the machine, very little
settlement of the material was evident. The material displaced by the
head, as it was lowered, was compacted in the area immediately surrounding

the head and left a hole when it was withdrawn. subsequent insertions

~42-

and withdrawals of the vibrator, while varying the frequency of vibration
of the equipment to its maximum and minimum limits, did not alter this
condition. Because of the voids developed with the equipment upon with—
drawal, further tests with it were abandoned.

A large bullet-nosed vibrator was tried in the next test. A View
of it can be seen in Figure 21. This vibrator is a completely enclosed
motor with a double unbalanced vibratory element mounted on the rotor
shaft. The motor is a 110 volt, and 5 phase as before. The full load
speed at 60 cycles ranges from 5.450 to 5.100 me. Horsepower impact
rating for this motor and equipment is 1.5 HP. The vibrator head is
18.5 inches by 6 inches in diameter. This vibrator was not allowed to
work itself to far into the material since too much of the material
would have been displaced which would have stressed the rubber coupling
connecting the vibrating head to the motor to failure upon trying to
withdraw it. The machine was initially designed for work in vibrating
' concrete with a high range of workability.

In the next test the paving tube actuating motor was clamped by
means of the paving tube motor brackets to the ends of two sections of
2-5/8 inch tubing which were 8 feet long. Figure 22. This tubing had
a 5/8 inch wall section and is used for the vibrating members of the
standard paving tube. This equipment was lifted to a vertical position
employing the crane. The tubes were allowed to penetrate to the full
depth of 6 feet. There was no visible settlement of the level of the
material in the bin although vibration could be felt for some distance
from the vibrator. Density tests taken indicated a high degree of com-

paction in the area immediately around the tube positions, but very

.45-

little increase in adjacent areas. As in the case of the flexible shaft
vibrator, voids were left upon withdrawal of the vibrator. Frequency
variations and time increment increases did not change the effects.
Following the tests with the penetration type vibrators, which
proved to be ineffective on the Large volume, further such tests were
abandoned with efforts turned to the experimentation and development of
a surface type vibrator which would impart a force to the layers of fill

material and effectivel consolidate it.

.F-v
‘.'l. I

   
 
   

 

 

 

 

v.“ . ' o
0"“ :‘— .-

’Figure 21. Large bullet—nosed Figure 22. Penetrating vibrator.
vibrator.

The first surface vibrator experimented with was a tamping type.

A View of this is in Figure 25. It is operated by a 110 volt, 5 phase,

60 cycle A.C. motor with a striking member fOrged integrally with rotor
shaft. This member delivers blows to the tappet, which in turn works
on a tamping bar to which a lO-inch conical tamping disc had been
attached. This machine was manually Operated over a small area in one
corner of the bin. There was a marked settlement of the material evi-
dent in the area immediately under where the machine was employed. A
density test indicated good compaction in the immediate area. Fol-
lowing this a platform vibrator, Figure 24, was tried and found to be
very effective. Although manually operated, a large area could be

effectively covered.

 

 

V

Figure No. 25. Surface tamper.

.45—

 

 

 

 

Figure NO. 24. Plathrm vibrator.
The results with the tamper and platform vibrator encouraged the

construction of a larger platform vibrator shown in Figure 25.

 

 

Figure NO. 25. Modified platform type vibrator.

—46—

The standard paving type vibrator motor was mounted on a seven
foot length of twelve—inch channel iron, modified type platform vibra-
tor which was mechanically handled. The motor was bolted to the chan-
nel in a crosswise position at the center of the channel length. After
the bin had been emptied and refilled to a level of two feet, the
machine was put in one end Of the bin and pulled over the surface of
the material by the crane. A cable bail had been attached to the
channel for the purpose of handling it. Skids of 3/16 inch steel, four
inches wide, were attached to the under side of the channel to aid in
keeping it from tipping as it was being pulled. The skids projected
about six inches from the face of the channel and were slightly curved
up at the ends to avoid digging in.

This equipment gave very satisfactory results when operated at fre-
quencies approximating 4,500 me.

The next type vibrator to be tried was made up of two 8—foot
lengths of paving tubes with the vibrator motor centered on the two
limbs. This vibrator was placed in the bin in a horizontal position
with a layer of gravel two feet deep placed over it. A view of this
vibrator is shown in Figure 26. There was an immediate settlement of
about two inches in the material in the area of the bin above the vibra—
tor. Vibration was continued for about three minutes but no further
settlement was evident. The vibrator was then slowly lifted by the
crane and the sling that had been attached to the equipment. After the
vibrator was lifted through the material, vibration continued with the
tubes on the surface gave evidence of little if any additional settle—

ments. The equipment was operated at approximately 4,000 vpm for the

-47-

entire test. Density tests indicated good degrees of consolidation.

 

Figure NO. 26. Paving tube vibrator.

The bin was now emptied and refilled to the top at one end with
the backfill material SIOping to the ground level at the Open end Of the
bin. The vibrator shown in Figure 26 was then placed on top of the
backfill material. As the machine was started the drag line pulled it
down the slope gradually in an attempt to move the material to fill in
against the planks on the Open end of the bin. The tubes were fOund to
have a tendency to float on the surface Of the material, thereby losing

the vibrational effect.

-48-

An attempt was made to fbrce the vibrator to travel deeper by
attaching four pieces of 2-1/2 - 2—1/2 - 1/8 inch angle iron to the for-

ward edge Of the tubes as shown in Figure 27. The fins were equally

4",! 3’5"

‘3.
G

$.:/

 

Figure NO. 27. Improved paving tube vibrator.
spaced and so attached that one side of the angle iron extended down from
the bottom of the tube. The machine was then employed as explained
previously.

Four passes of the machine were made, Figure 28, at a speed of
approximately 4,000 vpm with the speed of drag approximately 15 feet per
minute.

The modified model and the method employed seemed to be very effec-
tive on large quantities of material. The addition of the angle iron
fins caused it to dig in and effectively apply the vibrator.

A summary of all outdoor bin tests follows:

.49-

l. Bullet-nosed vibrators were found to be very inefficient since
they only proved to be effective over a very small area and tended to
loosen the material when they were withdrawn. It was too heavy for
manual handling but did not warrant mechanical handling.

2. Prong type or penetration type vibrators, both the plank with
spears and the vertical tube vibrator, were not satisfactory since they
had a very limited consolidation range and left holes in the material
when they were withdrawn. With the tubes or spears sixteen inches
apart, there was an unconsolidated zone between them. Both types require

mechanical handling due to their weight. NO settlements were observed

 

 

Figure No. 28. Paving tube vibrator in operation.

on the bulk of material with either type.
5. The puddler, both the standard type and the modified type,
gave good consolidation on the surface and to a limited depth. Either

could be handled manually and should be employed on very thin layers,

—50-

six to eight inches, and on small scale backfill Operations.

4. The vibratory tamper had a very limited range in its compactive
effectiveness. It gave good results on the surface but could not
effectively set the large mass in motion to settle the material other
than that consolidated by the striking force and that in a local area
under the tamping shoe.

5. The horizontal tube vibrator, as employed in paving Operations,
proved the most effective and efficient type. With the tubes being
pulled over the surface the material was consolidated to a slight
extent and very little settlement resulted. Much Of the vibration
effect was lost and the material was compacted on the surface only.

With the tubes buried in the material and allowed to vibrate in
this position, very good results were obtained and the maximum degree
Of settlement of all tests resulted. The vibrator was slowly lifted
from the material allowing time following each lifting for the vibrator
to consolidate the material filling the void left by the tubes and
motor. NO voids were left in the fill material when the vibrator was
lifted to the surface. The consolidation was uniform for the entire
,depth. The area affected was the entire width of the bin and approxi—
mately five feet wide. This type of vibrator could be Of varied
lengths and made up in sections twelve feet long, while the vibrator
under test was a special designed section eight feet in length.

When employing the horizontal tube vibrator as a drag to haul the
material down the inclined SIOpe and compacting it against one end of
the bin, it proved efficient and adaptable to backfilling operations.

Employed in this manner, it served a dual purpose in spreading the

-51-

material and consolidating it in one Operation. As the density results
indicate, each additional pass tends to further consolidate one to two

feet down. This action would be very favorable and reduce any tendency
toward surface fluffing of the materialr

6. The pressure cells were not too efficient since the material
was compressible enough that when the pressure was applied to the cell
it would force the material away from the face of the cell and obtain a
pressure reading. On releasing the pressure from the gauge and the
cell, the cell piston would return to its original position leaving a
void on the face of the cell when the material did not rebound. The
failure of the material to rebound would prevent any further readings
with the cells with additional manipulation of the mass of material.
Under this condition reliable readings could not be obtained in all
instances; however, sufficient readings were obtained to Qualify that
no additional or excessive lateral pressures were a resultant of the
vibratory equipment employed.

7. The frequency of vibration was varied on all tests and it was
found that best results were obtained in all cases with a frequency
varying between 4,000 me and 4,200 me. When higher frequencies were
used the same results were obtained with no apparent advantage. Lower
frequencies tended to slow the consolidation since the effectiveness of
the vibrator seemed to be in getting the entire mass of material in
movement at one time, and low frequencies would not satisfy this condi-
tion. Each type of vibrator has a natural frequency and when this is
maintained the best results were obtained. Higher than natural frequency

also tends to add load on the motor and damage it with continued use.

-52..

The above Observations clearly demonstrate the effectiveness of
the horizontal tube vibrator or paving tube vibrator as it is commonly
known. For work in excavations and around footings of abutments the
manually operated platform vibrator seemed to Offer the greater
possibilities.

_ NO further study was carried out on the paving tube vibrator for
field use, but the platform vibrator was modified and was employed on
a field construction project west of St. Ignace, on US-2. The work
consisted Of constructing a metal crib retaining wall shown in Figure
29. Each crib or cell was rectangular in shape with dimensions Of 10
by 10 by 20 feet. The crib was built up in sections and the backfill
was placed in each crib '
simultaneously with the
crib assembly.

Controlled density
was specified for the
backfilling Operation in
the cribs. Since the

contractor was using

 

tamping to consolidate

Figure NO. 29. Cut River metal crib.

the material it afforded
an excellent Opportunity to compare vibratory means Of consolidation
with hand tamped results.

The platform vibrator as it was modified for the work is shown

in Figure 50. Additional base plates that were also employed in this

-55-

study at Cut River are shown in Figures 51, 52 and 55. The modified

platform vibrator with the base plate shown in Figure 52 proved to be

the most effective.

 

Figure NO. 50. Modified platform vibrator.

The modified platform vibrator was found to travel under its own
Power, due to the vibrational motor, at a rate Of approximately ten
feet per minute. This eases the problem of handling and manipulation

to a great extent.

-54..

 

Figure NO. 51. Platform vibrator with sheepsfoot base plate.

The material used to fill the cribs was dune sand and was placed
in the cribs in layers of approximately nine inches. To effectively
cover the 100 square feet of area with the tamper it took the con—
tractor's equipment about fifteen to twenty minutes. With the vibra-
tory equipment the same results were attained in approximately three
to four minutes. Detailed results of the study can be found in
Appendix D.

The platform vibrator was also used on two other bridge abutment
projects in Michigan. The vibrator model was modified slightly for
ease of handling, but the platform principle and plates were the same

as used before. The modified units are shown in Figures 52 and 55.

-55-

 

 

 

Figure 32. Improved platform Figure 53. Improved platform
vibrator. vibrator.

With the modified platform vibrator, described above, there is an
effective radius of approximately twenty—five feet when it is operated

on a sand material. This conclusion was reached after Observations Of

the motion of the sand surface while the vibrator was in operation. It

~56-

can also be verified by readings with a reed tachometer placed at
various locations in the area surrounding the vibrational unit.

The most effective frequency of vibration for the granular mate-
rials experimented with On the field projects was found to range from
2700 to 4000 me. With the possibility of encOuntering materials that
'would fall out of this effective range, further study is now in pro-
gress to determine vibrational frequencies for varying materials.

With the equipment used in the field tests to date it was found
that layers of granular materials up to three feet in thickness could
be effectively compacted. However, with layers of this depth, the
duration Of the vibration time was considerable to fully consolidate
the material. This study has proven that layers Of approximately 12
inches can be compacted most efficiently by vibrational mediums.

On the majority Of the tests one pass Of the vibrator was suffi-
cient to Obtain densities Of 95 per cent of the maximum density or even
higher in some instances. Time density studies have shown that 95 per
cent Of maximum density can be attained in approximately 15 seconds.
Additional vibration will obtain higher density at a very slow rate,
but it is believed that densities equal to 95 per cent Of the "Cone" or
Vibrometer densities are satisfactory. NO detrimental settlements will
occur in granular backfills or embankments when compacted to this
degree.

Density determinations in the field tests were made by several
varying means including calibrated sand, heavy Oil, mold Of known vol-
ume, and rubber membrane density apparatus. The rubber membrane appara—

tus method was adopted as the most satisfactory under average field

conditions.

-57...

 

Although there is a definite need for additional study on vibra-
tional equipment and additional techniques in their operation, it has
been definitely proven from this study that to obtain satisfactory den-
sities and thereby eliminate the present—day settlements in backfills
or embankments there must necessarily be some Sacrifice in the speed of
placing granular materials. Some definite conclusions have been arrived
at from the foregoing studies and are outlined briefly in the following

paragraphs.
CONCLUSIONS

The gradation of backfill materials was found to be a very import-
ant factor in any location or materials where there is any possibility
of infiltration of adjacent material into or through the backfill. If
there is no possibility of any infiltration, gradation is not of major
importance, but will also facilitate in consolidation. With backfill
materials in adjacent materials of clay, which would be impervious to
water moving through it, gradation control would not be required. How-
ever, if a granular subbase is intended over the granular backfill mate—
rial, some precaution should be taken to prevent any infiltration of
the subbase fines down into the backfill material.

The necessity of consolidation for granular materials was defin-
itely answered by this study from the findings that volume changes as
great as 19 per cent by volume could and did occur in granular backfills
Which were placed without consolidation means. The amount of volume
change varies with types of material and consolidation mediums, but all

granular materials unless properly consolidated under the vibrational

~58—

effects of traffic and weathering thereby result in detrimental settle-

ments which will disrupt the surface treatment.

Conclusions as to the most efficient means of consolidating granu-

lar materials derived from the study are:

l.

Satisfactory consolidation can be obtained by hand tamping
or mechanical tamping if layers are restricted to six
inches or less. This method proved very slow and
expensive.

Penetration type vibrators; i.e. bullet—nosed concrete
vibrators, vibrating spears and tines, did not prove effi—
cient mainly because the vibrations are localized in a
very small zone.

Surface type vibrators; i.e. modified paving type vibrators
and self propelled platform vibrators, proved to be the
most effective and efficient type. The platform type
vibrator will give satisfactory densities on layers of
granular materials up to 12 inches in thickness. The
paving tube vibrator would be adaptable for large uncon—
fined areas such as deep bridge approaches, embankments,
and pavement subbascs. Platform vibrators are adaptable
to areas inaccessible to larger equipment. Modifications
of the platform vibrator for use on larger areas would be

to employ it as a "gang", made up of three or more units.

A controlling conclusion is the cost of placing backfill materials

when employing vibrational equipment is less than that when placed in

accordance with the requirements of current standard specifications.

APPENDIX A

DUNE SAND DENSITY STUDY

Bridgman Sand

Loose densities with varying moisture content.

 

 

 

 

 

Per Density Per Per Density Per
Test Cent Lbs./Cu.Ft. Cent Test Cent Lbs./Cu.Ft. Cent
No. Moisture (Dry) Voids No. Moisture jDry) Voids
1 0.09 95.50 45.7 9 12.2 75.65 54.5
2 1.15 78.00 55.0 10 15.6 75.54 55.8
5 .57 76.97 55.9 11 17.6 76.70 '55.8
4 5.50 75.56 54.6
5 4.50 75.25 54.7
6 5.90 74.71 55.4
7 7.40 74.71 55.0
8 11.90 75.65 54.5
Loose densities and densities after inundation.
Water introduced from top only.
Per Density Per Per Density Per
Test Cent Lbs. /Cu.Ft. Cent Test Cent Lbs ./Cu.Ft. Cent
No. Moisture jDry) Voids No. Moisture (Dry) Voids
l 0.00 95.5 45.6 0.00 95.65 45.6
25.75 96.06 45.5 22.58 98.14 40.9
25.60 96.06 45.5 8 1.12 77.08 55.6
26.50 96.54 45.5 27.70 88.65 46.6
2 0.00 94.04 45.4 9 0.00 94.97 42.8
22.12 96.5 42.0 25.51 97.94 41.0
5 0.00 95.59 45.6 10 1.25 75.77 54.4
25.04 96.54 42.0 24.16 86.10 48.1
4 0.00 95.59 45.6 11 1.04 77.06 55.6
22.90 96.54 42.0 25.26 87.86 47.1
5 1.75 75.5 54.6
26.05 88.95 46.4
6 1.09 76.5 54.0
24.47 88.76 46.5
7 1.99 74.15 55.5

 

___.____——

Loose densities and densities after indirect inundation.
Water introduced from sides of model into the sand. A load of
2-1/8 pounds per square inch was maintained on the surface of sand.

 

 

Per Density Per Per Density Per

Test Cent Lbs./Cu.Ft. Cent Test Cent Lbs./Cu.Ft. Cent
No. Moisture ((Dry) Voids No. Moisture ((Dry) Voids
1 0.00 95.59 45.6 5 10.64 _ 70.05 57.7
27.50 95.88 42.2 21.58 88.17 46.9

2 1.55 75.59 54.6 6 5.50 75.10 56.0
29.05 89.57 46.0 25.99 90.60 45.4

5 5.00 75.51 55.8 7 7.05 75.10 56.0
27.50 89.02 46.4 25.15 90.40 45.5

4 0.00 95.59 45.6 8 11.19 71.50 56.9
10.60 95.88 42.2 26.57 88.55 46.7

 

Dune sand densities as determined
by the cone method.*

 

No.0f No.0f

Per Density Per Blows Per Density Per Blows

Test Cent Lbs/Cu.Ft. Cent Per Test Cent Lbs/Cu.Ft. Cent Per
No. Moisture (Dry) Voids Layer No. Moisture (Dry) Voids Layer

' 1 0.00 104.1 57.2 15 15 7.50 102.8 58.1 15
2 0.00 104.7 56.9 15 16 7.40 105.5 57.8 20
5 0.00 104.7 56.9 15 17 6.90 104.2 57.2 25
4 0.00 105.1 57.9 15 18 10.20 105.1 57.9 15
5 0.00 104.9 56.8 15 19 9.80 105.2 56.6 20
6 0.00 105.5 56.6 15 20 9.80 104.7 56.9 25
7 0.00 105.2 56.6 15 21 8.50 105.5 56.6 20
8 0.00 105.6 56.4 20 22 15.60 105.0 56.7 15
9 0.00 105.5 56.6 20 25 15.70 105.4 56.5 20

10 0.00 105.7 56.5 25 24 11.70 107.0 55.5 25

11 0.00 105.9 56.2 25 25 0.00 104.9 56.8 15

12 5.70 101.5 58.9 15 26 0.00 107.0 55.5 20

15 5.60 102.4 58.5 20 27 0.00 107.1 55.5 25

14 5.50 105.5 57.7 25

 

* Cone method described in Appendix B.

2 (A)

Dune sand, densities loose and densities as determined
by different degrees of vibration on the vibrating

 

 

 

 

 

 

table.
Per Density Type Per Density Type
Test Cent Lbs./Cu.Ft. of Test Cent Lbs./Cu.Ft. of
No__Mgisture (Dry) Vibration No. Moisture (Dry) Vibration
1 0.00 95.70 4 Minutes 5 0.00 95.70 4 Minutes
Mild Hard
100.00 106.44
2 5.00 71.12 4 Minutes 6 5.09 71.55 4 Minutes
Mild Mild
96.00 99.01
4 Minutes
5 5.42 70.49 4 Minutes 106.41 Hard
Mild
96.59 7 5.12 71.99 4 Minutes
Mild
4 7.66 71.77 4 Minutes 97.19
Mild 4 Minutes
96.28 102.50 Hard
Dune sand densities as determined by
the Standard A.A.S.H.0. Test T—99
Per Density Per Per Density Per
Test Cent Lbs./Cu.Ft. Cent Test Cent Lbs./Cu.Ft. Cent
No. Moisture (Dry) Voids No. Moisture (Dry) Voidg
1 .58 104.15 57.5 9 1.24 100.5 59.5
2 5.00 105.71 57.5 10 2.75 105.9 57.4
5 6.16 105.46 57.7 11 4.66 105.5 57.6
4 8.56 105.46 56.5 12 6.42 104.1 57.5
5 12.22 105.58 56.5 15 8.58 104.4 57.1
6 11.78 105.88 56.2 14 11.11 105.0 56.7
7 15.08 105.85 57.5 15 15.02 105.9 56.2
8 0.00 104.20 57.2 16 15.89 105.5 56.5

 

 

5 (A)

-J_‘.J.A_l ____ - a

PEA GRAVEL DENSITY STUDY

Loose densities with varying moisture content.

 

 

 

 

 

Per Density Per
Cent Lbs./Cu.Ft. Cent Remarks

No. ___Mpisture (Dry) Voids ‘

1 0.0 94.54 41.7 Average of 10 tests

2 0.0 94.54 41.7 " " 11 "

5 0.5 96.29 40.7 " " 2 "

4 0.5 96.29 40.7 " " 2 "

5 1.0 95.06 41.4 " " 10 "

6 1.0 95.71 41.0 Max.Den." 10 "

7 1.0 95.71 41.0 Average " 4

8 1.0 95.07 41.4 " " 6

9 1.1 94.45 41.7 Min.Den." 10 "
10 1.1 94.45 41.7 One Test

11 1.7 95.44 41.2 Average " 2 "
12 1.7 95.44 41.2 " " 2

15 1.9 95.24 42.5 " " 2

14 2.2 92.50 42.9 One Test

15 5.0 95.08 42.6 Average of 5 "
16 5.0 95.08 42.6 " " 5 "
17 5.8 90.16 44.4 " " 5 "
18 4.5 90.00 44.5 " " 6

Pea gravel densities by inundation.
Per Density Per
Cent Lbs./Cu.Ft. Cent Remarks
No. Moisture (Dry), Voids

1 12.44 95.0 41.4 Indirect ponding
2 12.50 95.5 41.1 Direct ponding over surface
5 4.50 90.0 44.5 Saturated and drained
4 5.80 90.2 44.4 “ " "
5 1.90 95.2 42.5 " " " #

 

4 (A)

Pea gravel densities by vibration.

 

 

 

 

was-.m- __ ._ . . .

 

 

Per Density Per
Cent Lbs./Cu.Ft. Cent Remarks

No. Moisture (Dry). Voids
1 12.58 101.7 57.2 Mi1d Vibration
2 12.44 101.4 57.4 " ’ "
5 9.70 106.6 54.5 Mild vibration for 5 minutes
4 10.07 104.8 55.4 " " " 2 "
5 10.14 105.2 55.2 " ” " 4 "
6 9.65 104.9 55.5 " " " 5 "
7 10.88 105.5 56.5 " " " 2 "

Pea gravel densities by the Cone method*
Per Density Per
Cent Lbs./Cu.Ft. Cent Remarks

No. Moisture (DrY) Voids 11_
1 Dry 106.2 54.5 All samples consolidated
2 " 108.2 55.5 in three layers with 20
3 " 107.7 55.6 blows per layer.
4 " 106.5 54.5 '
5 " 108.8 54.1
e " 107.1 54.0 #

 

 

 

5 (A)

Nuu‘sxknu h§s) nu IIIIII

 

muvfikhtkflx 0 6 H I

 

0a.?

“(‘0 ."IV‘E>\ N‘k‘t I‘Q “\hl at:

§.§ 4‘\~N§ “hr“ k0 HN\N. Oklfl\

u\‘ .Q

 

 

 

11

 

Q‘quk‘t $.5de K0 NNxb R b\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    

«89% aghast}? no I

1 .I. 1 .11111 1

86
\Nu\§o§§~ our ..u IIHNI

“h.“ .1“ .‘Nk. oh 5‘
.V‘\E ”k V\..\ ‘0 .NN\n. *hux

“a.“ \Q HN\9 «5.9.
i‘tflR‘nstnNuuxv. ...G 38 fibx

seal

‘.§

 

 

00000

.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

    

 

 

 

 

 

 

 

 

 

I 1+_ 4 .0 0 0. PI.7
m M ..I" f .0 0 r. k. 0 1‘1 H.-
3 i I \l :0. .l
U G H 01700751 H
u. u u .11. - 1. u
I d I . I
m m w 10:..010. .+ 1 47014.. ..017 M.
w s ”0‘ m .3 0 . 00 .%0000 00 4 94.. ti +7 9 M
m M 9 1004+?0 i4+r4 1 .
0 0 0.»..00. 0. 0.74711. 7- .
. . 7 01.004.01.70? 004.
a c 90 4 7 030....0 .:. ?.17 «10%.4 1.0.
or .I. .
Y0. 0 F151 ..J. .4 . 0
.. .
c:— c .I..
c o I n u o n I I .. l .I I I t I I I
«mmuuon . um .mm. -- -m-u.m a m mmmm «unnum- . om mu ...- an. n... a w mum...
ill-1 ' Sill-a— E—i—2—1‘ .3 .5...:!.-
._ _ I . .. .. . 2 .n a. 0.. I ... In in ...-..I ..I ... _ .. .. ._ 2 on ... .... 3 c: an in Eiz}
E4 )5..— 1554 I... {.2 I 1 El II . Sautuua .I.; H ...I...—
o u.
c b- . . . I, . n u . . .I.? 1: II A
...... ...“... 091......“ N, 1 u... .. # .e .. .... . ............ .....u. 52...... ...; I H.
p . m 1 ”LI: n
7...»...27915 ...—..I. .... .5... E 5:... ...-ti. v I...:.._..7$._ . 1....7 ... 3 1...: 1.15.7 ._ :7 . .v
Q .
N. 8. KnMNN 5‘86 “\L 09$“ \ \\\ .5..th 3.3.x $.35...“
«W433 u... H l‘tssnxunuxh hast to has, «.8 . \1 .330 I 45...; .46..th 83. RR.
bu)“ IV Mn... Nhfltkfi HNxb Rb. Minixkbib 0‘0. 1 00.5 I §t§¥Kkfi NanRb.
.I.. 3: :
I ...-310.. .I.... . ..I
m u .m u
I ... u. H . N
m n n m
m . m m
.... m ... a
u .. m m ... m
H .. H u.
a in
. ' u ' .I . a. .I 5 ' . . '- C u .l 0| . Q ' I. . I . .-
-.uu... . .m gasses“... . In. m.u...... . .m.aem....: . I...
ilk-0J1 i]:
.... I a. I .... . 0. 8 0. .- ... In 08 [-1.5 ...... x r t 1 . 0. i 8 0. .- ... In 0k 5......
11y... .I....IcIE I]: I ...-1.. 3 ~ ...I...-
0 I
In: . O
14.....11 941. #11411...”— .. H .... .... ...... 4...... .I.. I0... nu. WM .. N .... n
6“: ...Itl - H ..I-1Hfl. {MU-1‘15. 1.... I. .9: -(I‘ I. 55.0“... ...-...I.- 1 1 Wm-

 

 

 

 

 

Mm..mi

w- a.

m

 

 

 

 

Nt§§X~ g :8 ‘Nfl‘
‘ . ... .

 

 

J: 7)) Jar”? N/ Jam“

 

 

hxfllh’ffi :J/gnuw éJ/VJD :” Q
\

a V g1 .‘S §8\
VV\‘ .‘3 RBI. ‘t Ki‘vkth.
QVK§§KhV

 

 

 

h\nfi\g
sgxlgg §b\%\\90s\9. ti!‘ 3
\. kixbbrghg 33905.; ...v 553...

 

g\K\\.U\L\\§ gs .OQV
\«\\\\%\\\§h’ \<\b\\\\.vR\MQ
bv\.\h.\§ \«k \K.>\I&

 

 

.I..

11

 

. M. \m. .3 .vq\\\“ Vw\b\\v§ \QQNQ

 

0 000000000

. . . 00040...

(IT4\I\L'«III 1V4 I "1.!

0000.077: .0 0

 

'|\|\\Vd l\l n n M

 

 

 

 

 

 

 

 

 

HOS I30.

‘9 .9‘ \KKWK AQKK‘Q‘NRV

5V 4‘8.ka v KN H

 

«32s.!» a .0... H

(I‘I\I\L'fll IV! I H M

V‘Q‘S ”2‘ KQ .‘Nsh. g
V‘Qfi‘t VINV\IM\ h\° ”N“. i .n.\

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\hl l\ 1 ' II M

"\l‘

APPENDIX B

THE COMPACTION AND DENSITY 0F GRANULAR SOIL

SCOPE

 

1. This density test for granular soil determines the dry weight
per cubic foot under a standard method of compaction.
APPARATUS

2. The apparatus shall consist of the following:

(a) Mold — A funnel shaped mold having a solid bottom in the
large end and equipped with a stopper for the small end. The bottom
Shall be so shaped that there will be no sharp corners inside the mold.
The base or large end of the mold shall be approximately 5-5/4 inch in
diameter and the small end shall be not less than 2-1/2 inches. The
mold shall be approximately 8-1/2 inches in height and shall have a vol—
ume of approximately 1500 cubic centimeters or 0.0459 cubic feet. (See
Figure A.)

(b) Balances — A balance or scale of 6000—g. (12 pound) capa—
city sensitive to 1.0-g or 0.01 pound; and a balance sensitive to 0.lO-g.

(c) Drying - Any approved method for drying soil samples.
PROCEDUBE

3. (a) A 5500—g sample shall be taken from the source of the mate-
rial to be used.

(b) The sample shall be thoroughly mixed, then compacted in the
mold in three equal layers, each layer receiving 25 blows. The blows

shall be delivered by raising the mold approximately 4 inches and

striking it sharply down on a concrete or heavy timber base. After the

third layer has been placed the blows shall be continued with the wood

stopper reversed and held firmly over the opening. Sand shall be added
at intervals to keep the mold full and operations continued until no
further consolidation occurs. The compacted soil shall be carefully
leveled off to the top of the mold and weighed.

(0) The weight of the compacted sample and mold, in pounds, minus
the weight of the mold, shall then be divided by the volume of the mold
in cubic feet and the result recorded as the wet weight in pounds per
cubic foot of the compacted soil.

The wet weight per cubic foot can also be obtained by dividing
the weight of the sample in grams by the volume in cubic centimeters and
multiplying by 62.4.

(d) The compacted mass of soil shall be removed from the mold
and a sample of approximately 100 grams taken from the center of the mass

shall be weighed immediately and dried to determine the moisture content.

QgLCULATIONS

 

4. The moisture content and the dry weight of the soil as compacted
shall be calculated by means of the following formulae:

Moisture, per cent = %Z% x 100

where -

A is the weight of dish and wet soil
B is the weight of dish and dry soil
0 is the weight of the dish

Dry Wt'/°u°ft' 0f compacted $01 Per cent moisture + 100

2 (B)

 

 

b -
. o
‘ N
I
'u
l
.
.
,(_
A
P
I | y'
I v
I. I, - -
3
, .
_ /

V ”D
I ‘ ' I V . .
. , r4 . - ‘

 

 

O VNH‘Q Zumlnmq

 

 

 

omooa H.mu om.N
.mumsma gs an emaama m.mm mm.¢
.cOpran> mmvsmpxm 0.0m H.mH mm.ma m.ma w.N mmam
mmooq H.5OH mo.H
panama >.¢NH wo.a mpdmmpmm<
.nowpmpnwb vmwzmpxm m.NmH H.ma mm.ma m.ma mo.a mmnmoo
omooq N.om us.¢ csmw
.nOflpmpnab Umpsopxm o.Hm m.moa u.wH o.mH m.ma >5.¢ amnspmz
= m.moa >>.N
mmooq o.mm m.NOH om.H
= = = : H.mNH mb.o
.mposea gs no emaaae H.mma om.¢
s a g H.maa m.NH mm.ma m.ma >>.N Hmbmpo
.cmpsupwb .Ho> Hmpoa m.vm m.>HH m.ma om.mH m.ma om.a mom
.mpmsma gs as soaama m.HaH mm.”
.mssaob Hobo» mo .nflb.pxm m.wva s.¢a oo.oa m>.HH mm.¢
.Hmpos mnflaaam mafia;
nowpmnnfib msosnfipcoo o.HmH oo.n
.msdao>
aspen mo nowumpnw> paw: u.HmH N.ma ow.ma m.ma mH.m Hmbwpw
mmooq 0.5N o.moa m.mH ma.m sum xcmm
mm-a..o.m.m.<.< . sesame auu- m.moa o.HH
sowpwh3H> wmvsmpxm m.mm v.moa N.o
nowpmnnwb UHHE m.mm m.ooa N.o
mmooq s.mv m.mm N.o wswm mafia
mUHo> Amway mwdmao dough .mwpo mndpmwoz1
mxHQSmm pnoo .pm.50\.mng madao> .Ph .50 pcmo Hmwpmvoz
pom mpwmson puma pom madao> mom

 

.pcmepdnw Houos CH opma mpmmp cowpmpflaomaoo

o xHszmm<

Eu‘ECEE-fifl II. c

EC EELEIFI"“ ‘lli

¢B¢Q 8mm .H.

 

ADV N

 

 

 

 

 

 

 

 

 

 

 

 

 

m.mm --- mm.¢ u: oa.ma --I .om.m :tuu ¢HN.H mas mmo.a a seam ca.
s.«ma --- mo.a -- oa.ma sin goo.m ---- cow.a was mso.m = .mma
omndoo m
m.mma --- ms.o u- oo.ma :uu =ma.m sun- Hmm.H owe Hmo.m g Hopwpo
mom m
m.sma --- ow.m u. om.OH ..- goo.m u.-- mH«.H new mam.a g seam s
m.m¢a 1.: ma.m -n mm.0a nu- =mm.m 11.: mmm.H was omm.a assess Ho>wno m
0.00 «.ms om.~ H.ma mm.ma om.mH =m Hana omm.a was mmm.a . seam m
H.NNH H.moa mo.a H.ma mm.ma om.ma gm Hana mam.a was a¢H.N z .mma
.wanHSmmn mmndoo v
masaob a“
«means on m.mHH m.m0a as.m H.ma mm.ma om.ma gm Haas omm.a owe m¢H.m g Hm>apo
spas amps mum m
Inapsm who:
mpmmp Has ¢.moa ¢.om sa.¢ 5.0H oo.na om.ma =mm.m Hana 55¢.H mmw «mm.H a seam m
aoflpauaaom
unoo umpoa ¢.¢wa o.mma no.» s.aH oo.oa ma.aa gm gs ma¢.H was mmm.a n.aonna> Hopano H
.Homaoo .Howaoo my: umno quwm wwda .Homnoo .Homaoo fimvnsomv nmvuson aoapmv
hmpm<wLnouommm Ipmflo: asao> Imago gowns mnommm Hobmpo xom Hobwho IHHomnoo Hmanmpds .oz
mxpmsmm puma ammo Naph.auu Mom no mo mo was mo puma
.pa.:o\.mng pom “mm masaos mos godmm snags: .p: mom mo soaps:
hwwmnon mowunzm Hobwuo onmamz

 

 

 

 

 

 

 

mom Hzmaabm4 Janos mma 2H mmomqm mA¢HmmB¢E

BZHMHhhHQ ho mmHBszma Qz< mcz<mo mSqu>

«Eda Hams

APPENDIX D

Consolidation of bank-run gravel in a bin
of approximately 1000 cu. ft. capacity.

 

Description of

Dry Densities in
Lbs. Per Cu. Ft.

 

 

Test Consolidation Method 1 Foot 2 Fodt— Remarks
No. Surface Depth Depth
1 Loose density 107.5(1) -—— 108.5(2) (1) Ave. of 7 tests
. (2) n n 5 n

2 Vertical spears or
tines on 2-foot layer 109.2 108.1 107.1 Averages of 6 tests

5 Paving tube vibrator
in vertical position 116.0 112.6 —~- Averages of 5 tests

4 Paving tube vibrator 10 Min. vibration. if
dragged over the sur- Before modification A
face. 111.5 107.5 107.5 of vibrator.

5 Same as No. 4 108.2 109.7 110.1 20 Min. vibration.
Before modification
of vibrator.

6 Same as No. 4 and 5. Vibrator buried

(See remarks) 115.6 115.7 115.5 under 2' of fill.

7 Tube vibrator dragged Vibrator modified
down sloping face of ‘as shown in Figure
fill. 110.9 115.9 115.5 27.

8 Tamped 117 .2 111. 3 ---

9 Platform vibrator 109.2 110.5 106.0 Original platform

vibrator.
10 Platform vibrator. Ave. 109.4 after one pass.
(Modified) Ave. 111.8 after three passes.
11 Filled one end of bin 111.0 -115.7 115.5 Placed a 2' lift and

 

to the top allowing
it to flow down at
natural angle of
repose. Dragged vi-
brating tube down
through fill mate-
rial, pushing some
ahead of the vibrator
and compacting it
against the opposite
end of the bin which
simulates the abutment
face.

 

 

 

 

 

 

 

 

vibrated 20 minutes.
From these tests
this method appears
to be the most
effective and adapt—
for large volumes

of backfill.

 

Consolidation of sand backfill in metal
crib retaining wall.

 

 

Moisture
Description Content Dry
Test of Per Density Remarks
No. Test Cent Lbs/Cu.Ft.
1 Laboratory 12.2 108.1 Tamped
2 Natural Sand 5.6 98.8 Undisturbed
5 Field Check 5.6 111.5 Prolonged tamping
4 " " 5.5 105.6 Normal "
5 " " 5.8 108.2 " "
6 " " 6.5 102.5 " "
7 " " 5.0 105.5 " "
8 " " 5.5 105.8 " "
9 " " 4.9 102.5 Vibrated - 5" plank base with metal
base plate having large corrugations.

lO " " 5.8 99.9 Same as No. 9.

ll " " 6.8 101.6‘ Same as No. 9 except metal base plate
with small corrugations.

12 " " 6.9 95.6 Same as No. 11. _

l5 " " 5.4 105.5 Vibrated - Plank base and sheepsfoot
metal base.

14 " " 5.9 105.5 Same as No. 11.

15 " " 5.1 102.2 Same as No. 9.

16 " " 5.7 104.4 " " " "

17 " " 5.7 109.0 Same as No. 9 except two passes were
made.

18 " " 6.8 110.6 5" plank removed for remainder of
tests. Base plate with small
corrugations.

l9 " " 5.6 107.4 Same as No. 18. Vibrated 4 min. one
pass at 5600 v.p.m.

20 " " 6.1 104.8 Same as No. 18. Vibrated 5 min. at
2400 v.p.m.

21 " " 8.4 105.5 Same as No. 18.

22 " " 8.2 114.8 Prolonged tamping with sheepsfoot base
plate.

25 " " 8.5 101.5 Same as 18, vibrated only 2—112 minutgs.

 

2 (D)

-‘*——-‘-' Lieu. -m.‘

 

FFNQIrl

 

Amv m

.Umgmpwaomsoo ma «mow esp op :m:osnp

pnwflh pmmma pmpfim opp scum .ucoaxcmnSm spawns mnp mobs pmswmpno mowpflmsmp

esp mm moapflmmmp poom mm :H pHSmmh pom woos mowpmpmmo mo mama mama .hmpws

amps: vmomad mm; HHHM on» mo pooh moans .poppmpm was soflpmpwaomsoo whommp
poomaa Haamxomn mo boom xwm haopmsflxonmmm mm: chomp pommoam mflnp so

 

.Hcmmltomv
# A=QNI=NHV A=OHI=NV Azm

 

ngmv so mspame aopm.pmaa m.m m.soa = g . ma
gma a massannfl> smpom.mess m.HH N.OHH = = = «a
mama ma s .sa.na.ma same 4.5H o.mHH = a = ma
.Azmmlsmmv w A=ONI=NHV v.HH H.0HH a = = NH
Agoaugmv Acmu=mv so assume o.m s.soa = = a HH
SCAM Q PMflH :mm m Go mvds O.NH $.0HH : c = OH
who: HH 5 oa.m.m spams a.HH 4.0HH g = = m
.A=omu=mav a A=m n =mv m.m o.aoa w.maa m.mOH m.mm m
.Agmucmv mo mspame scam a.ma m.aoa c = = a
a pmsfl gma a no mess mama m.ma s.Hoa g = = m
a s.m.m mpmm9.A= mngmmvs m.sa o.moa = = = m
A=omn=mav.fl=au=mv.Azau=mv m.m m.sm g = g a
mo wapame sons swamp m.ma o.moa a g a m
mhmk Una pMHH zmm w so ¢.©H O.¢OH a = c N
mesa who; a a m.m.a mamas H.ma s.moa m.saa m.mm m.mm H
dwmmo pmmq\‘ (Amsoou ANHQ mmooqv
Haas ca spamamm .pa.:o\.mpq .pa.su\.mpq .pa.so\.mna .oz
mthsmm oawqxoOpm Uamam hpflmsmm hpamcmm mfimema pmme
mmmxsflnsm shops knows QHHQ
nbopaq nachos soopm

 

.Hauaaosp mmeutn samsmsss so coasmeHHquoo

 

 

_|