11

Compre s sibility and S ettlement

I*ss than 70 years ago the Foandatian Commtttee of

a well-known engineering society decidcd, at one of
its meetings, that the word "settlement" should be
avoideil in public discussions, becouse it might
disturb the peace of mind of those vho arc to be
sened by the engineering profession-

Karl Terzaghi (1939)

Many civil engineering projects include pl4cing loads onto the ground, which produce
corresponding increases in the vertical effective stress, o.'. These increases are important
because they induce vertical strains, e., in the soil, and thus causo the ground surface to
move downward. We call tliis downward movement settlement. When settlement occurs
over a large area, it is sometimes called subsidence.
In a mechanics of materials course you leamed that stresses in any material always
produce strains. Therefore, whenever o.' increases, there always will be a corresponding
settlement, 6. The issue facing a geofechnical engineer is not i/settlements will qp-cur, but
rather the magnitude of these settlements and how they compare with tolerable limits.
This chapter discusses the various factors that influence settlemdnt and presents
methods of predicting its magnitude. Chapter 12 continues these discussions and addresses
the rate of settlement, and Chapter 17 applies these methods to the design of structural
foundations. Engineers use the results ofthese settlement analyses to design structures and
other civil engineering projects. For example, if the analysis indicated the weight of a
proposed building uould cause excessive settlement in the soils below, the engineer may
decide to place the building on pile foundations that penetrate through the soft compressible
soils to deeper, harder strata.

36'6
Case Studies ' 367

Gase Studies

Some of the most dramatic examples of soil settlement are found in Mexico Ciry. Parts of
the city are underlain by one of the most troublesome soils in any urban area of the rvorid,
a very soft lacustrine clay that was deposited in the former lake Texcoco. Is engineering
properties include (Hiriart and Marsal, 1969):

Mositure content, w Average 28 I Vo, maximum 5 \OVo

Liquid limit, w. Average 289, ma,ximum 500
Plastic limit, w" Average 85, maximum 150
Void ratio, e Average 6.90 :

A comparison of these values with the typical ranges described in Tables 4.5 and .1.8
demonstrates that this is an extraordinary soil. For example, the very high void ratio
indicates it contains nearly seven times as much water as solids! Another of its important
properties is an extremely high compressibility.
As the city grew, municipal water demands increased and many wells were installed
through this clay and into deeper water-bearing sand layers. These activities resulted ia a
significant drop in the groundwater levels which, as we will discuss later in this chapter,
caused an increase in the effective stress. Because the clay is so compressible, and the stress
increase was so large, the resulting settlements became a serious problem. Between 1898
and 1966, parts of the ciry settled 6 to 7 meters (Hidaft and Marsal, 1969)! At times, rhe
rate of settlement has been as great as I mm/day. Fortunately, Mexican geotechnical
engineers, most notably Dr. Nabor Canillo, recognized the connection between groundwater
withdrawal and settlement, and convinced govemment authorities qo prohibit pumping in
the central city area.

',.':
,'r,..,;;1;y;q,'';1;: : |
..,tr,r: ,a:
:.i,] _:,.r ,. .",
,:.: !
'
..'..).1 i; . --
;:-.

-., ,;,
:-'
i':. -,.;
::
lt;.
.i.'-
r. r: ',',.,:
; i,, ' ,. ' ;
:ri:.t- :.,:: ri.:rr.
.,:: ".i..';'-:-,'
i:i.:.i:.-
_
,.j. rj:,r,.,t.rr,,;1r,,.
tl ..:.:
-:., j { :, ,::t+,
!..:;. rr :::,:: :r,' iri;,':,ri'
' I ;r::..::. -.Ji' .. ! i: I .
r,

Figure 1I.l By 1950, rhe

Palace of Fine Arts in
Mexico City had seuled
abor'! 3 m more than the
surrounding streets.

In addition to the widespread settlements due to groundwater withdrarval, local

sttlements also have occurred beneath heavy structures and monuments. Their rveight
increased the stress in the underlying soil, causing it to settle. Oae example is the Palacio
de las Bellas Artes (Palace of Fine Arts), shown in Figure I I . L It was builr betr,r'een 1904
ad 193.1, and large senlements even before it w
 368 Compressibility and Settlement . Chap. 11

palace and the immediately sunounding grounds were about 3 m lower than the adjacent
streets (Thomley et al., 1955). As a result, it has been necessary to build stairx'a1-s from the
street down to the building area.
As a result of these problems, geotechnical engineers in Mexico City have dereloped
techniques for safely supporting large structures without the detrimenal effects of ercessir-e
settlement. One of these, the 43-story Tb,wer Latino Americana is discussed in Forordation
Design: PrinciplesandPractices (coduto, 1999),thecompanionvolumeto&isbooft- This
building is across the street from the palace of Fine Arts, and has been perfuming
successfully since irs completion in the mid-1950s.
The Tower of Pisa in Italy is another example of excessive settlement. In rhis casq
one side has settled more than tlie other, a behavior we call dffirential settlement, vtbi616
gives the tower its famous tilt. Foundation Design: Principles and Practices also explores
this case study.
Settlement problems are not limited to buildings. For example, the highway bridge
shown in Figure 1 1.2 is underlain by a soft clay deposit. This soil is not able to support the
weight of the bridge, so pile foundations were installed through the clay inio hardir soils
below and the bridge was built on the piles. These foundations protect it from large
settlements.
It also was necessary to place fill adjacent to the bridge abutments so the roadway
could reach the bridge deck. These fills arq very heavy, so their weight increased o., in thl
clay, causing it to settle. when this photograph was taken, about twelve years aiter
the
bridge was built, the filt had settled about I m, as shown by the sidewalk in the ffrreground.
./

1'.1 PHYSICAL PROCESSES'.

The three most common physic.al processes that produce settlement

in soils are:

' consolidation-settle-mrent (also knorvn as prtmary consolidation setilerrcn:j.

6n
occurs rvhen a soil is'subjected to an increase in s.', as discussed
in tbe pmnicmus
sectioa, and the individual particles respond by rearr:"nging iaro a *cbfier
poctimg:
 Sec. 11.1 Physical Processes 369

This process causes a decrease in the volume of the voids, V" . If the soil is saturated
(S = l00fc), this reduction in V" can occur only if some of the pore $'ater is squeezed
out of the soil. All soils experience some consolidation when they' are subjected to
an increase in o.', and this is usually the most important source of settlement.
. Seconilary compression settlement,6,, is due to particle reorientation, creep, and
decomposition of organic materials, and does not require the expulsion of pore water.
Secondarl'compression can be significant in highly plastic clays, organic soils. and
sanitary landfills, but it is negligible in sands and gravels. Unlike consolidation
settlement, secondary compression settlement is not due to changes in o.'.
. Distortion seftlement, 6r, results from lateral movements of the soil in response io
" changes in o,'. These movements occur wheu the load is
confined to a small area,
such as a structural foundation, or near the edges of large loaded areas, such as
embankments.

The settlement at the ground surface, 6, is the sum of these three components:

6=6.*6"*6a (l l.l)

Other sources of settlement, such as that from underground mines, sinkholes, or

tunnels, also can be important, but they are beyond the scope of our discussion.

11.2 CHANGES IN VERTICAL EFFECTIVE STRESS

Most settlement is due to changes in the vertical effective ltress,. so we will begin by
examining these changes . The initial vertical effective stress, a ;, at a point in the soil is
. . the value of o.'before the event that caqses settlernent occurs. ,Thefin4l vertical ffictive
stress, ary' ,is the value after ttre event has occurred and the settlement process is complete.
Notice how settlement analyses are based on changes in effective stress, not total stress.
r.: : -The value of o5. may be computed uqing the techniques described in Chapter 10.
Usually ttre initial condition consists of geostatic stresses only, and $11s is evaluated using

. , ., . Ihe method gf cgmpu$rg of. oo.,dgpends e-n ]he kind qt evgfl!:+3! iq calsing the
stresses to increase. The most common events are: placement of a fill, placement of an
extepal load, and changes in tlre gotinA,ryatertable elevation. ' r ,l'. i,,'.',f'-
t'"""
::ltt""''
"' "

Stress Changes Due to Placement of a Fitl

When a fiIl is placed on the groundl, oi:: inthe underlying soil increases due to the rveight
of *re fill. If the length and w'idth of the fill are large compared to the depth of the point at
r*'hich we rvish to compute the stresses, and the point is beneath the central area of the fill,
then *'e compute aj by sirnply adding another layer to the l1FI of Equation 10.34.
Therefore.
370 Compressibility and Sbttlement' Chap. 11
Sc.

oO = o|n+TluHpu (11.2)
Gq
trnal
perfi
where:
tlrat
o.o' = initial vertical effective stiess
elevi
o.-l' = final vertical effective stress
char
Y;, = unit weight oi the fill now
fluu = thickneqs of the fill
prob
and
Unless stated otherwise, you may assume all of the fills described satisfy these criteria and
befo
that Equation I1.2 is valid.
rEti
If the width or length of the fill are less than about twice the depth to the point at h
which the stresses are to be computed, or if this point is near the edge of the fill, then we
q
need to evaluate the fill as an area load using the tedhniques described in chapter 10.
oq

Stress Changes Due to Placement of an External Load

Extemal loads. such as structural foundations, also produce increases in o.'. In this case,
sSfr
o"" is:
Som
./ mult
lt simrl
a'd -- aro+ (or)ua*"a (l1.3)
' i j -: : r: --l ':l:r-.a: r ,\ .. ."a ,, - : t. i 3 niet
-'!-:;
| :.-,t, ::. Si | ".. i
c(xE
form
imrn
ItrbE
hcf
oq
Pd-
 372 Compressibility and Settlement. Chap. 11
.:
Commentary

The placement of this fill rvill eventually.eause o -' at Pbint A to increase from 322lb/ft 2 to
932 lbftaz. The value of o-'at other depths in the natural soil also will increase, causing a
vertical strain - As a resuit. the tcp of the natural ground will sink from elevation 10.6 ft to
some lower elevation. Thus, the placqtnent of a 5.0 ft thick fill will ultimately produce a ground
surface that is less than 5.0 ft higher t\an the initial ground surface elevation.

1'!.3 CONSOLIDATION SETTLEMENT PHYSICAL PROCESSES

-
We use the term consolidation to describe the pressing of soil particles into a tighter packing
in response to an increase in effective stress, as shown in Figure 11.4. We assume the
volume of solids remains constant (i.e., the compression of individual partictes is
negligible); only the volume of the voids changes. The resulting settlement is known as
consolidstion settlement, 6.. This is the most important source of settlement in soils, and
its analysis is one of the cornerstones of geotechnical engineering.

g
rri

-t
,H

_L

j-:i

--f-
V, = (e',-A,e)l'"
--r--
Y

I .
Figure 11.4 Consolidation
of solid particles-under the --.

-. influence ofan increasing

verticat;ffectivestress. :
'':'" i
'i ; ,; .

' Consolidation analyses usually focus on saturated soils ("S = 100%), which means the
voids are completely fiilea *ith *ater. Both the water and tt. roiiAs are'vinually
incompressible, so consolidation c.4+ ogcur g,r.rly al some of the"r,,,ater is squeezed out of the
voids. We can demonstrate this process bi iaking a Saturated kitchen sponge and squeezing
it; the sponge compresses, but only as the water is pushed out. This retationship between
consolidation and pore rvater florv rvas qualitatively iecognized as early as 1809 when the
British engineer Thomas Telfoid placed a I 7 m deep surcharge fill over a soft clay "for the
pulpose of squeezing out the w'ater and consolidating the nrud" (Telford, 1830; Skempton,
1960). The American engineSr William SooySmith also recognized that "slow progressive
 Sec. 11.3 Consolidation Settlement
- Physical Processes 323

settlements result from the squeezing out of the water from the earth" (SooySmith, i89l-r.
The first laboratory soil consolidation tests appear to have been performed arourd
l9l0 in France by J. Frontard. He placed samples of clay in a metal container, appliec a
series of loads with a piston, and monitored the resulting settlement (Frontard, 1914i.
Although these tests provided some insight, the underlying processes were not )'ir
understood. About the same time, the German engineer Forchheimer developed a cruce
mathematical model of consolidation (Forchheimer, 1914), but it was not very accurate a:d
did not recognize important aspects of the problem.
Karl rezaghi, who was one of Forchheimer's fomrer sfudents, made tbe majoi
breakthrough. He was tffighing at a college Lq Istanbul, and began studying the scil
consolidation problem. This worh which he conducted between 1919 and 1923, produced
the fint clear recogpition of the principle of effective stress, which led the way :o
understanding the consolidation process. Terzaghi's theory of consolidatian fferzag;r",
1921, 1923a 1923b, 1924, 1925a. and 1925b) is now recognized as one of the
major milestones of geotechnical engineering. Although this theory includes several
simplifications, it has been verified arid is considered to be a good representation of the fieid
processes. We will study it ir this chapter and in Chapter 12.

Piston and Spring Analogy

To understand the physical process of consolidation and its relationship to the flow ofpcre
water, let us consider the mechanical piston and spring analogy shown in Figure I 1.5a. Ti,is
device consists of a piston and spring located inside a cylinder. The cylinder is filled wirh
water and small drain holes are present in the piston. All of this represents an element of
soil aI some depth in the ground, wift ee spring representing the soil solids, the warer
representing tbe porre water, md fte holes representing the soil voids through which the pore
water must flow. . .

' , We wiil begin with the piston in sadc equilibrium under a certain vertical load, P, as
: shown in Figure 11.5a. The assembly is submerged in a tank, so the water is subjected ro
a hydrostatic pressure that represents tbe hydrostatic pore water pressure, un, in the soil (see
Equation 7.7). ln addition, the water prcssues on the top and bottom are equal, so the
applied load on the piston is canied entirely by the spring. This loadrdivided by rhe cross-
sectional area of the cylinder represents tbe initial vertical effective stress, o;.
Then, at time =io, we uppty un additional load AP ,o tf," piri* * rtfi* in Figure
' I 1.5b. This represents the additional total vertical stress Aq in a soil, such as tnar inaiceO
, " by a'new fill. Il causes a very small downward.movementoi the pir*;,i"i fr;;""**
':
-- ;i5 resisted by both the spring and.the water.i The water is much stif'fer th.an the spring, so it
I , carries virtually all of this additional load and the water pressure increases. trls aOaliionat
'ii:-rr. pressure is
known as excess pore iater pressufe, a,. Thus, the water pressure! zn inside *re
cylinder now equals u^+ u,.
The water pressure (and the total head) inside the cylinder is now greater than that
outside, so some of the water begins to flow through the holes. These holes are very small,
so the flow rate through them also is small, but eventually a certain quantity of watei passes
through. This allows the piston to move farther down, thus compressing the spring and
 {'74

r,:r.":"S.fi j**:#di#dfm
ilT*?rniif
*ryff**:*:**iliii*"'il*fi
!h[,il":#,:+'ff
Drain loles

Il
= k|l

,r=rrh*rr"

tt=ut+k
z
u=uh

Figuretl.s&.,o1,f)unarpdnganalogy -ry,

g.tr#mfffi}""'ffi --""trt or soil ' r nts ch

processes ;n o. n.ra.
in the Fietd
Sec. 11.3 Consolidation Settlement - Physical Processes . 375

the vertical effective stress, Ao,'rises. Eventually, ail of the nerv load is carried by the
solids, the pore water pressure retums to its hydrostatic value, and the flow of pore watI
ceases. :: 1,'. .- ,: !i::._i .:. . .

This transfer of load from water to solids is one of the most impditant processes in
geotechnical engineering.

Example 11.2

The element of soil at point A in Figure 11.6 is initially subjected to the following sresses:

a-o=EYH
= (1s.7 t<rwm1(1.0 m) + (19.0 kN/m3i(2.0 rn) + (16'5 kN/m3)(a.s rn)
= l36kPa

il = \nZn
= (9.8k].r-lmr)(6.8m)
= 67 kPa

ofr = oz0- u
= l36kPa - 67 kPa
= 69kPa

;.'i.r,

Figure 11.6 Soil profile foi

Example I 1.2.

These conditions are illustrated on the left side of the plots in Figure I 1.7. Then. we place a

5.00mdeepfillrhathasaunitweightof 19.5ki\'/inr. Thisincreasesthevertical rctal stressto:

376 Compressibilifo and.settlement Chap.lt (}

aJ=a,o+.{pHp = 136kPa + 119.5kN/m3)(5.0m) =2341<pa

Notice the jumps in these curves in Figure I 1.7. Initially, the applied load is carried entirety
by the pore water, so the pore water pressure becomes:

tt = uh+ u" = uh *,t5uh611 = 6?kPa +(l9.5kN/m3)(5.0m) = 165kpa

but the vertical effective stress remains unchanged ac

o" = o'-u
= 234kPa- 165kPa
= 69kPa
ouEsrr
250 :>l [+ Construction Period II
2N
0i 150
v
b 100

50

2W
150
d
* 100

50
t= "t
o aa
I
I
I
--.: cd
.:
&n*:-: '
=T
i
,':l
..1= Ir-4 C
+l
-E
-4
* I

B
.:=
'4 l

.L .l
.l

u = ud+ u" = 67 kPa r okPa = 67kPa

 Questions and Practice Problems 377

o.= o:-!t
234Lla - 67 kPa
167 kPa

Commentary

This example illustrates the process of consolidation and how it is intimately tied to the buil+up
and decay of excess pore water pressures. It atso illustrates why this process could nor be
" properly understood until Terzaghi developed the principle ofeffective stress.

OUESTIONS AND PRACTICE PROBLEMS

I 1.1 or' at a ceftain point in a saturated clay is I 8 I kPa. This srdl is to be covered srith
The current
a2.5mthickfillthatwillhaveaunitweightof l9.3kN/m1 Whatwillbethevalue of aiat
this point immediately after the fill is placed (i.e., before any c,onsolidation has occuned)?
What will it be after the consotidation setrlement is completed?

ll.2 A4.0 m thick fill with a unit weight of 20.1 kN/n3 is to be placed on the soil profile sho*n in
Figure 10.7. Develop plots of o, ald or' vs. depth. The plot should extend from the original
ground surface to a depth of 10.0 m.

1t.3 Thegroundwatertableatacertainsitewasatadepthofl0ftbelowthegroundsurface,andthe
vertical effective stress at a point 30 ft below the ground surface was ZZnoWt#. Then a series
of wells were installed, which caused the groundwater table to drop to a depth of 25 fr below
the ground surface. Assuming &e unit weight of the soil above and below the groundwater
table are equal, compute the new o.' at &is point.

11.4 A l.O0 m3 element of soil is lmred below the groundwatet table.i'Il&en a new compressive
load was applte4 this element consolidated, producing a vertical strain, of 8.57o. The
r,
horizontal strain was zero. Comprte the volume of water squeezed out of this soil during
consolidation and express yor answer in liters.

'' ' .'..'" :

1 1.4 CONSOLTDATTON {OEDOMETER} TESis' ?'.',n''j -nr:i,

.

To predict consolidation settlement in a soil, we need to kno-w its stress-strain properties

(i.e, &e relationship between q'and e.) This normltty inuofi;;s bril'ging d'ioit sampte to
the laboratory, subjecting it to a series. of loads, and measuring the corresponding
settlements. This fesi is essentially the ;ame as'thoid bonducted by Frbntard in tito, uuI
now we have the benefit of understanding the physical processes, and thus can more
effectively interpret the results. The test is liown'ut'ur"^"iU.ri"";iil;;;;;;
oedometer test), and. is conducted in a consolidometer (or oedometier) as shown in
Figure I1.8.
We are mostly interested in the engineering properties of natural soils as they exist in
the field, so consolidation tests are usually performed on high-quality "undisturbed"
samples. It is fairly simple to obtain these samples in soft to medium clays, and the test
374 Compressibility and Settlement Chap. 11

results are reliable. However, it is virtually impossible to obtain high-quality undisturbed

samples in uncemented sands, so we use empirical correlations or in-situ tests instead of
consolidation tests to assess the str6Ss;strain properties as discussed in Section I 1.6.
It also is important for samples that were saturated in the field to remain so during
storage and testing. If the sample is allowed to dry, a process we call desiccation, negative
pore water pressures will develop and $ay cause irreversible changes in the soil.
Sometimes engineers need to evahate the consolidation characteristics of proposed
compacted fills, and do so by performiirg consolidation tests on samples that have been
remolded and compacted in the laboratory. These tests are usually less critical because well-
compacted fills generally have a low compressibility.

Soil sample

ri : .i..'

I:
t
t
I .j t{
: I
.J .

I
A.'
i

-t

i1
Sec. 'l 1.4 Consolidation (Oedometer) Tests 379

The tesr begins by applying a vertical norrnal load, P. It produces a vertical .e.$e,giire
StrESS Of:

rP
a-=--u
.A (11.+)

where:
o.' = vertical effective stress
P - applied load
A = cross-sectional area ofsoil sample
4 =pore water pressure inside soil sample

The water bath barely covers the sample, so the pore water pressure is very small compared
to the vertical stress and thus may be ignored. Thus:

(11.5)

Then the sample is allowed to consolidate. While conducting his early tests, Frontard
noted "one of the most interesting facts which have been revealed is the great length of tine
required for the escape of the excess water." During this period some of the water is being
squeezed out of the voids, and must pass through the soil to reach the porous stones.
Because we are normally testing clayey soils" the hydraulic conductivity is low and the
water flows several hours or more may be required for the sample to
slowly. Thus,
consolidate. We determine when the consolidation is complete by monitoring ttre dial gage.
The vertical strain, o upon completion of consolidation is: ,

(11.6)

'':- ' ' .

The strain is expressed using the sign convention defined in Cfup1er 10, where positive
stain indicates compression. We now have one (o.', e.) data point '
The next step is to increase-the normal load to some higher.value and allow the soil
to consolidate again, thus obtaining q secorid (o.', eJ data point. This process continues
until we have obtained curve ABC in Figure I L9. The stress-strain curve in soil is
decidedly nonlinear when shown in an arithrnetic plot. However, when presented on a
sernilogarithmic plotras shown,'the data;i3' much easier:'to interpret.' Finally, ws
incrementally unload the sample and allow it to rebound, thus producing curve CD. Cun'e
AB is known as the recompression curve,BC is the virgin cuwe, and CD is the rebound
cun'e.
3as CompressibilityandSettlement Chap. 11

Methods of Presenting Gonsolidation Test Results

Geotechnical testing laboratories use two'd#fei6rrt methods of presenting consolidation test

results: a strain plot or a void ratio plot. The test results that arrive on a geotechnical
engineer's desk could be presented iq either or both forms, so it is important to recognize
the difference, and be able to use both"methods.

t.6

1.5
0.05

1.4

0.10
1.3

e. 0.1-5
L2e

1.1
0.20

1.0
0.25
0.9

0.30
l0 100 , 1000
o.'(kPa) .

Figure I 1.9 Resuls of laboratory consolidation test. The initial

void ratio, e, is 1.6O.

r-T-
(r +eo)rz"l

.VL
I
lY' E
I eov" t-*''"'-t

Initial
e-V"

Figure 11.10 PhiSe' diagram for

derivation of Equations I l.?-:l 1.9.

The second method presents the data as a plot qf void ratio, e, vs. log q' as shown in
the horizontal and right axes in Figure 11.9. This was the method Tirzaghi used,
presumably because it emphasizes the reduction in void size that occurs during
consolidation. To compute the void ratio at various stages of the test, we need to develop
an equation that relates void ratiqwith strain. Using Figure 11.10:
 Sec. t 1.4 Consolidation {Oedometer) Tests 381

eV, = eoVr- AeV, (1 1.7)

Le V"
c,--
' 1l *eo)V"
( l r.8)

Combining Equations ll.7 and 11.8 gives:

e = eo-Le
= eo - e.(l + eo)
(11.e)
= I +eo-er(l +eo) - I
=(l-z)(l+eo)-l
where: . .l
e= void ratio
eo = initial void ratio (i.e. the void ratio at the beginning of the test)
. = vertical saain
Ae = change in void ratio during test= eo- e
V"= volume of solids

The initial void ratio, e" is usually computed from the moisture content using Equadon 4.26.

As Figure 1 1.9 illustrates, these two methods are just different ways of expressing the
sarre data. Both methods produce the same computed settler,.lents.

': "
Plastic and Elastie Deformations
. i'

Allmaterials deform when subjected to an applied load. . If all of this deformation is retained
wben the load is released, it is said to have experi enced plastic defonnation. Conversell
if the material returns to its original size and shape when the load is released, it is said io
have experienced elasrtc deforynation. For example, we could illustrate plastic deformatlon
bq-lending a copper wir.e, and elastic deformatiln by ben-ding 4'rubbei hose. The copper
willretainnearlyallofitsdeformation,whiletherubbeiwillnot;,:,',.,.'...'..
I :
Soil exhibits both plastic and elastic deformations, which is why we see two slopes
in the loading and t'nloading curves in Figure 11.9.,:, To understand this behavior, leius
consider soil element A in the profile shown in Figure l1.l l, and consolidation data for this
- clement, also shown in Figure I l.l l.

I on the plot. The effective srress is low and the void ratio is high. Then, additional
deposition occurs and the element becomes progressively buried by the newly deposited
roil. The effective stress increases and the void ratio decreases (i.e., corisolidation occurs).
At this stage, both plastic and elastic deformations are occurring. Point 2 in Figure I l.l lb
38? Compressibility and Settlemeii Chap. 11

descr$es the conditions that existed just prior to our drilling and sampling effort.
.fhen, we drill an exploratory boring, obtzun an undisturbed sample from Element A,
and take it to the laboratory. This process removes the overburden stress, so the effective
stress drops and the sa.'nple expands blightly (i.e., the void ratio increases) as described by
Point 3 in Figure I i .1 lc. This expansion reflects the elastic ponion of rhe compression that
occurred naturally in the fleld. Aithough there has been some elastic rebound, most of the
compression was plastic, so the unloading Curve 2-3 is much flatter than the loading Curve
I-2.

rl *-^
<a:tr:\ r-.
Grcund surface
-Mlir{l
iif-!*:;-.--.--=-!,'l
Laboratory
coasolidometer

o.'(log) o.'(log) :,t :.: .'bz'.(log) ..:,: ' :o1'.(log)

'(a) (") (d)
figule.|f,,11 Slitnfnlean!cbnsolidarionhiStoryfor.anelementofsoil.. ,:- :. .:,' ....,- ,;i .

: :. ::1.,' r '.,, .. :..'., ..:..:t:*-.,. . :: iti i,: ... - :... ..:..:i, y.'. i..'.
.a,
': ' Once the sample has been instalted in the laboratory consolidometer, we'once again
load itanil produce ctirver3-4-5 as shown in Figure l l.lld. ,The initial part, Curve 3-4, has
alieady been defined as the reld{iding-criive,.'It is nearly parallel to Curve l-2.and reflects
elastic'compression only. The effective stress is less than th6 maximum p?st effective stress,
so no new plastic deformation occurs-: Howbver,lwhen the curveirbAches Point 4, its slope d

suddenly changes, and Curve 4-5 reflecti.new plastic deformations, whiih occui only when
the effective stress is higher than ever before. .This is the virgin curve defined earlier.
Finally, we unload the sample in the lab and form curve 5-6, the decompression curve. It
is nearly parallel to curves 2-3 and 3-4, and reflects the elastic component'only.
Thus. scils behave one way if tlrc vertical effective stress is less than the past
nra.rimum, and another if it is increasing beyond that maximunt.
 . 1 ;Compressibility and Settlement . Chap. 11
I

The Casagrande procedure determines the pr'econsolidation stress, o.', from laboratory
data. Irnplement this method as follows; and as iilustrated in Figure I l.l3:

l. Locate the point of minimum radius on the consolidation curve (Point A).
) Draw a horizontal line from Point A.
J. Draw a line tangent to the laboratoiy at Point A.
"o*"Steps 2 and3.
4. Bisect the angle formed by the lines'from
5. Extend the straight portion of the virgin curve upward until it intersects the line
formed in Step 4. This identifies Point B, which is the preconsolidation stress, o.'.

ii t ''' . or'
tl:
.' - . . or'
Figure 11.13. Casagrande's method of finding tbe
preconsolidation Stress.
I
Sample disturbance also affects the slope of the curves, so the Schmertrnann procedure I
is an attempt to reconstruit the field consolidation curve (as illustrated in Figuie I l.l4). il
i
This procedure is performed as follows: :
;
;t

l. Determine o.' using the Casagrande procedure. .-' fr

2. Compute the initial vertical effective stress, o/, at the sample depth. This is the i
:
, . vertical effective stress prior to placement of the proposed load. ", , ]
3. Draw C horizontal line ai e 1 e. 0) from the vertical axis to o.,. This locates
"oio, =
:
', Point c.
4. Beginning at Point C, draw a line parallel to the rebound curve. Continue to the right
until reaching o"'. This foniiiiPoint D. In some cases, o.' " oo1, so this step becomes
unnecessary.
5. Extend the virgin curve downward to e = 0.42 e, thus locating Point E. If no void
ratio data is included on the consolidation plot, locate Point E at e r- 0.42, which is
 Sec. 11.4 Consolidation(OedometerlTests 385

the same as e = 0.42 eo when eo =2 and sufficiently close for other initial void ratios
(i.e., locating Point E more precisely has very little impact on the results of
Schmertrnan' s construction).
6. Draw a line connecting points D and E. This is the reconstructed virgin curve.

The final result of the Casagrande and Schnrernnan constnrctions is a bilinear function when
plotted on a semilogarithmic diagram.

Reconitructed recompression curve

Cl

0.42 es
oz1' , oi ...... :.., , ,. .: :

Fieg.le 11.14 Schmertmann's methodof 'djustingconsolidatiol,,.,,i,.,

testresults.IfvoidratiodataisavaiIable,thenlocate'.PointEata
t'.-
' of 0.42'er'ronly strain data ii aviildgri 1i1,, n9 voi{.,".,,
- ";il.utio
ratiosaregiven),thenlocatePointEate,=0.42.,Ifi;ftlv6idi;",;.rrr;'::'r:
,

.- t:. ratio alrd strairi data are give,n, use the void ratio data to tocate f i :rii f i: :

: ..'PointE,eventhough thispointmaynotconespondto e;1 0.42.'i..,,;., .:

,.

The slopes on the consolidation plot ieflect the compressibitity o.f the'soil. Steep slopes
-' mean a given increase in o.' will cause a large strain (or a large dtrange in void ratio), so
such soils are said tobe highly comprgs;iblg. pgnvgrgely, shallow sb-'pes indicate the same
increase in o.' will produce less strain, so the soil is slightly compressible. Although we
could use graphical constructions on these plots to determine the strain that corresponds to
a certain increase in effective stress, it is much easier to do so mathematically, as follows:
The slope of the virgin curve is defined as the compression ind,ex, C
":

^ =-
L' de
(l l. l0)
,'ror,;,
 386 Compressibility and Settlement Chbp. 11
There is a potential point of confusion here, because geotechnical engineers also use the
variable C" to represent the coefficient of curvature, as defined in Equation 4.36. However,
these are two entirely separate parameters: The compression index is a measure of the
compressibility, while the coefficient of curvature describes the shape of the grain-size
distribution curve. \,

The reconstructed virgin curve is ri straight line (on a semilogarithmic e vs. log o.'
plot), so we can obtain a numerical value for Q. by selecting any two points, a and b, on this
line and rewriting Equation 11.10 as:

o-b
C"= (1 1.1 l)
(log oj)o - ltog oj)"

Alternatively, if the data is plotted only in e . - o,' form (i.e., no void ratio data is given),
then the slope of the virgin curve is:

cr= (e,)u - (e,)o

l+eo (1 l. l2)
ltog ot)6 - ltog oj).

where the parameter C"l(|.+e)is called the compression ratio.

Ifthe
reconstructed virgin curve is sufficiently long, it is convenient to select Points
a and b such that log (a,')r= 10 log (o.')o. This makes the denominator of Equations I l.l 1
and 11.12 equal to 1, which simplifies the computation. This also demonstrates that C.
could be defined as the reduction in'void ratio per tenfold increase (one log-cycle) in
effective stress, as shown in Figure 11.15. Likeyise, C"l(l+e) is the strain per tenfold
increase in effective stress.
In theory, the recompression and rebound curves have nearly equal slopes, but the
rebound curve is more reliable because it is less sensitive'to sample disturbance effects.
This slope, which we call lhe recompression index, Q, is defined in the same way as c" and
can be found using Equation 11.13 wi& Points c and d on the decompression curve:
.i.r+..i;r1". a--
ji:::.: '. ' . :,. .: : :
,-r i .'r.

If the data is plotted a strain diagram, then the slope is

'recompression ratioi

C, (e.), - (e.).
(l l. l4)
7+eo ltog oj)a - ltog oj).
3A8 Compressibility and Settlernent. Chap. 11

Example ll.3
A cbnsolidation test has been performed on a sampte ofsoil obtained from point j
A in Figure
I I '6. The rest resurts a1 sho_wn in Figure I r .9. compute ,d
o.' using c*"g*"a"', method, then
adjustthetestresultsusingSchmerimann'smethod.Finally,.o,ipu."i"andC,
I
Jl
Sotution
""{
Stresses at sample depth:

From Example I1.2: .l

oL = 69kPa
J
From the Casagrande construction (Figure t L I 6): f!
h1

oj = fCO kPa - Answer

:-
 Ouestions and Practict, Problems

o- b
C"=
(togo), - gogoj),
r.43 - 0.84
log 1580lb/ft2 - log25}lblft2
= 0.74 o Answer

" - e
d
cr=
(toEa:)d- ltogoj)"
1.58 - 1.46
log 1000lbift2 - log l00lb/ft2
_ 0.12
t
= 0.12 * Answer

OUESTIONS AND PRACTICE PROBLEMS

11.5 A consolidation test is being performed on a 3.50 in diameter saturated soil sample that had an
initial height of 0.750 in and an initial moisture content of 38gVo.
a. Using G,= 2.69, compute the initial void ratio, qo.
b. At a certain stage of the test, the normal load P was 300 lb. After the consolidation at
this load was completed, the sample height was 0.690 in. Compute o j (expressed in
lb/ft'?), e., and e. ,

,
r1.6 A consolidatisn test on a sample ofclay produced the following data:

The initial void ratio was 1.21, dnd or' at ttre sample depth'was 4O kPa-
' :t it,. : ,i..
a. Plot this data on a semilogarithmic diagram similar to that in Figure I1.9.
,p.,UgingCasagrande'smgthod,4ndg.', ; ., : .....) -. irr :

c. Using Schmertmdnn's method, adjust the test.results.. .:: r. . ., .

. ,, .-.. , '.
,:

A. OetermineC,andC" ,,i-
e. The soil has a plasticity ina.i of Zf."Sased oh Kulhawy
theconsolidationtestresultsseemreasonable?' '
t t ind Mayne's
: ''t
conelarions, do

Normally Consolidated, Overconsolidated,

and Undercoilsolidated Soils

When performing consolidation analyses, we need to compare the preconsolidation stress,

o"', with the initial venical effectiye stress, o{'. The former is determined from laboratory
390 Compressibility and Settlement - Chap. 11

test data, as described earlier, while the latter is determined using Equation 10.34 with the
original field conditions (i.e., without the new load) and the original hydrostatic pore water
pressureS {i..e., Equation 7.7). Both values must be determined at the same depth, which
normally is the depth of the sample on,which the consolidation test was performed. Once
these values have been determined, we need to assess which of the following three
conditions exist in the field:
\
. If o.o' = o"', then the vertical effective stress in the field has never been higller than
its current magnitude. This condition is known as being normally consolidated (NC).
For example, this might be the case at the bottom of a lake, where sediments brought
' in by a river have slowly accumulated. In theory these two values should be exactly
equal. However, in the "real world" both are subject to error due to sample
disturbance and other factors, so the values obtained from our site characterization
program will rarely be exactly equal, even if the soil is truly normally consolidated.
Therefore, in order to avoid misclassifying the soil, we will consider it to be normally
consolidatedifo,,ando,,areequalw!t!rinabouti20pgrgent'....

. If oo' < o"', then the vertical effective stress in the field was once higher than its
current magnitude. This condition is known as being overconsolidated (OC) or
preconsolidated. There are many-processes that can'cause a soil to become
overconsolidated, including (Brumund, et al., 1976):

- Extensive erosion or excavation such that the ground surface elerdtion is now
much lower than it once was.
- Surcharge loading from a glacier, which has since melted.
Surcharge.loading from a Structure, such as a storage tank, which has since been
removed.
Increases in the pore water pressure, such as from a rising groundwater table.
Desigcal!91({ry1ng.9}"g"19_*eJ?p93ti9lplg!.'ogtlil!-{,oltp_rprggplrys,which
produces negative pore water pressures in the soil (Stark and Duncan, 1991).
Chemical changes in the soil, such as the accumulation of cementing agents.

: . If oro') %', the soil is said to be underconsolidated, which means the soil is still in
the process of consolidating under a previously applied load. We will not be dealing
1': -' "'-' ' '-
with this case.
,t
Table 1l.l
gives a classificitio.n'of soil compressibility.based on C.{l+er) for
normally consolidated soils or C ,l(l+e ) for overconsolidated soils.
 ,; ..

and Settlement '

l', !
3g2 Compressibility Chap. 11

The overconsolidation margin should be approximately constant throughout in a

stratum with common geologic origins', Therefore, we can_compute.the preconsolidation
srress ar btlter dpttis in'that stratum by using Equation 11.17 with od' at the desired depth.
Another useful parameter is the overconsolidation ratio or OCR'-

\r
OCR = :: (l1.18)
o'ro

Unlike the overconsolidation margin, the OCR varies as a function of depth, and therefore
cannot be used to compute q' at other depths in a stratum. For normally consolidated soils,
OCR = l.
TABLE 11-2 TYPICAL RAryGES OF OVERCONSOLIDATION MARGI\S
0verconsolidation Margin, o-'
Classification
(kPa) 0b/ft1
0 0 Normally consolidated

0- 100 0 - 2000 Slightly overconsolidated

100 - 400 2000 - 8000 Moderatelyoverconsolidated /

> 400 ' >80O0 '.. Heavily overconsolidated
1

1.6 COMPRESSTBTLTTY OF SANDq,4ND GRAVELS

The principles of consolidation apply to all soils, but the consolidation test described in
Seqtiol I L4 and the methods of assessing consolidation status in the field, as described in
Section 11.5, are primarily applicable io clays and silts. It is very difficult to pertbrm
reliable consolidation tests .on most sands because they are more prone to sample
disturbance, and this disturbance has a,significant effect on the test results-.. Clean sands are *+-
especially troublesome.' Gravels havd S-fi#iiti;pie aiiti,iriaira; prbbl#a; bfis ttreir targe
1

would require very large sample; and a very largq consolidometer. -'
L
!
. _, .grain-;lze
' ,* Foitunately, sands and gr4ygts,gqbjected to staticloads are much lgss compressible t
I T
,' -.
ihiir silib and clays, so it often is*3-uffrrcient to use estimated values of C ror C,in lieu o[
- laborator!' tests. For sands, these estimates can be based on the data gathered by Burmister
(1962) as ioterprcted in Table I 1.3. He pgrfoiined a series of consolidation tests on samples
reconsrituted to various relative densities: Engineers can estimate the in-situ relative density
"l(l+e )
using tlre merhods described in Chapter 4, then select an appropriate C from this
table. Note tlrat all ofrhese valus afe '1very slightly compressible" as defined in Table I l.l.
Sec. 1 1.6 Compressibility of $ands and Gravels 393

TABLE 11.3 ryPICAL CONSOLIDATION PROPERTIES OF SATUMTED NORMALLY

CONSOLIDATED SANDY SOILS AT VARIOUS RELATIVE DENSITIES (Adapted from Burmisler,
1962)

C,l (l+eo)
Soil Type

Medium to coarse sand,

some fine gravel (S$f

Medium to coa$e sand

(swsP)

Fine to coa$e sand (SW) {; o.ooe 0.002

Fine to medium sand

0.010 0.003
(swsP)

Fine sand (SP) 0.013 0.003

tiijti-:1,,r:!':l::,i
Fine sand with tracc fine
to coarse silt (SP-SM)

Find sand with linle fine

to coarse silt (SM) t:d::

Fine sand with some fine

to coarse silt (SM)

For saturated overconsolidated sands, C ,l (I+e o) is typically about one-third of the

values listed in Table I1.3, which makes such soils nearly incompressible. Compacted fills
can be considered to be overconsolidated, as can soils that have clear geologic evidence of
pieloading, such as glacial tills- Therefore, Dily settlement analyses slmply consider the
compressibiliry of such soils to be zero. If it is unclear whether a soil is normally
consolidated or overconsolidated, it is conservative to assume it is normally consolidated.
. ,,, ,,Y"y,few consoJidation tests have,beeq performed on'gravelly soils, but the
compressibillty of these sqilp ip probably lgfal to or fess than those fo1 pqd, as listed in
Tablel1.3..l:,'':,j.-.-:,.:
Anottrei ctraracidiiiti"c'of sands and gravels'is th6ir high hydraulic coirductiviry, which
'Thus.,
means any excels p'gre water dlains very-quickly. the rate ofconiolidatioo i, u"ry
faqt, ana typically occurs ngarly as fasl aS,ttr6 !g1d is applied. Thirs, if ttrc load is due to a
fill, the coniolidation of theie ioils may hCve iittie practicat signincdie:i :.' .

Hovrever, there are at least two cases where consolidation of coane-giained soils can
be very important and needs more careful consideration:

l. Loose sandy soils subjected to dynamic loads, such as those from an earthquake.
They can rxperience very large and irregular settlements that can cause serious
damage. Kramer ( 1996) discusses methods of evaluating this problem
 A94 Compressibility and Settlement Chap. 11

2. Sandy or gravelty soils that support shallow foundations. Structural foundations

are often very sensitive to settlenient, so we often conduct more precise assessments
of compressibility. These are usualiy done using in-situ tests, such as t}re SPT or
CPT, and often expressed in terrhs of a modulus of elasticity, E, instead of C.or C.,
Special analyses based on in-situ test results are available to predict such settlements,
as discussed in Chapter 17.
:

ESTIONS AND PRACTICE PROBLEMS

ll,7 A consolidation test has been performed on a soil sample obtained from Point B in
Figure I1.19. The measured preconsolidation stress was 88 kPa. Determine whether the soil
is normally consolidated or overconsolidated, then compute the overconsolidation margin and
overconsolidation ratio at Point B.

Note: These computations are based on the initial conditions, and thus should not include the
weight of the proposed fill.

11.8 A saturared, normally cons<ilidated, l00o-year-old fine-to-mediunr sand has an SPT Nr= 12
at a depth where the vertical effective stress is about 1000 lb/ft:and D ,;
0.5 mm. Using the
rechniques described in Chapters 3 and 4, determine the relative density of this soil, then
estimate C"l (l+eo) based on Table I l-3.

1 1.9 A consolidation test has beetr performed on a sample obtained from Point A in ngirr" f f .f . m"
measured preconsolidation stress was l5OO lblft1
a. Determine if the soil is normally consolidated or overconsotidated
" b. Compute the overconsolidation margin anp the overconsolidation ratio
'
, . , ,,-.'r "i,,.co*ePJ99i.atPoilB :.i; .,, , :, ,.

1.7 CONSOLIDATION SETTLEMENT PREDICTIONS

' I fne purpgse of performing consolidation tess is io define the stress-strain properties of the
'i: 'soil hnd thriS attow us to predici consolidationseitlements'in the field.: We perform this
'- 'rcompriAtion by projecting the'hbdiatory test'results (aS contained in the parameters C"; G,
' eo, arnd o.') back to the field conditions. For simplicity, the discussions of consolidation
t -*+-
gi
ai
' setttemdnt 'predictioirs''in'thir' inapiei donsiOdr'only'the'case :of one-dimeosional titi
' ',!con3o!ida{9n, and *et*ill Ue computing o3l1 the ultimate'cdnsolidation seitler-ngnt.
' '::t I tonej'di*ensional
coryolidoiton i"*ri iinff v6rtiial'strains b6cur"iri thi ioil (i.e.;
fl
t'r :
,= et= 0). We can reasonably assume this is the case when at least one of the following
'' conditions exist (Fox, 1995):

1; The.width of the loaded area is at least four times the thickness of the compressible
strata.
Z. me depth to thb top of the compressible stratd is at least twice the width of the loaded
area.
or 3. The compressible strata lie between stiffer soil strata whose Presence tends to reduce
the magnitude of horizontal strains-
 Sec. 11.7 Consolidation Settlement Predictions 395

In this context, "compressible strata" refers to strata that have a C"or C,large enough to
contribute signific.antly to the settlement.
The most coflrmon one-dimensional consolidation problems are those that evaluate
settlement due to the placement of a long and wide fill or due to the widespread lowering
of the groundwater table. Maay other problems, such as foundations, also may be idealized
as being one-dimensional.
Tlne ultinnte consolidation settlement, (6.L, is the value of 6. after all of the excess
pore water pressures have dissipate4 which may require many years or eYen decades.
Chapter 12 explores this topic in more detail" ad presents me&ods of developing time-
settlement cun'es.

fIonnally Gonsolidated Soils lo,; = s '1

lf on' = o.', &e soil is, by definition, normally consolidated. Thus, the initial and finat
codititm ue as shown in Figure 11.17, and the compressibility is defined by Co the slope
of the virgin curve.

-r-

I
Le

." i;

Figure 11.17 Consolidation of normaily

consolidated soils.

Rewriting Equation ll.l0 gives: :r i:,.r1 a -: r;l - .. -

de = -C,dlogat, (1 1. l9)

Ae -- -C, (11.20)
'"[*J
Sec. 11,7 Cons$lidationSettlementPredictions 397

(6J.r, = t, i^"l"-[*) (rt.?4)

o$ ad a"
ar'0qgcr-)

::
.:
'"
I

::

. or'(log scale)
Figure 11.18 Consolidation of overconsolidated soils.
398 Compressibility and Settlement Chap. 11

Overconsolidated Soils - Gase ll la,i < o! < a.il

Ifthe consolidation process begins on the recompression curve and ends on the.virgin curve,
as shown in Figure I I . 18, then the analysis must consider both C" and C,:

(6"),u = .&"*[9 (l l.2s)

'l*"';[*)
This condition is quite common, because many soils that might appear to be normally
consolidated from a geologic analysis actually have a small amount of overconsolidation
(Mesri, Lo, and Feng, 1994).
When using Equation 11.25, o.o', o.', and of must be computed at the midpoint of
each layer. This means o.' will need to be computed using Equation 11.17..

Ultimate Consolidation SettlerRent Analysis Procedure

Use the'following procedure to compute (&)r,:

l. Beginning at the original ground surface, divide the soil profile into stratgs'kre each
stratum consists,of a single soil type with commion geologic origins. 'For example,
one stratum maf consist of a dense sand, while another might be a soft-to-medium
clay. Continue downward with this process until you have passed through all of the
compressible strata (i.e., until you reach bedrock or some very hard soil). For each
stratum, identify the unit weight, y. Note: Boring logs usually report the dry unit.
weight, 1r, and moisture content, p', but we can compute y from this data using
, Equation 4.27. Also define the location of ihe grotndwaterrable.
2. nacn clay or iiltltlatu* must have results from ai least one consolidation test (or at
least estimates of these results). Using tire techniques described in Section 11.4,
defermine if each srratum is normally consolidated or overconsolidated, then assign '
values for C.l(l+eo)and/or C,/(l+"0). Fo. each overconsolidated stratum, compute
o-'using Equation ll.17 and assume it is constant throughout that stratum. For , .'

normally consolidated soils, set o O.

^' =
3. For each sand or gravel stratum, uiiign u value for C, I (l+eo\ ot C, I (l+e) based on
the infonnation in Section I1.5.
4.' For any very hard stratum, .such as bedrock or glacial till, tliat is I'irtually "
5. Working downward from the original ground surface (i.e., do not consider any
proposed fills), divide the soil profile into horizontal layers. Begin.a new layer'
l

whenever a new stratum is encountered, and divide any thick strata into muldple
layers. When pertorining computarions by hand, each strata should have layers no
more than 2 to 5 m (5 to 15 fr) thick. Thinner layers are especially appropriate near
the ground surface, because the strain is generally larger there. Computer-based
Sec. 11.7 ConsolidatisnSeEl.ementPredictions 399

computaiions can use much thinner layers throughout the entire depth, and achieve
slightly r,iore precise results.
6. Tabulate the following parameters at the midpoint of each layer: :r : ,,: :

For normally consolidated strata:

a ro'
a-rr'
C,l(l+eo)
H

For overconsolidated stata:

' Qfl'
aj
. 6"' = oro' + a^'.
C,l(l+eo)
C,l(l+eo)
H

It is not necessary to record these parameters in incompressible strata.

Normally we compute ozo' and o.t'using *re hydrostatic pore water pressures
(Equation 7.7) with no significant seepage force,:and this is the only case we will
consider in this book. However, if preexisting excess pore water Pressures or
significant seepage forces are present" they should be evaluated. Sornetimes this may
require the installation of piezometers to obtain accurate informalion on the in-siru
pore water pressures.
7. UsingEquation ll.23,ll.24,or ll.25,computetheconsolidationsettlementforeach
layer, then sum to find (6")r. Note that each layer will not necessarily use the same
equation. If o.' is only sligtrtty greater than o, | (perhaps less than 20 percent greater),
then it may not be clear if the soil is truly overconsolidated, or if the difference is only
an apparent overconsolidation due to uncertainties in assessing these two values. In
such cases, it is acceptable to use eitherEquation 11.23 (norrnally consolidated) or
I 1.25 (overtonsolidated case II).
.i:. . .: . ; ,'.t :.,,.
'.'::" '.
',,..!,,,. ... ..
' '
,r.,''1,r,,i - -,.i:, ,..:,,,i;
. :'i.: ' ,

,'" ',' . . :: :' ': i::''tr:ir'i;l:1 t,'?'i:l ,11.."

A 3.0 m deep compacted fill is to be placed over the sgil profrle. shown in Figure 11.19. A
consolidation test-on a sample from point A prodo:..,1$.."-,lolfgwine rlsutts: I , ,
1ir* ',
,

C"=A.40 : f l'' .,i,:,.-11

,

C, = 0.08 :

eo=l)A 't. ,' . i'- t' :, i,;.-

o.' = 70'0 kPa

This sample.is representative-of the entire soft clay stratum. Compute the ultimate consolidatioa
$Iement due to the weight of this fill.
 {
I

*..JI
Sec. 11.7 ConsolidationSettlementPredictions . 4O1

At the sample a', = oto clay is normally consolidated

If the soil at the sample depth is normally consolidated, and the sample is truly
representative, then the entire stratum is normally consolidated.
Assume the sand also is normally consolidated, srhich is conservative. For the sand
strata, use C.l(l+e) = 0.008, per Table I1.3.

At midpoint of layer
H t, c" Eqn.
(6.)r,
l-a.ver
(m) or'(kPa) ou GPa)
I +co (mrn)
Eqn 10.35 .

..-.r

3 ' 3.0 56.4 I14.0

..: , 0.19 lt23 174

4 3.0 75.0 ., 132.6 r' , 0.!9 1t.23 141

5 ,'.4.o 96.7 154.3 0.19 11.23 154

(6J,,= 483

Round off to:

(4Lo = 480 mm - Answer

Notice how we have used rhe same analysis for soils above and below the groundwater table,
and both are based on sanrared C. / (l+eu) values This is conservative (although in this case,
very slightty so) because tbe soils abore the groundwarer rable are pobably less compressible.
Section ll.9 discusses unsaturated soils in more detail.

ExamPle 11.6 ::ij,'i '

' :.
-', : 1, ::::1:: ': i-'l': -':'-:: l''-. I '1, . i
: An 8.5 rn qiep compacted fill is to be placed o'ier'the soil. profile shown in Figure 11.20.
Consolidation tests on samples from points A and B pro&rced the following results:
: :" :r i.,.:iliif;tJ;i, i
". :,ri:,:,: : ii;r.. r: ! ;:; !!, _,t,:::.1 i,i.::!:: j :,: :

Sample A Sample B

C, 0.25 , ' 0-20

c, 0.08 0.06

eo . 0.66 'o.45
a"' l0l kPa 5 I0 kPa

Compute the ultimate consolidation settlement due to the weight of this fill.
4o,2 Compressibility and Settlement Chap. 11

ffgsre f f2O Soil profile fa E:<aqile 11.6.

Solution
-;
UsingEquation 11.2:

.t
I
t

t
I
T

t
rf

o'rr = oL+ l72.6pa -

= 55.0kPa + l72.6kPa
- = 227.6kPa

a-o' < o,' < oo' .'. overconsolidated case II

 -. 403
Sec. 11.7 ConsolidationSettlementPredictions

a'. = - o'r = lol.-55 = 46kPa

"1"

Therefore, o"' at any depth in the stiff silty clay straruro is equal to o,' + 46 kPa.

At sample B:

a'*=ly4-u
= (18.3 kN/m3) (2.0 m) + i19.0 kl{/n3) {?.0 m)
+ (19.5 kN/m3) (10.0 m) - (9.8 kNeE) (17.0 m)
= 198.0 kPa

alu = o'ro*yra,
= 198.0 kPa + 172.6 kPa
= 370.6kPa

aro' 1o"' and oo' < a"' .". overconsolidated case I

At midpoint of layer
I.a;yer H C, c" Eqn- (SJ*
(m) or'(kPa) o"f(kPa) : or" (tfa) l+eo l+eo (mm)
Eqn. 10.34 Eqn. 11.17 , '.

4 4.0 120.4
:
r .5 4.0 159.2 5l
6 " 5.0 zo2.g 53

,7 45

. . i ; ..t i:;. .:.j,!=-.i:. :. ._.:

-i:,:,'
^
.
='830Fm'" ; :'AnsYf :'::: : : '
'(6')" ,
"':-'

Notice how most of the compression @curs in the upper $ratum, which is overconsolidaFd
case tr (i.e-, som of the compression'occurs along the viryl'n curve). The lower str-atum, which
is overconsolidated case II, has much less compression even *rough it is twice as 6ick because
it is overconsolidated case I and all of the compression occun on the recohpression cuwe.
 :i
;i
4o,4 Compressibility and Settlement Chap. 11

Example 11.7

The groundwater table in the soil profile shown in Figure I l.2l is currently atlhe elevation
labeled "initial." A proposed dewatering project will cause it to drop to the elevation labeled
"final." Compute the resulting ultimate consolidation settlement.
A consolidation test performgd on a sample from point A produced the following results:
\,
C,l (l + e) = O'14
C,l (1 + er) = 0.06
o"'= 30001b/ft2 '

Figure ll.2l Soil profile foi Eiample 11.7.

I

Solution
I l. . i
At sample A :
j
i
{
:. :j.. :

tn
a'ro=lyH-u I
:

(l2r lb/ft3) (7.0 ft) + (r25 lb/ft3) (6.0 f0 + 027 tbtftr(l8.0 ft) i i
3) (5.0 I
+(I l0 lb/ft fr) .- (62.4 lb/ft ) (29.0 ft)
= 2623lbtft2 -t

a!r' = o,' . ..' normallY consolidated

 Sec. 11.7 ConsolidationSettlementPredictions 405

At midpoint of layer c" (6i*

Layer ,1, l*do
Eqn
.{in}
'0b/o '0b/fr'?)

tI.23
r:::.
r: i::-

7210 0.01I ,1123

I 2126 0.009 11.23

|-
i,
0.009 11.23 0.2
.-!'-,. j . l.ara/:*:ii1:,i:r9,1::j::.::i5i I : :: . i a -r

: o.t+, , . lt.il :, .2.1 ,

:r' 1:',iri: l.=i.rr?itrr,:i::'::,':-':. : :i.i :'l

:
0.14'.r1tr.,;i!.23,: ;:,
;i.:...2-1 ::1....
: r r! i aj:;:::i- ::li:i.ri,r; :-

(6Jr, = 4-6

(6.)4 = 4.6 in - Answer

Notice how most of the compression occurs in the normally consolidated clay, even though it
remains below the groundwater table. The cause of senlement is an increase in effective stress,
not drying, and the clay is most susceptible to this increase because has the highest it
C,l(l + e).

Example 11.8

After the settlement due to the fill described in Example 11 .5 is completed, a 20 m diameter,
l0 m tall cylindrical steel water tank is to be built. The bonom of the iank will be at the top of
the fill, and it will have an empty mass of 300,000 kg. Ultimatety, the water inside will be 9.5
. m deep. Compute the ultimate consolidation settlement beneath ttle center of this tank due ro
the weight of the tank and its contents. ,Assume the new. fill is overcorasolidated wirh
': C,l(l+e) = 0.002.
.' :"
Strategi

So\tion r'i

Compute weights:

Vr* = Mg

= (300,000ke)(e.8 n/. r)
.\f l+l
loooN/
= 2900kN
dtos Compressibility and Settlement Chap. 11

Wnor", = V,onl\*
n B?H
= ttu
4
:r (20.0 rn)2 (9.5 m) (9.g kN/m3)
- -^, 'i4 t,

= 29,200 kli

The weight of the water is much gieater than thai of the empty tank, so it is reasonable
for us
to arrunra the bearing pressure q is constant across the bottom of the tank.

q= tv
A
2900 kN i 29;200 kN
n(2o.am)2 t4
102 kPa

At midpoint of iayer

Qa' (a,).a*,a ar! Cc c, (6J,,

H (kPa) (kPa) | 'ro fun.
Layer (kPa) zl I +eo (mm)
{m) Eqn (m) Eqn. Eqn.
10.34 10-25 I i.-1
1.5 lot-? t3t).5 0.002 tt-24 7'4

r 14.0 8.0 7'1:r 191.1 '' 0.19' , 11.23 r28

::; :. i- ';:: ; , .: i: i

s,3.0 132.6 1t.0 oo.r i.

193.3 0,,!9 , .:.1:,.:ll:23 93

6 r4.0 154.3 14.5 45.1 199.4 0.19 r 1.23 85

(o.)u, = 319

, , i,.. . : .;:;,;.':::-r,; 1i i11"-;ili '),i,:";:.; il";.::':'.r:. :l::'i

,. i;.,. . i, ,
.:'1:.:,.e:f.:.1a.!i,].ii*::glj ;':1.;:.t:::
i''ri r-ri;]ti."; !l
:: ::.4.,-:;l l:..
. .,* ti,r.
i
(6J..,. = 320 mm
.

'c'{u , -,.;-.'.i fiiiw.er

.iii .t i :'i i

,,,,-:,..,,,,.:.;.:.t"iir.r,iiir'lii:r'i'?l;!'lr}!ii';;.i,:..:,,::;11:;-)'.1:t.::::

Comrnentary r
:

l - If thertank were buitt immediately after the fill was placed, then o.r'would be the same
as in Exampte 11.5, and er-erything else would re.main unchanged. Such a solution
'of stresses.
s oul,J illustrate the use of superpositiotr
' , 2. The values of (O");"a,i..*beneath the edge of the tank are less than those beneath the centel
(see Figuri tO. fbi.
'iius, the consoiiclation se[lement atso will be less and the bottom
oi the tank will settle into a dish shape. The difference between these two settlements
is called tliJferential settletnent. We wiil discuss drfferentiai rcttlemcnts in more detail
in ChaPter l7'
Sec. 11.7 ConsolidationSettlementPredictions 407

Program FILLSETT

'Progrim FILLSETT, which is in geotechnical analysis software package that accompanies

this book, computes the ultimate consolidation settlement due to the weight of fitls. It uses
the analysis method described in this chapter, except that it divides each soil srratum inr+
50 layers, thus obtaining a slightly more precise solution.
To use this program, you must first download the geotechnical analysis softwa:e
package from the Prentice Hatl web site and install it onto a computer. See Appendlr C ior
computer system requirements, downloading information, and installation instruction-s.
Then select FILLSETTfTom the main menu.
. Once the FILLSETT screen appears, select the units of measurement and enter the
requested data. If fewer than five soil strata are present, then enter only the dara for those
strata and Ieave the lower fields empty.
Finally, click on the CALC[iI"ATE button to perform the compu[ation. The results are
shown on the screen, and printouts may be obtained by clicking on the pR${T button.

Example 11.9

Compute the ultimate consolidation settlement due to the weight of the fill sho*n ip
Figure I1.22.

-r-
I
12 ft
I

{
l-
8ft
l--
i C../(l+ee)=0.00i

I
nft
I

-L
I
I
tr3 ft

l
I

Solution
a
Using pro_eram FILLSETT see screen capture in Fi*eure I 1.23
-
Final result: (6"),,, = 1.f73 ft (round off to 1.1 ft) * Answer
 404 Compressibility and Settlement Chap. 11

'lt ,':.!J

Figure 11.23 FILLSETT analysis for Example 11.9.

OUESTIONS AND PRACTICE PROBLEMS

11.10 A 5.0 ft thick frrll is to be placed on the soil profile shown in Figure I 1.3. A consolidation test
performed on a sample obtained from Point B produced the foltowing results: C.= 0.27,
C, = 0.10, eo = 1.09, o"' =760lb/ff . Compute the ultimate consolidation settlement due to the
weight of this fill and determine the ground surface elevatirn after the consolidation is l
. complete. Check your answers using program FILLSETT. ' ' t: i,
I
j

Note: The first layer in your analysis should extend from the original ground surface to the

.
groundwater table.
f
11.11 A 4.0 m thick fill is to be made of d soil with a Proctor maximum dry unit weight of 19.4 kN/m]
' and an optimum moisture content of I3.09o. This fill will be compacted at optimum moisture
content to an average relative compaction of 927o. The underlying soils are as shown in Figure
I 1.24. Consolidation tests were performed at Points A and B, with the fo"llowing results:

Sample C, C, ,1 o.'(kPa)
A 0.59 0.r9 l.90 113 I

B 0.37 ;0.14 1.21 100

The silt_l sand is normrlll consolidated. Using hand computations, determine the ultimate
consolidation settlement due to rhe rvei-eht of this llll. Then, {rheck your ansrver using program
FII,LSETT.
 410 Compressibility and Settlemeni Chap. 11

process is called dessication and causes these soils to become overcorsolidated.

Thus,

prohles that contain crusts have les.s settlement than identical profiles without crusts.
The presence of crusts has a significant impact on settlement computations' even
if
than the underlying compressible soils. In addition, variations in the
they are much thinner
crust thickness across a site can be a significant source of differential settlement.
Thus' site
studies need to carefrilly evaluate the thickness and compressibility of
characterization
crusts.
eI.2.0 m

eI.0.0 m

el.-0.8 m

el.-5.0 m
:

Figure 11.25 Ty'pical crust near the ground surface in an otherwise normally
consolidated claY.

11.9 SETTLEMENT OF UNSATURATED SOILS -

Thus far rve have treated unsaturated soils using the techniques developed for saturated
soils, except rve have set il = 0 (e.g., Example 11.5). However, some unsaturated soils
are
, , ;;;;;;" oir,.. t ina, of settlemeni problems, especially if they belomc wetted somedme
after construction.
One of these problems occurs in certain kinds of clay'that are known.as- expansfue f
soi/s. These clays expand when they becomq wetted, and contract ryhgn dried.. Another
kind of problemati. ,oit it called, a'iollapsible soil, which comprelses whgn it is
wetted'
Both types are discussed in Chapter 18.
Although the soft saturatei soils generally have the worst probleqs.with
settlement,
expansive and collapsible soils also can be problematic, especially
in arid and semi-arid
climates. '

1 1.1O SECOSJDARY COMPRESSIOilI SETTLEMENT

Settlement ceases'
Once the excess pore water pressures hal'e dissipated, consolidation
This additional settlement is due to
However, sone soiis continue to settle anywaY.
Sec. 11.10 Secondary Compression Settlement . 411

secondary compression and occurs under a constant effective stress. We don't fui-y
parti'-ie
understand the physical basis for secondary compression, but it appears to be due to
rearrangement, creep, and'the decomposition of organics. Highly plastic clays, orgadc
soils, and sanitary landfills are most likely to have significant secondary compressiin.
However, secondary compression is negligible in sands and gravels'
The secondary compression index, Cn,defines the rate of secondary compression. It
can be defined either in terrns of either void ratio or strain:

,-c
^de= -- (1 l.:5) .

dbgt

Cn de,
= (r1.3i)
| * t, dlogt

where:
Co = secondary compression index
e= void ratio
p= yoidratio at end of consolidation settlement (can us , = q without
inroducing much error)
e. = vertical strain
/ = time

Design values are normally determined while conducting a labontory consolidation test.
The consolidation settlement occurs very rapidly in the lab (because of the short draine
distance), so it is not difficult to maintain one or more of the load increments beyond the
completion of consolidation settlement. The change in void ratio after this point can be
:
plotted against log time to determine Cn
Another way of developing design values of Co is to rely on empirical data that relates
it to the compression index, C. This data if in Table 1 1.4.
,llqryariryd
''

TABLE 11.4 EMPIRICAL'CORRELATION BETWEEN C"-

AND C" ffenaghi, Peck, and Mesri. 19961
Materiii", c,/c,
Granular soils, including rocHill 0.02 t 0.01
Shale and mudstone 0.03 + 0.01
a
Inorganic Clays and siks 0.Mt0.01
Organic clays and silts 0.05 * 0.01

Peat and muskeg 0.06 t 0.01

412 ompressibility and Settlement Chap' 11

The settlement due to sebondary conrpression is:

' c / ,\
53 = " Hloel-:-l (11.28)
l+p "ltr]

where:
6, = secondary compression settlement
-F1= thickness of compressible strata
/ = time after application of load
lp= rime required to complete consolidation settlement (in theory this is infinite,
but for practical problems we can assume it occurs when 95 percent of the
consolidation in the field is complete.

We assume the secondary compression settlement begins at time lo'

Usually secondary compression settlement is much smaller than consolidation
settlement, and thus is not a major consideration. However, in some situations, it can be
very important. For example, the consolidation settlement in sanitary landfills is typically
complete within a few years, while the- secondary compression settlement continues for
many decades. Secondary compression settlements on the order of I percent of the refuse
thickness per year have been measured in a lO-year-old landfrll (Coduto and Hliric, 190).
Significant structures are rarely built on soils that have the potentialfor significant
secondary compression. Hou'ever, highways and other transportation facilities are
sometimes built on such soils.

Example 11.10 :

.":
The soft clay described in Example 11.5 has C 0.018. Assuming the consoliduion
"l(l+e o\=
setrlement will be 95 pereent compiete 40 years after thq fill is placed, compute the secondary
compression settlement that will oucur over the next 30 years. .

,i Solution

"d
d'lrt
{

This is approximatell' one-tenth of the consolidation settlement of 480 mm, as compuied in

Example I 1.5. t
A
coinplete.exarirple including both consolidation and secondary compression
settlements is included in Chapter 12.
 Sec. 11.1'l Distortion Settlement 413

11.11 DISTORTION SETTLEMENT

When heavy loads ar3 a-pplied over a small are4 the soil can deform laterally, as shown in
Figure I 1.26. Similar lateral deformations also can occur near the perimeter of larger loaded
areas. These deformations produce additional settlement at the ground surface, rvirich *'e
call dis tortio n s ettl ement
Distortion settlement is generally much smaller than consolidation settlement, and can
usually be ignored. However, it is sometimes considered in the design of spread footiag
foundations, as discussed in Fottndation Design: Principles and Practices (Coduto, 1999).

Figure 11.26 Distortion settlement

beneath a small loaded area-

77.12 HEAVE DUE TO UNLOADING

Our discussions thus far have considered only settlement of soils in response to an if,creased
load. Another possibility is heave (negative settlement) due to a decreased load, such as an
excavation. In this case, o j <s;. The soil will heave according to the rebound curve in
Figure 11.9, which has a slope of C,, so we compute the heave using Equation ll-24.
Because the soil is expanding, the excess pore water pressure is negative, causing pore water
to be drawn into the vo1{s." Thjs is,!!e opposite of the process described in Figure 11.7.
., Therd are oiherprocesses that also can cause heave in soits- tle'moirhotable'one is
the swelling of expansive clays, which is discussed in Ch4pter 18.
't-

11.13 ACCURACY OF SETTLEMENT PREDICTIONS

As with all other geotechnical analyses, settlement predictions are subject to many errors.
'.Theseinclude: : ',; .-
..:
. Differencqs- between profile-used in the analysis and the real soil profile,
.' $e s.o_il : ':
especially the proper identification of crusts
Differences between the engineering properties of the soil samples and the average
properties of the strata they represent (i.e., are they truly representative?)
Sample disturbance
 414 Conrpressibility and Settlement Chap. 11

Errors introduced due to testing techniques in the laboratory

Errors in assessing o"'
a The assumption that consolidation in the field is one-dimensional (i.e., there is no
horizontal strain)
Differences between Terzaghi's theory of consolidation and the real behavior of soils
Su
in the field

The compression index, Q, and the recompression index, C,, vary widely within soil
deposits, even those that appear to be fairly uniform. Ktrlhawy, Roth, and Grigoriu (1991)
reported coefficients of variation in C" values of 26 to 52 percent. This means C" values
from a single randomly obtained soil sample would have only a 30 to 56 percent probabiliry
of being within 20 percent of the true C" of the stratum. This uncertainfy can be
significantly reduced by testing more than one sample from each stratum, but it still
represents an important source of error in our analyses.
Fortunately, settiement analyses consist of a summation for multiple strata, which
introduces an averaging effect on test uncertainties. Even so, the error in consolidation
settlement predictions is typically on the order of +25 to 50 percent, even rvith careful
sampling and testing. Analyses of secondary compression settlement are even less accurate,
having errors on the order of about +75 percent (Fox, 1995). We need to consider these
potential errors when setting allowable qettlement values, and incorporate an appropriate
factor of safety in these allowable values. These margins of error also underscore the
usefulness of monitoring the actual settlements in the field, comparing them to the predicted
settlements, and, if necessary, modifying the designs accordingly.

OUESTIONS AND PRACTICE PROBLEMS

11.15 Using t}re data from Example 11.10, develop a plot of secondary compression settlement vs.
time for the period 40 to 100 years after completion of the fill. Is the rate of secondary
compression settlement increasing or decreasing with time?
: ,.. :,

11.16 Point C in Figure I l.l9 was originally at elevation 12.00 m, but it dropped io elevation I 1.52 m
as a result of the consolidation settlement described in Example 11.5. Now that the
consolidation is complete, the fill is tc be removed. Compute the new elevation of Point C after
the natural soils rebound in response to the fill removal. Ignore any secondary compression
settlement.
. ,,' ':-

11.17 A cross-section through a tidal mud flat area is shown in Figure 11.25. This site is adjacent to
a bay, is subject to varying water levels according to tides, and is occasionally submerged u'hen
heavy runoff from nearby rivers raises the elevation of the water in the bay. For analysis
purposes, use a ground"vaGr table at the ground surface, as shown. A cnrst has formed in the
upper 0.8 m of soil due to dessication (drying) and is stiffer than the underlying soil. This crust
is overconsolidated case I, and the soils below are normally consolidated. The proposed fill is
required to protect the site.from future flooding, and thus permit construction of a cbmmercial
development. Using progranl FILLSETT, determine the ultimate consolidation settlement due
to the weight of this proposed fill. Then, consider the possibility that the crust was not
recognized in the site characterization program, and perform another FTLLSETT analysis using
 Summary 4113

C"(l+eo)=0.20andy=14.0kN/nf fortheendre5.0mofclay. Comparetheresultsofthe'e

two analyses and comment on the importance of recognizing the presence of crusts.

SUMMARY

Maior Points
l. Settlement can be caused by several different physical processes. We have considered
three: consolidation settlement, 6",.secondary compression settlement, 6", and
distortion settlement, 6d.
2. Consolidation is usually the most important type of settlement. It occurs when the
vertical effective stress increases from an initial value of o -n' to a final value of a ,i .
This change causes the solid particles to move into a tighter packing, which results in
a vertical strain and a corresponding settlement.
3. If the soil is saturated, which is the case in most consolidation analyses, the applied
load is first carried by the pore water. This causes a temPorary increase in the pore
water pressure. This increase is called an excess pore water pressure, uc The
presence ofthis pressure induces a hydraulic gradient in the soil, forcing some pore
water to flow out of the voids, thus relieving the excess pore water pressures. After
some period, which may be years or decades, rq-0, the applied load is transferred to
the solid particles, and or' - or.'.
4. We measure the stress-strain properties of a soil by conducting consolidation tests in
the laboratory on undisturbed samples. The test results are expressed in the following
parameters:

o. = preconsolidation stress
eo = initial void ratio
C. = compression index
. C,=recompression index ,, , ,..- ;. ,.

'' i '
'
The preconsolidation stress, !q th9 grgatesl.vertical effective stress th9 soil has ever
experienCed at the point where the sample was obtained. The parameters.C ,and C"
define the slope of the consolidation curve at stresses less than and greater thar o"',
respectively.
5. Using the parzrmeters from the consolidation test, the changes in effective stress in the
field, and other data, ]ve gaq"compule.thg consolidation settlement.
6. The greatest consolidation settlements occur in soft clays. Sandy and gravelly soils
are usually much less compressible. In addition, it is nearly impossible to obtain
sufficiently undisturbed samples to conduct reliable consolidation tests on sands and
gravels, so the compressiblity of these soils is determined by empirical cgrrelations
or by in-situ tests.
7. Normfily consolidated soils are those that have never experrc'cd ? vertical effective
stress significantly greater than the present value of o.'. Conversely, overconsolidated
soils are those that have experienced higher stresses.
 416 Compressibility and Settlement . Chap' 'l 1

g. Normally consolidated soils iiften have an overconsoljdated crust near the ground
surface. It is important to recogllize the presence of these crusts in the site
characterization program. ' r" r'

9. Secondary compression settlement is the result of particle rearrangement, creep'

decompoiition of organic materials, and other processes. It produces settlement bven
though o.' remains constant, and settlement continues at an ever-decreasing rate
\is
with time.
10. Distortion seftlement is due to the horizontal movements of soil, and occurs primarily
when the loaded area is small and the bearing pressure is high, such as structural
foundations.
11. If the soil is unloaded, o.' decreases and oU < o.', so the soil heaves instead of
settles.
I 2. Settlement predictions are subject to several sources of error. Even careful
predictions
settlement typically have a precision on the order of +25 to 50
of consolidation
perceot. Predictions of secondary compression settlement typically are even less

accurate, with errors on the order oft75 percent'

Vocabulary

collapsible soil expansive soil recompressron curve

compressibility heave recompression index
compression index normally consolidated recompression ratio
compression ratio oedometer secondary compiession
consolidation one-dimensional index
consolidation settlement consolidation secondary compression
consolidation test overconsolidated settlement
consolidometer overconsolidation margin settlement
crust overconsolidation ratio subsidence
'theory of consolidation 1

desiccation plastic deformation

differential settlement porous stones ultimate consolidation
distortion settlement preconso!idated settlement
. elastic deformation
excess pore water Pressure
preconsolidation
rebound curve
stress vrgln curve
.:, a:: i:
:
I
r' j . ,. r j : 1
j' ': ; i:l -
:1;i, ,ti', r,,', r.U.t'
''.;
COMPiEHENSIVE OUESTIONS AND PRACTICE PROBLEMS
t.
":

soils, and give

11.18 Explain the difference betw".n no.*Jly consolidated soil and overconsolidated
examples of geologic conditions that would form each type'
.j
why? which are
11.19 What types of natural soils are best suited for testing in a consolidometer?
not well suited? WhY?

results from a conso-lidation test, the preconsolidation stress for

a certain soil
'^'"t According to the
11,20
stres; the sample location is 797 lb/fi' and
;;;;:"aio ritrit. rrr" in-situ vertical effective at
Comprehensive Ouestions and Practice Problems 417

theproposedloadwillcauseo.toincreaseby500 lb/ft':. Whichequationshouldbeusedto

compute the consolidation settlement, 11.23,11.24, or I1.25?
r*ty?

11.21 A 3.0 m thick fill with a unit weight of 18.1 kN/mr is to be placed on the soil proiile shown in
Figure 11.24. Consolidation test results at Points A and B are as stated in hc5lern ll.ll'
except that o"' at point B is now 200 kPa. Using Equation I l.2l with Q or Q. as appropriate,
develop a plot of vertical strain, e. r's. depth from the original ground surface to the tcp of the
glacial till. How does this curve vary within a given soil stratum? Why? Does it suddenly
change at the strata interfaces? Why?

11.22 Considering rhe variation of strain with depth, as found in Problem I 1.21, does a I m thick
layer near the top of a sEatum contribute more or less to the consolidation settlement than a I m
thick stratum near the bottom? Explain. Does this finding support the statement in Section 11'8
that "The presence of crusts has a significant impact on setilement computations, even if they
are much thinner than the underlying compressible soils?" Explain.

11.23 A shopping center is to be built on a site adjacent to a tidal mud flat. The ground surface
elevation is +0.2 m, and the grounilwater table is at the ground surface. The underl-ring soils
consist of 7.3 m of medium clay with C"l(l+e) = 0.18, C /(l+e = 0.06, oJ = 0, and
)
y = l5.l kN/m3. The clay stratum is underlain by relatively incompressible stiff soils.
In order to provide sufficient flood control protection, a fill must be placed on this site
before the shopping center is built, thus maintaining the entire site above the higbest flood level.
This fill will have a unit weight of 19.0kN/m3.According to a hydrologic study, the fill must
be thick enough so that the ground surface elevation is at least +1.8 m after all of the
consolidation settlement is complete. Using program FILI,ScII, determine the required ground
surface elevation at the end of construction. Assume all of the setlement occurs after
construction.

l1J4 A consolidation test has been performed on a sample of lodgement till from a region that was
once covered with a glacier The current vertical effective stress at the sample location is
I 800 lb/ft'? and the measured preconsolidation stress is 32,500 lb/ff -

a. Assuming the glacier was in place long enough for complete consolidation to occur, and
assuming the ground surface and groundwater table elevations have remained unchange4
r compute the maximum thickness of the glacier- The specific gravity of glacial ice is
about 0.8?.
b. Although glacial ice was prsent for a very long time, it atso extended c ver very large
qreas, so the required drainage distance for the excess pore water was very long. As a
result, all of the excess pore water pressures may not have dissipated (Chung and Finno,
1992). Therefore, our assumption that complete consolidation occurred may not tte
accurate. If so, would our computed thickness be too large or too small? Explain.

11.25 A highway is to be built across a wetlands with the soil profile shown in Figure I I .27 belo*'.
These wetlands are subject to flooding, so a fill must be placed to keep the pavement above the
highest flood level. According to a hydrologic analysis, the roadway must be at elevation 7.0 ft
or higher to satisfy this requirement. Sandy fill material that has a compacted unit weight of
l22lblftl is available from a nearby bonow site.
A subsurface exploration program has been completed at this site, and laboratory tests
have bJen performed. The results ofthis program are tabulated below:
418 Compressibility and Settlement . Chap. 11

DeDth
Drvunit Moisture
-'
(it) serght content C,l(t+e) C,t(l+e) q'(lb/fi?)
(rb/ir') (7,)

2.0 95 28.6 0.t3 0.06 3000

7.5 89 33.0 0.16 0.06 550

13.0 92 30.5 0.12 0.05 850

21.0 93 29.9 0.14 0.07 4800

All depths are measured from the original ground surface.

will settle under the weight

The natural soils
of the proposed fill.
Agproximately 25 years will
pass before this settlement is complete. Therefore,
the road must be built at an elevation higher than
7.0 ft so that after the settlement is complete it is at
7.0 ft. The pavement thickness is 0.5 ft, so the top
of the fill must remain at or above elevation 6.5 ft.

a. Using program FILLSETT, determine the

required elevation of the roadway
immediately after construction. _ Assume
that no settlement occurs during
construction, and the pavement has the same
unit weight as the fill.
Hint: As the filI thickness becomes
greater, the settlement increases. Thus, this
problem requires a trial-and-error solution.
You will need to estimate the required fill
thickness, then compute the settlement and
final roadway elevation. Try to have one
trial that.produces a road elevation that is
. too high, and another that produces one too
I

.low. Then interpolate to hnd the required

fill thickness. :

Note 1: As the fill settles, the lower {

r.. , . portion will become submerged below the a:i --T...-

. - groundwater table, so ool will be less than Figure Il27 Cross-section for Problem
predicted by Equation I1.2. However, we
r,., . ,.. will ignore this effect. .:i.. , : i
Note 2: Laboratory tests have been
performed on two samples of the soft clay. Combine these two sets of test results, then
assign y, C.l(l+er), C,l(l+e), and o-' values that apply to the entire stratum.
Note 3: We dgnot have any unit weight data for the portion of the crust above the
groundwater table. Therefore, assume it is the same as that below the groundwater table.
In this case, this assumption should introduce very little error.
b. Is our assumption rdgarding the submergence of the fill (per note I in part a)
conservativ or unconservative? Is this a reasonable assumption? Explain.
Comprehensive Ouestions and Practice Problems 419

11.26 An engineer has suggested an alternative design for the proposed highway in Prc.blem I 1.25.
This design consists of using geofoam for the lower part of the fill, as shown in Figure 6.39.
' It will be covered with at least 1.0 ft of soil to provicle a buffer between the pavement and the
geofoam. Compute the minimum required geofoam thickness so that the roadway will always
be at elevation 7.0 or higher.

Hint: The ge6foam is an extra "hidden" layer between the fill and the natural ground surface.
If we assume the geofoam has y = 6, then it can be ignor{ when using program FTLIJET'T.

11.27 Aseries of prefabricated dual-bore steel tubes similar to the one in Figure 11.28 are to be
installed in an underwater trench to form a tunllel. The trench will be in seawater, which has
. a unit weight of 64.0 tbfft1 The tubes will be floated into position, and sunk into place by
temporarily flooding the interior. Then, non-structural concrete will be placed irrto chambers
along the tube to act as ballast, and the inside will be pumped dry. The completed nube will be
80 ft wide, 300 ft long, and 40 ft tall, and weigh 32,000 tons exclusive of buoyant forces.
Finally, the tube will be covered with soil, producing the cross-section shown in Figure I 1.29.

a. The interior of the tube will be dewatered after the concrete is placed, but before the
uench is bacldilled. Once this is done, will the tube remain at the bottom of the trench,
or will it float up to the water
surface?
b. After the trench is backfilled,
what will be the net Ao. in the
soft clay? Assume Ao-is constant
' with depth.
c. Using the final cross-section,
compute the ultimate
consolidation settlement or heave
of the tube due to Ao.' in the soft
clay. Assume no heave occurs
during construction.
d. The weakest parts of the completed - ,
tunnel will be the connections berween
the tube sections. In order to avoid
excessive flexural stresses at these
connections, the structural engineer has
specified a maximum allowable
differential settlement or differential
heave of 5 in along the lengttr of the tubb:;i

I (the term "allowable" indicates this

value already includes a factor of
safety). An evaluation ofthe soil profile
.. suggests the differential settlement or
heave rvill be no more than 50 percent of
the total. Has the structural engineer's
criteria been met? Figure I lJ8 This prefabricated tunnel section is part
of the Centril Artery Project in Boston. It was floated
t into position, then sunk to rhe bottom of the bay
(Photograph by Peter Vandenvarker, courtesy of the
Central Artery/Tunnel Project).
 Compressibility and Settlement ' Chap. 11

V
I

35 ft Sea water

-.,1"!: !r

Figure 11.29 Final cross-section for underwater tunnel as described in Problem 11.27.

11.26 A proposed building is to have three levels of underground parking, as shown in Figure I 1.30.
To construct this building, it will be necessary to make a 10.0 m excavation, which will need
to b temporarily dewatered. The natural and dewatered groundwater tables are as shown, and
the medium clay is normally consolidated. The chief geotechnical engineer is concerned that
this dewatering opration may cause excessive differential settlements in the adjacent building
and has asked you to compute the anticipated differential settlement across the width of this
building. Assume the wall is perfectly rigid, and thus does not contribute to any settlement
problems, and that the maximum ailowable differential settlemen! from one side of the building
to the opposite side is 50 mm. Neglect any loss in o.'below the existing buildiig due to the
removal of soil from the excavation. Discuss the implications of your answer.

l-<_zem___-] l-_rn1
Existing building
Proposed
building

'.'1=1
":i!;i.::i
1O3D]':
:; ;:.', ::. .h .:l l

1,".:"''j"l rc
.,-.r: Figure ll3{l Cross-section for Problem I l -28.

11.29 The i'alacio de las Bellas Artes in Mexico Ciry, shown in Figure I l.l, is an interesting example
of large consolidation settlement. It is supported on a 1.8 to 3.0 m thick mat foundation which
' is approximately 65 m wide and l 15 rn.long, The average bearing pressure between the bottom
of this mat and the supportingsoil is I l5 kPa (Ledesma, 1936).
The soil conditions beneath the palace are too complex to describe in detail here.
Ho*'ever, we can conduct an epproximate analysis using the following simplified profile:

I
Cornprehensive Questions and Practice Problems 4?1

C,
Depth v
Description 1-eo
(m) (kN/mr)

0-5 Sandy filI t7.5 0

5-45 Normall;- consol i dated soft clays I 1.5 0.53

>45 Stiff soils 0

For our simplified analysis, use a groundrvater table at a depth of 5 m, and assume the bottom
of the mat is at the original ground surface. In addition, assume the ftll has been in pace for a
very long tinie, so the consolidation settlement due to the rveight of the fill is co.nplete.
Divicie the soft clay zone beneath the center of the building into five layers of equal
thickness. Then, compute (a.)ra*,a at the midpoint of each layer using the methods described
in Chapter 10. Finally, compute the ultimate consolidation settlement beneath the cenrer of the
palace due t0 its own weight.

11.30 Develop a spreadsheet that can compute one-dimensional consolidation settlement due to the
weight of a flll. This spreadsheet should be able to accommodate a fill of an1' unit weight and
thickness, un,lerlain by multiple compressible soil strata. It also should be able to accommodate
both normally consolidated and overconsolidated soils. Since the computer does all of the
computations, the spreadsheet should use at least 50 layers. Once the spreadsheet is completed,
use it to solve Examples i 1.5 and I 1.6. Submit printouts of both analyses.

11.31 A fill is to b.e piaced at a proposed construction site, and you need to determine the ultimate
consolidation settlement due to its weight. Write a 20O-300 word essay describing rhe kinds
of field expl,oration, soil sampling, and laboratory testin,e you will need to perform to generate
the information needed for this analysis. Your essay should describe specific things that need
to be done, i,rnd what information will be gained from each activity.