Documente Academic
Documente Profesional
Documente Cultură
8.1 Definition
Compaction is an artificial process in which through mechanical means soil
particles are constrained to pack more closely resulting in an increase in density (unit
weight) and a consequent decrease in void ratio.
From what has been stated above it is obvious that a well graded soil achieves
greater compaction than a uniformly graded soil for the same energy input, since in
the former, particles of increasingly smaller size are able to occupy and fill the void
space enclosed by the successively larger particles (see Fig 6.1).
A cohesionless soil such as sand, and a cohesive soil such as clay, are both
discrete systems composed of particles. They, however, exhibit a difference in terms
of continuity. Even though discrete, a clayey soil is a more continuous medium
because of the cohesive bond existing between the particles, while in the case of sand,
continuity is merely the result of the mechanical contact between the particles. A
cohesionless soil with some clay content to hold the particles together – normally
specified as fill material for embankments such as for highways and railways – is more
amenable to mechanical compaction than highly cohesive soils where the energy input
may be insufficient to break the cohesive bond between particles.
The input of mechanical energy can be in the form of tamping or ramming (Fig 8.1),
running rollers (Fig 8.2) or even by vibration (Fig 8.3) all of which enable the soil
particles to move into closer packing. Such movement is aided in the presence of
water, which serves as a lubricant in this case, enabling the easy movement of
particles gliding past one another into a state of higher density.
3. The decrease in the void ratio, e, makes the soil less permeable, resulting in
lowering k, the coefficient of permeability.
At this stage, and before proceeding further, it shall be profitable for us to examine
the basic difference between consolidation and compaction.
Consolidation Compaction
1) natural – occurs when pressure is artificial – occurs by the input
Note:
‘Density’ is a term used in Physics where it means mass per unit volume (kg/m 3).
In geotechnical engineering we use weight per unit volume (kN/m 3) which is therefore
called ‘unit weight’. Actually the term density in relation to compaction is a misnomer,
as we are not dealing with mass, but only weight. The use of the term density here
follows a popular practice, which will continue to remain so, much as we may wish
otherwise. (It is important to realise in this context that ‘mass’ and ‘weight’ are entirely
different quantities (Kurian, 2005 – App.G, also Topic 52). While mass is a base
quantity (along with length and time), weight is a derived quantity as it is a force, viz.
the force with which earth attracts the mass of the object. Since f = m x a, weight =
m x g, where g is the acceleration due to gravity.)
3
Notwithstanding the terminology we use the same symbol 𝛾 (gamma) for unit
weight as the symbol for density, which in a sense helps us to overcome the confusion
created by the terms.
As against this, the total unit weight (γt ) which we shall call wet density (γwet ) in
relation to compaction,
w
𝛾𝑤𝑒𝑡 = 𝛾𝑡 = v
ww w- ws w
w= = =w -1
ws ws s
W W
Therefore, w+1=W or Ws =
s 1+w
γwet
Dividing by V, we get, γd = (8.2)
1+w
wet density
or, dry density = 1+water content
WS
Ws WS VS
Further, 𝛾𝑑 = =V = VS VV
v S+ VV +
VS VS
𝛾𝑆
= (8.3)
1+𝑒
= γdry , which from basic Soil Mechanics we know is the unit weight of dry soil, which
we now see is the same as γd , the dry density. Even though quantitatively the same,
there is a significant physical difference between the two terms. When we use γdry we
mean a completely dry soil; on the other hand, when we say γd we actually mean a
wet soil in which we are interested in only the amount of solids (as a measure of
compaction) and are not concerned with the presence of water.
As regards γt , the total unit weight (or wet density, γwet , for us)
w ws + ww
𝛾𝑡 = =
v vs + vv
w
ws (1+ w )
ws
= vv
vs (1+ )
vs
1+𝑤
= 𝛾𝑠 (8.4)
1+𝑒
from which
4
𝛾𝑡
𝛾𝑑 = , (same as Eq.(8.2))
1+𝑤
which shows that wet density (γwet) is the same as the total unit weight (γt), both
quantitatively and physically. (Note further that γd (Eq 8.3) is obtained by putting w
= 0 in Eq. (8.4).)
On the dry side the added water creates a coating of a thin film of water on the
surface of the particles imparting a lubricating effect on them which enables them to
move closer to each other under the influence of the compaction effort. This results in
the reduction of air voids which goes on till the peak is reached at which we obtain the
maximum dry density(γd max.) the corresponding moisture content being called
optimum moisture content (O.M.C.). Beyond the peak the added water starts
occupying the pore space preventing the soil particles from moving into that space. As
a result the dry density starts decreasing from the peak value. The corresponding
moisture content is an optimum and not a maximum value unlike dry density which is
a maximum value. The curve of Fig 8.5 is called the ‘moisture content-dry density
curve’ or simply the compaction curve. The moisture content corresponding to the
ascending part of the compaction curve is marked as the dry side of optimum and that
corresponding to the descending part as the wet side of optimum.
𝐺.𝛾𝑤
𝛾𝑑 = 𝐺.𝑊 (8.5)
1+
𝑆
Substituting S = 1 (100%) we can obtain γd for various values of w and plot the result
which is also shown in Fig 8.5 marked as ‘zero air voids curve’ or the saturation curve.
It is noted from the figure that zero air voids is a theoretical limit which can never be
reached in practice, signifying thereby that it is impossible to fully expel the pore air or
make the soil fully saturated in the compaction process.
Eq.(8.5) can be used to plot the w-γd relationship for different values of S such as
95%, 80% etc. One can also determine the S corresponding to γ d max_OMC or at any
point on the curve defined by the pair of values of γd_w.
Our discussion so far pertained to one constant value of the compactive effort, but
without specifying it.
Procedure
Measure the inner height and diameter of the mould. Take the weight of the empty
mould with base.
Air dry the sample soil, break lumps and sieve through a 4.75 mm sieve. Collect
about 15 kg (mass) of the sieved material for conducting a full test for a minimum of
five values of w. Take the first portion of 2.5 kg from the above, add little water and
mix thoroughly using both hands. Fill it in the mould in three layers delivering 25 blows
per layer after placing each layer by simply lifting the rammer and letting it fall freely,
through the cylindrical sleeve, on the fill vertically. The surface of the layer below must
be scarified for ensuring good bond with the succeeding layer. At the end of
compaction detach the collar, cut off the protruding soil at the top and take the weight
with the base. The difference with the empty weight gives the wet weight which when
divided by the volume of the mould gives the wet density. Determine moisture content
using three samples collected from different heights of the compacted soil removed
from the mould with the help of a sample extractor. The dry density is determined
from the wet density and water content.
Now take another 2.5 kg of the sample soil and add about 2 to 3 % more water
over the previous quantity and repeat the test as before. Continue with the test till the
6
weight reaches a maximum and falls by two or three values. Plot γd vs. w and join the
experimental results by a smooth curve, locate the peak and determine γ d max and
OMC (Fig 8.8). (Note: In general the peak will lie between two experimental results
and will not coincide with any one result. This explains the need for plotting the result
rather than go by the maximum value from the tabulated results.)
The mould of the modified Proctor test has an effective height of 127 mm, the other
dimensions remaining the same. The rammer weighs 45 N and falls over a height of
450 mm. The mould is filled in five layers and 25 blows are delivered per layer. The
compactive energy is of the order of 2726 kNm/m3 of soil against 605 kNm/m3 in the
Standard test.
Because of higher compactive effort the compaction curve will lie above the
standard test curve (Fig 8.6) giving rise to a higher γd max and a lower OMC, both of
which are advantageous from the field point of view, even though the gain is at the
expense of higher energy input.
The repeated raising and shifting positions of the rammer manually, is a strenuous
experience in the conduct of the test. To overcome this difficulty, electrically operated
machines have been developed which simulate the manual input of the compactive
effort to the great relief of those conducting the test.
We shall first discuss two tests for the determination of field density.
The apparatus consists of a cutter, dolly and rammer (Fig 8.9). The cutter weight
and dimensions must be noted prior to the test. The cutter with the collar fitted on top
is driven down by a weight falling on a wooden cushion (dolly) placed on top of the
collar.
When the driving is complete the side soils are removed by a spade. The cutter
with the soil is cut at the base and brought out. The collar is removed and the soil
protruding at the ends is trimmed off. The cutter with the soil is now weighed so that
the difference with the empty weight gives the wet weight of the sample which when
divided by the volume gives the wet density. The soil specimen is removed from the
cutter and samples taken for the determination of w. From γwet and w we calculate the
dry density.
The method is not suitable for dry sand as the same cannot be retained in the cutter
while lifting, besides the compaction induced by the vibration generated by the weight
falling on the dolly, which affects the natural density.
To obtain the field (wet) density we need to determine the volume of this hole which
constitutes the rest of the work.
This makes use of an equipment in the form of a sand pouring cylinder with a
calibrating can (Fig 8.10b). The volume of the cylindrical can can be determined by
measurement. The can is typically of dia. 100 mm and its depth is supposed to be the
same as the depth of the hole excavated in the soil. The sand pouring cylinder has a
conical (upright) funnel-like (inverted) part at the bottom with a stopper at the junction
which allows/closes the flow of sand from the cylinder.
The sand pouring cylinder is nearly filled with dry uniform sand (of size within 600
– 300 µm range). In order to determine the bulk unit weight of the sand the can is
placed at the bottom of the cylinder and sand is allowed to flow down filling the can
and the conical part. The sand in excess of what is needed to fill the can is removed
by running a scale or spatula horizontally across the top of the can. From the weight
and volume of the sand filling the can its unit weight is obtained.
8
Our next effort is to determine the weight of sand filling the excavated hole so that
dividing the same by the unit weight of sand gives us the volume of the hole.
For this we must first determine the weight of sand filling the conical part. For this
we weigh the cylinder with sand and keeping it on a level surface (Fig 8.10c) we run
sand to fill the cone, close and weigh the cylinder again. The difference gives the
weight of sand filling the cone.
In the final phase of the work, we weigh the pouring cylinder nearly filled with sand
and bring it on top of the cylindrical hole excavated in the ground and empty it until it
fills the hole and the conical part (Fig 8.10d). Now close the stopper and take the
weight. The difference in weight gives the weight of sand filling the cylindrical hole
and the cone. From this the weight of sand filling the cone already determined is
subtracted which gives us the weight of sand filling the hole alone. Dividing it by the
unit weight of sand gives the volume of the excavated cylinder. Knowing the weight
and volume of the excavated soil so obtained we can determine the field density, and
from the water content, the dry density.
We shall now look at a somewhat indirect method for the rapid determination of the
water content in the field which is sufficiently accurate as a routine measure for
controlling the amount of water to be added to reach OMC and the corresponding γd
max.
The first part of the test is the calibration of the needle. This is carried out at the
time of the Proctor test itself, in which each specimen in the mould is at a particular
water content. Before extracting the specimen from the mould, the needle is pushed
down vertically at the rate of 12.5 mm/s to a depth of 75 mm. The force needed to
cause this penetration, which is proportional to the compression of the spring is
marked on the plunger, which when divided by the area of the point gives us the
penetration resistance (N/mm2). The test is repeated on specimens at different water
contents and a plot is made between penetration resistance and water content (Fig
8.11b). This is the calibration curve of the Proctor needle. The spring of the spring
loaded plunger is a permanent feature, whereas the point at the bottom of the needle
can be changed to suit the hardness of the soil.
The field part of the work consists in bringing the field soil and compacting it in the
Proctor mould following the same specification as in the test. Now push the needle
into the specimen and note the penetration resistance against which we can obtain
the water content from the calibration curve as shown in the figure.
9
It is to be noted that penetrating the needle in the field will not give us the result as
the field density need not be the same as the laboratory density against which the
calibration has been prepared.
At small moisture contents the thin water films around the grains keep them apart
by surface tension, rather than bringing them closer. The result is that density
decreases reaching a minimum value at A. With increasing water contents this
situation reverses with the curve rising until reaching the peak at B, which corresponds
to 100% saturation, with the density decreasing thereafter with increasing water
contents.
Apart from the above, a change in the compactive effort has a much lesser effect
on cohesionless soils than on cohesive soils. Thus, in general, the compaction
characteristics of cohesionless soils are very different from those of cohesive soils.
The best means of compacting cohesionless soils like sand is imparting vibration.
For vibration to be effective the sand must be either fully dry or fully saturated. Density
index (relative density) is the best measure of expressing the state of compaction of a
cohesionless soil like sand. Vibro-compaction of soils, with its different possibilities, is
dealt with in greater detail later under Topic 31.
Rollers used for compaction of soil are divided into four types:
1) Smooth wheeled rollers,
2) Rubber tyred rollers,
3) Sheepsfoot rollers, and
4) Vibratory rollers.
10
This has one large wheel in the front and two smaller wheels at the rear. Since the
rear wheels are placed apart at a distance equal to the length of the front wheel (Fig
8.13b), a single pass compacts the entire width from end to end of the rear wheels.
These have pneumatic tyres fixed on two axles - with one number less on the front
axle - spaced in such a way (see Fig 8.14b) that together they cover the complete
width, as in the smooth wheeled rollers.
In this case the drum (wheel) incorporates projections which go their entire depth
down compacting the soil locally with the drum compacting the full width. (The name
comes from the resemblance of the projections to a sheep’s foot.) Fig 8.15 shows two
types, (a) the club foot type, and (b) the tapered foot type. These are normally towed
by pneumatic tyred tractors, which can tow up to three rollers in tandem.
4) Vibratory rollers
Weights placed eccentrically across a rotating shaft attached to the drum produce
vibrations in the vertical direction (Fig 8.16).
With regard to the selection of the type of roller, while all the types are suitable for
cohesive soils, the vibratory type is most effective in cohesionless soils.
Vast deposits of dry loose sands, away from human habitation, have been
effectively compacted by detonating charges of dynamite or TNT buried at depth in
boreholes which are backfilled after placing the charge (Fig 8.17). The energy
released in the explosion brings the loose soil particles to rearrange themselves into
a denser mass, which is evidenced by the depression formed on the surface, Fig 8.17.
The extent of densification and the amount of depression depend upon the quantity of
11
the charge and the state of looseness of the soil. Repeated detonations can be tried
for additional densification. A set of charges in a grid pattern can be detonated
together for faster results.
In a significant study conducted in Tamil Nadu, India, Gandhi et al. (1999) have
tried the above technique to densify flyash existing in a wet state in an ashpond where
it has been brought in a slurry form from the nearby Mettur Thermal Power Plant, close
to Salem. Apart from distributing charges at many locations, the same were placed at
different elevations at the same location. Saucer shaped depressions were noted in
all cases and the results are presented in quantitative form pointing to the
effectiveness of the technique.
Reference
Soil Mechanics for Road Engineers, DSIR – Road Research Laboratory, HMSO,
London, 1952, xxiv + 541 pp.
Pictures
Some pictures taken at the site during the work reported in the above Reference are
included here, courtesy Prof. S. R. Gandhi.
Pic. 8.1
Pic. 8.2
Pic. 8.3
Pic. 8.4
Pic. 8.5
Pic. 8.6