Source: GEOTECHNICAL ENGINEERING

17 Time-Rate of Settlement

17.1 OVERVIEW

17.1.1 Time Is Timely

A settlement analysis not only should indicate the amount of anticipated
settlement, but also should give an indication of the rate at which it will occur.
Measurements in laboratory consolidation tests indicate that after each increment
of pressure is applied, consolidation is nearly complete after 24 hours, whereas the
time for foundations to reach equilibrium is measured in months or years. Some
type of scaling factor is necessary.

17.1.2 Terzaghi’s Research

Terzaghi’s consolidation apparatus included small standpipes to measure
excess pore water pressure developed during compression of soil, and pore
water pressures were observed to quickly increase after a compressive load was
applied, and then gradually decrease as water drained out and allowed the soil
to compress. Sand was found to compress much more rapidly than clay,
indicating that the soil permeability might be a controlling factor––but why
should so much more time be required in the field compared with in the
laboratory?

17.2 TERZAGHI’S THEORY

17.2.1 Development of a Theory

Terzaghi’s theory of consolidation is considered one of the cornerstones of
geotechnical engineering. His theory is based on (1) continuity of fluid flow and
(2) an application of Darcy’s Law. The derivation is highly mathematical, but as is

438 Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 439

the case with many classical engineering formulas, results have been simplified to
make them easier to use. It is suggested that the student follow through
the theoretical development in order to gain a better understanding of the basis
for the theory.

The principle of continuity is illustrated in Fig. 17.1, where the flow rate of water
upward into the bottom face of the differential element, qz, equals the area, dx dy,
times velocity, v:
qz ¼ v dx dy ð17:1Þ
The amount of water departing from the differential element is increased as a
result of vertical consolidation of the element:
 
@v
qzd ¼ v þ dz dx dy ð17:2Þ
@z
In eq. (17.2) the partial differential term, @v/@z, represents the increase in flow rate
in the z direction due to consolidation of the element. Subtracting the volume of
water departing from the volume entering gives the volume of pores lost, or the
amount of consolidation, per unit of time:
@v @n
qzd
qz ¼ dx dy dz ¼ dx dy dz ð17:3Þ
@z @z

Figure 17.1
Differential
element of
compressed soil.
The volume
change of
saturated soil
equals the volume
of water ejected.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

440 Geotechnical Engineering

@v @n
¼ ð17:4Þ
@z @z
In eq. (17.4), @v/@z is the change in velocity across the element and @n/@z is the
change in porosity, n (expressed as a fraction) with time. This is the continuity
equation for one-dimensional consolidation. It also may be expressed in terms of
the void ratio, since n ¼ e 7 (1þe):
 
@e 1 @v
ð17:5Þ
@z 1 þ e @z
That is, the higher the rate of change of fluid velocity with respect to vertical
distance, the higher the rate of change of the void ratio.

17.2.2 Consolidation and Darcy’s Law

The next step is to introduce Darcy’s Law, q ¼ kia, or
v ¼ ki
where k is the coefficient of hydraulic conductivity and i is the hydraulic gradient.
This may be written
@n
v¼k ð17:6Þ
@z
where @h/@z is the rate of change in head with respect to z within the element.
Changes in elevation head are compensated by equal and opposite changes in
static pressure head, so h signifies excess pressure head. Multiplying by the unit
weight of water converts this to excess pore water pressure, u:
u ¼ h
w
where h is the excess pore pressure expressed as head and
w is the unit weight of
water. Then
k @u
v¼ ð17:7Þ

w @z
Differentiating eq. (17.7) with respect to z gives
@v k @u2
¼ ð17:8Þ
@z
w @z2
Combining eqs. (17.7) and (17.5) gives
 
@e 1 k @u2
¼ ð17:9Þ
@t 1 þ e
w @z2
For small strains the change in void ratio @e is proportional to the change in pore
pressure @u:
@u
av ¼ ð17:10Þ
@e

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 441

where av is a coefficient of compressibility. Substitution for @e in eq. (17.9) gives

@u kð1 þ eÞ @u2
¼ ð17:11Þ
@t av
w @z2
If we define a coefficient of consolidation, Cv:
kð1 þ eÞ
Cv ¼ ð17:12Þ
av
w
then
@u @u2
¼ Cv 2 ð17:13Þ
@t @z
This is Terzaghi’s consolidation equation, and indicates that the reduction in excess
pore pressure with time is a function of the rate of change of pore pressure with
distance. The longer the distance, the slower the rate of drainage. This is the scaling
equation that enables laboratory time to be converted into field time. By definition,
compression that depends on the rate of ejection of pore water is primary
consolidation. It is important to note that the coefficient of consolidation, Cv, is a
time function, not a pressure function as in the case of the compression index, Cc.

17.2.3 Solving for Primary Consolidation

Drainage time is instantaneous at the boundaries of a consolidation layer and
increases with distance from a boundary. Solution of eq. (17.13) therefore requires
evaluating and summing excess pore pressures at different levels within a
consolidating layer. The distribution of pore water pressure in a consolidating
layer between two drainage faces is illustrated in Fig. 17.2, which is the drainage
geometry in a laboratory consolidometer. This same arrangement occurs in the
field where a consolidating clay layer is sandwiched between two sand layers and
the drainage directions are vertical, both upward and downward.

The solution of eq. (17.13) is obtained by defining boundary conditions and

integrating between those boundaries with Fourier series. This solution is
presented by Taylor (1948), Jumikis (1962), Holtz and Kovacs (1981), and
others. Jumikis develops a useful identity to define the percent total pore pressure
dissipation U in a consolidating layer, and hence the percent of total consolidation.
In Fig. 17.2, U is the enclosed area labeled ‘‘pore water pressure,’’ and can be seen
to decrease as consolidation goes from 50 to 90 percent completion. The series is
8
Nt 1
9Nt 1
U¼1
ðe þ e þ e
25Nt þ . . .Þ ð17:14Þ
2 9 25
where t equals time and
2 Cv
N¼ ð17:15Þ
4 H2

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

442 Geotechnical Engineering

Figure 17.2
Distribution of
pore water
pressure and
effective stress in
a consolidating
clay layer
between two sand
layers.

where H is the maximum vertical distance to a drainage face. Thus, in a clay layer
that drains in two directions, upward and downward, H equals one-half of the layer
thickness. Equation (17.14) quickly converges toward a final value but in theory
never reaches that value. The relationship can be simplified by defining a time
factor, T, as
Cv t
T¼ ð17:16Þ
H2
Substituting in eq. (17.15) and solving for N,
2 T
N¼
4 t ð17:17Þ
2
Nt ¼ T
4
which is the negative exponent in eq. (17.14).

17.2.4 Importance of the Time Factor T

Transposing eq. (17.16) gives
2
H
t¼ T ð17:18Þ
Cv
Consolidation time therefore is proportional to T and inverse to the coefficient of
consolidation, Cv. The coefficient of consolidation in turn is proportional to k, the

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 443

coefficient of hydraulic conductivity, in eq. (17.12), so the higher the permeability,

the shorter the time required for consolidation. Terzaghi’s analysis demonstrates
why consolidation of sand is many times faster than that of clay, other factors
being the same.

17.2.5 Importance of H 2
The maximum drainage distance, H, is particularly significant because it is
squared, so doubling the drainage distance increases consolidation time by a factor
of 4. The drainage distance is doubled if drainage can only occur from one face
instead of two, which is the reason why a large foundation on a consolidating layer
will be placed on a bed of sand, so that consolidation can occur more rapidly.

Because the drainage distance is squared, it can explain the huge difference
between consolidation times in the field and those that are measured in a
laboratory consolidometer. For example, if a soil layer 2.5 m (8 ft) thick in the
field is represented by a laboratory consolidometer sample 2.5 cm (1 in.) thick,
the thickness ratio is 250/2.5 ¼ 100, and the ratio of consolidation times is
(250/2.5)2 ¼ 10,000. It is this relationship that gives practicality to the consolida-
tion test and the use of thin samples to reduce testing time.

Figure 17.2 is a schematic representation of drainage from a clay layer between two
sand layers. The pore pressure at any depth z is u, and the percent conversion from
pore water pressure to effective stress is indicated by the area U. Two solutions are
shown, one for an area ratio indicating 50 percent completion of consolidation,
and the other for 90 percent consolidation. It will be noted that because of the
double drainage geometry H is one-half of the clay layer thickness.

17.2.6 Values of T
Table 17.1 shows values of T. Rearranging eq. (17.18) gives
H2
Cv ¼ Tp ð17:19Þ
tp
where the subscript p designates a particular percent completion of consolidation.
Equation (17.19) is used to evaluate Cv for a particular soil and load increment
from time-compression data in the consolidation test. Once Cv is known,
eq. (17.18) is used to solve for time t for a given percentage of consolidation in
the field. Results typically are plotted as a settlement versus time curve. This is used
to predict the time that will be necessary for primary consolidation settlement to
reduce to an amount that is manageable.

Example 17.1
A laboratory consolidation test gives Cv ¼ 2.5 m2/year. How much time will be required for
80% consolidation of a 1.5 m (4.9 ft) thick layer of this clay that is underlain by shale and
overlain by sand?

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

444 Geotechnical Engineering

Table 17.1 Percent of total settlement, U  100 Percent remaining Time factor T
Calculated time
0 100 0.000
factors T for various
5 95 0.002
values of U
10 90 0.008
expressed as
15 85 0.018
percentages
20 80 0.031
25 75 0.049
30 70 0.071
35 65 0.096
40 60 0.126
45 55 0.159
50 50 0.197
55 45 0.239
60 40 0.286
65 35 0.340
70 30 0.403
75 25 0.477
80 20 0.567
85 15 0.684
90 10 0.848
95 5 1.13
98 2 1.50
99 1 1.78
99.9 0.1 2.71
100.0 50.05 3.00

Answer:
Step 1. With single drainage H equals the layer thickness, or 1.5 m.

Step 2. T from Table 17.1 for 80% completion is T80 ¼ 0.567.

Step 3. Substituting these values into eq. (17.18) gives
ð1:5 mÞ2
t80 ¼  0:567 ¼ 0:51 year ¼ 190 days
ð15:26Þ

17.3 DETERMINING CV

17.3.1 Taylor’s Square-Root-Of-Time Method

In order to determine Cv for a particular soil and loading condition, it is
necessary to fit experimental consolidation data for a particular soil to eq. (17.19).
Taylor (1948), at MIT, noted that if the percent completion of primary

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 445

consolidation determined by Terzaghi’s theory is plotted versus the square root of

the time factor T, the graph initially is a straight line, as shown in Fig. 17.3. Since
time t is proportional to the time factor T, the same relationship may be expected
to hold for consolidation data, which has been confirmed experimentally.

In Fig. 17.3 the linear portion of the theoretical relationship between percent
p
consolidation and T is extended from O to B. A point then is established on the
theoretical curve that represents 90 percent consolidation, and a line is drawn
through that point, from O to C. The horizontal distance measured from the
ordinate is AC ¼ 1.15AB.

The same construction can be applied to experimental data, as in Fig. 17.4, to

p
define the t for 90 percent primary consolidation. This graph uses measured
deflections and does not require calculating void ratios. The value of Cv then is
determined from eq. (17.19) by substituting appropriate values for drainage
p
distance H, the measured t, and T90 ¼ 0.848.

The Taylor or ‘‘square root of time’’ procedure for estimating Cv is as follows:

1. Plot timed consolidometer compression data on the y-axis, versus correspond-

p
ing values of t on the x-axis.
2. Draw a straight line through the linear portion of the plot, omitting the first
point(s) if necessary as they may be influenced by elastic compression.
3. Draw a horizontal line from the y-axis to intersect the linear plot, measure
the distance to the y-axis, and mark off that distance times 1.15. Then draw
another line from that point to the origin at the upper left of the graph. The
p
intersection of this line with the data plot defines t90 for 90 percent primary
consolidation, and is read from the graph. Square to obtain t90.
4. Substitute t90 in eq. (17.19) and calculate Cv.

Figure 17.3
Graph of
theoretical percent
consolidation
p
versus T that is
the basis for
Taylor’s
construction to
determine Cv.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

446 Geotechnical Engineering

Example 17.2 p
The Taylor construction in Fig. 17.4 indicates that t90 ¼ 5.3 min.0.5, so
2
t90 ¼ 5.3 ¼ 28.1 min. The sample thickness at the beginning of this episode of consol-
idation was 1.383 in. (35.13 mm) and at the end was 1.315 in. (33.40 mm), giving an average
of 1.349 in. (34.26 mm). (1) Calculate Cv, and (2) predict the time for 95% consolidation of
a clay stratum 3 m thick with double drainage.

Answer: (1) The determination of Cv is very sensitive to the sample thickness so the average
thickness before and after consolidation is used. The next step (and the one that is most
often overlooked) is to divide the thickness by 2 to obtain H for a double-drainage
situation: H ¼ 34.27/2 ¼ 17.13 mm. Then from eq. (17.19),

H2 ð17:13 mmÞ2
Cv ¼ T90 ¼  0:848 ¼ 8:86 mm2 =minute
t90 28:1 min
The preferred units for application to a field situation are m2/year or ft2/year. To avoid
mistakes, the series of conversion factors should be written down so that units can be
cancelled. Thus in the first two terms, minutes cancel minutes, which converts the time to
hours, and so on, until the remaining units are m2/year:
 



2 60 min: 24

mm hours




365 days 1 mm 2
cv ¼ 8:86     ¼ 4:66 m2 =year
min : 1 hour




1 day 1 year 1000 mm




(2) The value of Cv is substituted into eq. (17.18) with the appropriate values for H and T
for the field situation, where the maximum drainage distance H again is one-half of the
layer thickness:
H2 ð1:5 mÞ2
t95 ¼ T95 ¼  1:13 ¼ 0:55 years ¼ 7 months
Cv 4:66 m2 =year
Note that the final answer is given to no more than two significant figures because that is
the limit of accuracy of reading the graph, so to include more would be to claim an
unwarranted precision in the results.

Figure 17.4
Application of the
Taylor method to
data for the
pressure interval
1 to 2 tons/ft2 in
Table 16.1.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 447

Question: Upon careful inspection of the soil boring logs and samples, a 2-inch
layer of sand is discovered that divides the clay layer into two equal layers. How
might that influence settlement?
Answer: The total clay thickness is about the same but the sand divides it into two
layers, thereby reducing the drainage distance of each by a factor of 2. As the
drainage distance is squared, time will be divided by a factor of 4, reducing the
time requirement from 7 months to 7 weeks.

17.3.2 Accuracy of the Taylor Method

The data in Fig. 17.4 deviate from the ideal theoretical curve as consolidation
progresses toward completion. This is reasonable because Cv depends on the
coefficient of hydraulic conductivity, which decreases as soil compresses. This can
be confirmed by determining Cv for different load increments. However, will this
p
affect the determination of t90?

In Fig. 17.4 the total amount of compression is 0.061 in. Ninety percent of that is
0.055 in., which closely agrees with the amount predicted by the intersection of the
data plot with the 1.15-sloping line.

17.3.3 Logarithm of Time Method

An empirical procedure suggested by A. Casagrande at Harvard University
involves plotting compression dial readings versus the logarithm of time.
This is shown in Fig. 17.5 for the same data as in Fig. 17.4. The inter-
section of the two straight lines through the data, at 40 minutes, was proposed
to define the end of primary consolidation, although this is more accu-
rately defined from measurements of pore water pressure. One-half of the
compression at this point is 0.0300 in., which as shown in Fig. 17.5 corresponds
to a time of 6 minutes. This value and T50 are substituted in eq. (17.19) to
obtain Cv.

Figure 17.5
The log-time
method applied to
the same data as
in Fig. 17.4.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

448 Geotechnical Engineering

Example 17.3
Evaluate Cv for the previous example using the log-time method. Note that in this case time
t ¼ 6 min is read directly from the graph and is not squared.

Answer:
H2 ð17:13 mmÞ2
Cv ¼ Tp ¼  0:197 ¼ 9:6 mm2 =minute
tp 6 min
which is very close to the estimate of 8.9 mm2/minute from the Taylor method.
The initial data points that are not on the line occur during the first 3 minutes and
indicate a quasi-elastic response, discussed in the next section.

17.4 FORE

17.4.1 Overview
In many natural phenomena the rate of approach to an end condition is
proportional to the departure from that condition, similar to an automobile
gradually slowing to a stop. This concept of a steadily diminishing rate is
formalized in a ‘‘first-order rate equation’’ of chemical kinetics. Its application in
geotechnical engineering recently was given the acronym ‘‘FORE’’ (Li, 2003).

A requirement of FORE is that the primary thesis must be met, that a rate of
change is proportional to the departure from some end condition. Often
conditions will change and require definition of a new end point and a new rate
equation, such as the shift from primary to secondary consolidation.

In some cases a change may be continuous and closely coupled with the primary
rate, in which case it will not be discerned by the rate equation (Handy, 2002).

17.4.2 FORE for an End Value

Let D equal an unknown end value of y that in theory will never be fully achieved.
The amount of departure from the end value at any time is (y – D).

A first-order rate equation states that the rate of change d(y – D) with respect to an
independent variable x is proportional to (y – D). This can be stated mathemati-
cally by
dðy
DÞ
¼
kðy
DÞ ð17:20Þ
dx
where k is a constant and the minus sign indicates a declining rate. Rearranging,
dðy
DÞ
¼
k dx ð17:21Þ
ðy
DÞ

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 449

from which
In ðy
DÞ ¼
kx þ C ð17:22Þ
where C is a constant of integration. This is the basic rate equation expressed
in the form y ¼ ax þ b, which means that if the rate equation method is appli-
cable, data inserted into eq. (17.22) must plot as a straight line. Converting to
logarithms to the base 10, and in recognition that there is no logarithm of a
negative number,
log10 j y
D j ¼
k10 x þ C10 ð17:22aÞ
The final value, D, is not known so it is determined by trial and error, by
substituting trial values of D into the equation until it yields a straight line. If D is
not correct, the line will curve and a new value should be tested. If a single D value
does not meet this criterion it may be necessary to departmentalize the data and
define two separate rate equations.

As D is varied, convergence to develop a straight line is rapid and is easily

accomplished in a minute or two with a computer spreadsheet. Both y and D must
have the same sign convention: if compression y is negative, the maximum amount
of compression D also must be negative.

17.4.3 When Conditions Change

The rate equation changes if the system changes. Such a change can be expected
as primary consolidation that depends on the rate of water being expelled from soil
yields ground to secondary consolidation that is much slower and related to soil
creep. Test data should include some overlap because drainage and the initiation of
secondary consolidation should begin immediately in soil next to a drainage face.

The procedure to find D is as follows and is illustrated in Table 17.2:

1. Set up a spreadsheet with x and y data, for example time and compression or
settlement amount data in the first two columns.
2. Reserve the third column for trial values of D, which are entered at the top and
automatically copied to other cells down the column.
3. In the fourth column calculate log10jy – Dj where jy – Dj is the absolute or
positive number.
4. Plot log10jy – Dj from column 4 versus x from column 1.
5. Visually observe linearity of the plot and vary D to obtain a straight line.

For convenience the graph can be displayed on the spreadsheet to instantly

display effects of changes in D. A simultaneous calculation of the statistical R2
allows fine-tuning of D to obtain the best linearity and highest R2, shown at the
lower right in Table 17.2. Regression coefficients a and b also are calculated and
are used to write an equation for settlement versus time.
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

450 Geotechnical Engineering

Figure 17.6 shows a solution for D from the data in Table 17.2, and the effect of
varying D by  0.01, or in this case  1.7 percent, illustrating the excellent
sensitivity of this procedure.

Error messages in Table 17.2 result from attempting to calculate the logarithm of
a negative number, because the best-fit D ¼ 0.061 is less than the actual final

Table 17.2 Time in minutes Dial, inch Compression y in inches Trial D in inches |y
D |
Spreadsheet for
0 4478 0.0000 0.061 —
application of FORE
0.25 4400 0.0078 0.061
1.274
to primary
1 4343 0.0135 0.061
1.323
consolidaton data for
2.25 4296 0.0182 0.061
1.369
the 1 to 2 tons/ft2 load
4 4230 0.0248 0.061
1.441
increment in
6.25 4173 0.0305 0.061
1.516
Table 16.1. For
9 4116 0.0362 0.061
1.606
convinience a graph
12.25 4059 0.0419 0.061
1.719
such as shown in
16 4011 0.0467 0.061
1.845
Fig. 17.6 can be
25 3946 0.0532 0.061
2.108
plotted on the
36 3900 0.0578 0.061
2.495
spreadsheet to show
49 3880 0.0598 0.061
2.921
changes in linearity
64 3872 0.0606 0.061
3.398
with different ultimate
100 3860 0.0618 0.061 ERR
trial values
225 3838 0.0640 0.061 ERR
400 3821 0.0657 0.061 ERR
1440 3793 0.0685 0.061 ERR
R2 ¼ 0.9992
a¼
0.0333
b¼
1.3090

Figure 17.6
D values are
varied to obtain a
straight-line plot.
Sensitivity is to
50.001 in.
(50.025 mm). For
convenience in
testing D values
this graph can be
plotted directly on
a spreadsheet.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 451

measurement of 0.0685. This indicates that there is a discontinuity in the data,

which is the anticipated shift to secondary consolidation as time exceeded
64 minutes. Longer times therefore were excluded from the determination of the
first trial D and R2, and then treated separately to obtain another D end value for
secondary consolidation.

17.4.4 Prediction Equations

After a value for D has been evaluated and the regression coefficients obtained,
the values are entered into eq. (17.22a) solved for y:
log ðD
yÞ ¼ ax þ b
ð17:23Þ
y ¼ D
10 expðax þ bÞ
The equation for primary consolidation for the data in Table 17.2 is
y ¼ 0:061
10 expð
0:033t
1:309Þ ðinchesÞ
where y is the amount of compression in inches and t is time in minutes. The same
procedure is used with SI units. Applying the procedure to the remaining data that
represents longer times gives
ys ¼ 0:069
10 expð
0:0014t
0:199Þ ðinchesÞ
for secondary consolidation. The two respective end amounts are 0.061 and
0.069 in. (1.55 and 1.75 mm), indicating that approximately 89 percent of the final
compression is attributed to primary and 11 percent to secondary consolidation.
These two equations, one for primary and the other for secondary consolidation,
are shown by heavy solid lines in Fig. 17.7, and show convergence and the
anticipated overlap.

The time at which primary gives way to secondary consolidation is indicated to be

between 50 and 60 minutes, compared with 40 minutes in Fig. 17.5.

Figure 17.7
The solid lines
show results from
a FORE analysis
for primary and
secondary
consolidation, and
show close
agreement with
measured values.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

452 Geotechnical Engineering

17.4.5 Initial Compression

As in the other methods, there are initial deviations from the ideal equation during
the first 3 minutes. This rapid response suggests elastic compression. Substituting
t ¼ 0 in the FORE equation gives a starting value for primary consolidation, which
then may be compared with the initial zero-load specimen thickness to obtain
a modulus. The most accurate determination will be if the soil has been preloaded
to the overburden pressure and then unloaded, as discussed in the preceding
chapter.

It should be noted that the modulus measured during compression in a

consolidometer is a confined modulus because lateral expansion is prevented.
The confined modulus generally is higher than the Young’s elastic modulus where
lateral expansion is allowed, but is applicable to a field compression situation
under a broad loaded area.

Example 17.4
Calculate the confined modulus from the initial response to loading.

Answer: When t ¼ 0, y ¼ 0.061
10 exp 1.309 ¼ 0.0119 in. (0.48 mm). The specimen
thickness at the beginning of this load increment (see Example 17.2) was 1.383 in.
(35.1 mm), so elastic strain was " ¼ 0.0119/1.383 ¼ 0.00859.

The load increment is 1 ton/ft2, or 13.9 lb/in.2 (95.8 kPa). Therefore Ec ¼ 1/0.00859 ¼
120 tons/ft2 or 13.9/0.00859 ¼ 1600 lb/in.2 (11 MPa). Note that calculations carry three
significant figures, but the answers are reported only to two.

17.4.6 Transforming Thickness and Time Scales to

the Field
In order to apply laboratory findings to a field situation, the settlement scale is a
simple ratio of the soil layer thickness in the field to the sample thickness in the
consolidometer, H/h.

According to Terzaghi’s theory and as in eq. (17.18), time is proportional to the

square of the drainage distance, or
 2
H
tf ¼ t ð17:24Þ
h
where tf is time in the field, t is the time for a comparable amount of primary
consolidation in the laboratory, and H and h are the corresponding field
and laboratory maximum drainage distances. Thus, for a double drainage
situation the multiplier for time is the square of the ratio of field and labo-
ratory thicknesses, but for single drainage situation in the field it is four times
that ratio.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 453

Example 17.5
Prepare a graph showing percent primary consolidation settlement versus time for a
3 m (10 ft) thick layer of the clay represented in Table 17.2. Assume double drainage.

Answer: The ratio of field to laboratory thickness is 3000 mm/34.27 mm ¼ 87.5, so the scale
factor for thickness is 87.5, and for time is (87.5)2 ¼ 7660. Both the thickness and time
conversions are accomplished by adding two additional columns on the spreadsheet.
A transformed graph for primary consolidation is shown in Fig. 17.8.

Question: How does this result compare with that from the Taylor method?
Answer: The Taylor method (Example 17.2) indicated that 95% consolidation
would be completed in 7 months, which agrees with the FORE analysis. FORE
indicates that 2 months will be required for 60% primary consolidation, compared
with the Taylor method determination of
ð1:5 mÞ2
t60 ¼  0:286 ¼ 0:138 years ¼ 1:7 months
4:6 m2 =yr

17.4.7 Consolidation Test of Quick Clay

Consolidation tests and simultaneous pore pressure measurements of a quick clay
by Crawford (1964), of the National Research Council of Canada, gave a puzzling
result that appeared to conflict with Terzaghi’s theory, as pore water pressures
declined to zero before primary consolidation was complete. Might an
explanation be hidden away somewhere within the clay?

A FORE analysis of the primary consolidation data showed that pore water
pressures reached zero; exactly as predicted by Terzaghi theory (Fig. 17.9). A
FORE analysis applied to long-term data then left a large gap shown in the figure,
and a speculation that consolidation involved three instead of two stages:
(1) primary consolidation during which pore water pressure, which gradually
increased pressure on the structure, caused (2) a time-dependent breakdown of the
structure, followed by (3) secondary consolidation as soil grains came into more
intimate contact.

Figure 17.8
Transformed
thickness and time
scales from
Fig. 17.7 for
primary
consolidation
settlement in the
field for
Example 17.5.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

454 Geotechnical Engineering

Figure 17.9
FORE predicts the end time for primary consolidation of a quick clay, and suggests two other
stages of compression. Data are from Crawford (1964). (From Handy, 2002, with permission of the
American Society of Civil Engineers.)

17.5 CORRECTING SYSTEMATIC ERRORS AFFECTING

CONSOLIDATION TIME

17.5.1 Temperature Correction

Because Cv is directly proportional to the coefficient of hydraulic conductivity,
which in turn depends on viscosity of the pore fluid, temperature differences
between laboratory and field can introduce systemic error. The groundwater
temperature in the field generally reflects the mean annual temperature, so in
temperate climates Cv may be overestimated and time underestimated from
laboratory tests. Times therefore should be corrected by multiplying by the ratio
of water viscosities at the field and laboratory temperatures:

ktemp ¼ ð17:25Þ
f
where ktemp is the temperature correction factor for primary consolidation time
and  and f are respective viscosities of water in the laboratory and in the field.
Viscosities are shown in Fig. 17.10.

Example 17.6
The mean annual temperature obtained from weather records is 588F (14.48C) and the
laboratory temperature during a consolidation test is 728F (22.28C). What correction
should be made to primary consolidation times?
Answer: Inserting respective values of  from Fig. 17.10 gives
ktemp ¼ 1:22=0:95 ¼ 1:3
That is, primary consolidation times in the field should be multiplied times 1.3 to account
for the lower viscosity of water compared with the laboratory. This correction should be

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 455

Figure 17.10
Viscosity of water
in relation to
temperature.

made regardless of the analysis method. In this case, not making the temperature
correction would introduce 30% error which, as it would underestimate consolidation time,
would be on the unsafe side. The correction only applies to primary consolidation where
viscosity is a controlling factor, and not to secondary consolidation where chemical
bonding controls. A correction to secondary consolidation rate would be on the basis of the
absolute temperature, that is, the temperature range calculated relative to
2738C, so there
is little percent change.

17.5.2 Drainage Rate, Varved Clays, and Vertical Drains

The consolidometer models compression of a layer of soil in the field sandwiched
between two drainage layers. Some soils such as varved clays are composed of
regular interlayers of clay and silt in relatively thin layers that greatly reduce
consolidation time provided that there is lateral drainage—some place for the
water to go.

Vertical drains frequently are installed to hasten primary consolidation

settlement. These may be vertical borings filled with loose sand, called sand
drains, or a more modern equivalent is constructed by stitching in vertical
channels of plastic, called wick drains, with an apparatus that operates like a
massive sewing machine.
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

456 Geotechnical Engineering

Where vertical drains are used it is imperative that they not be covered up without
some provision for a drainage outlet at the ground surface. If rapid drainage is
anticipated, vertical drains will be connected to drain pipes that run laterally
offsite. Slower drainage can be through a layer of sand that may contain
perforated drain tiles for draining large areas.

The drainage rate of soil between the vertical drains is governed by horizontal
instead of vertical permeability, and cannot reliably be predicted from a
conventional one-dimensional consolidation test. The design of vertical drains is
specialized and is beyond the scope of this text. After drains are installed, the
progress of settlement can be predicted from settlement measurements and
FORE. An example is given later in this chapter.

17.5.3 Imperfect Modeling by the Consolidation Test

A consolidometer models compression of a broad layer of soil in the field, but also
incorporates side friction that does not exist in the field. Side friction is minimized
by smooth sides on the inside of the consolidometer ring, which may include a
Teflon liner. Side friction is eliminated in the triaxial compression test discussed in
the next chapter, but specimens are much thicker, drainage times many times
longer, and air pressure trapped in the pores can become significant.

A second difference between field and laboratory conditions is a much higher

hydraulic gradient in the laboratory test, which could be troublesome if it results
in movement of clay particles. If that occurs, clay will intrude into the porous
stone disks used to confine the soil layer. The hydraulic gradient is reduced by
using smaller load increments, which increases the time required for the test.

Another aspect of the modeling is lateral confining pressure, which will be reduced
by lateral bulging of soil under a small or narrow foundation area, which tends to
increase settlement. The effect of bulging can be duplicated in triaxial stress-path
testing, which was discussed in the previous chapter. Allowable foundation
pressures are based both on acceptable settlement and on bearing capacity failure,
which employs a generous factor of safety to minimize lateral bulging and prevent
shear failure.

17.5.4 Comparison of Measured and

Predicted Settlements
Settlement of buildings normally is limited to less than 1 inch (25 mm), so
comparing measured with actual settlement requires accurate elevation measure-
ments to stable benchmarks. A much larger amount of settlement can be tolerated
in earth dams, and one such comparison is shown in Fig. 17.11, which shows best
and worst settlement predictions for points in the Ft. Randall Dam, an earth dam
constructed in South Dakota on the Missouri River floodplain.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 457

Figure 17.11
Best and worst
predictions of
settlement at
Ft. Randall Dam.
The main
discrepancy at
Gauge 36 is from
immediate
settlement that
should not be
considered as
primary
consolidation.

In both cases in Fig. 17.11 settlement initiated almost immediately when

construction began, suggesting an initial elastic response that was not considered
in calculations of primary consolidation settlement. This was not included in
the settlement prediction and can explain the large discrepancy at Gauge 36, where
the immediate settlement was approximately 2.3 ft (0.7 m). If that amount is added
to the predicted primary consolidation settlement, the result closely agrees with
the measured total settlement. Settlement at this gauge had not stopped after
14 years, indicating a contribution from secondary consolidation. Therefore,
despite simplifications in the development and applications of the theory of
primary consolidation, if care is used in sampling and testing, predictions can
be reliable.

17.6 SECONDARY CONSOLIDATION

17.6.1 Overview
Secondary consolidation used to be like the crazy aunt in the attic: everybody
knows she is up there but don’t quite know how to deal with it. Unlike primary

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

458 Geotechnical Engineering

consolidation, there has been no guiding theory except to recognize that the
behavior is time-related and in some ways is similar to soil creep.

17.6.2 Empirical Procedures

The conventional method for analyzing secondary consolidation assumes that a
linear second part of a void ratio-log time curve extends indefinitely. The
coefficient of secondary consolidation is

e
C ¼ ð17:26Þ

log t
where
e is the change in void ratio over the time interval represented by
log t.
In order to calculate C, strain measurements must be reduced to equivalent void
ratios. C is readily evaluated from the change in void ratio during one log-time
interval. A more complete discussion of this method is presented by Holtz and
Kovacs (1981).

A more direct procedure is to define a ‘‘modified compression index,’’ C",

which is expressed in terms of strain instead of void ratio. The modified
compression index is determined by reading the change in strain over one log-time
interval:

"
C" ¼ ð17:27Þ

log t
In eq. (17.27),
" is the increment of strain over the time interval represented by

log t.

Example 17.7
According to eq. 17.27 and the data in Fig. 17.5, (a) what is the modified compression
index, and what will be the percent increase in strain (b) 1 year and (c) 10 years after
primary consolidation is completed?

Answer: (a) The change in strain over one log time interval in Fig. 17.5 gives C" ¼ 0.007.

(b) Equation (17.26) is rearranged to solve for
". If compression is taken as positive,
after 60 minutes or 1 hour and at the end of primary consolidation "0 ¼ 0.06.
One year ¼ 8160 hours, so the change in
log t ¼ (log 8160
log 1) ¼ 3.94. Multiplying
times C" gives
" ¼ 0.028, which added to "0 ¼ 0.06 gives strain after one year, " ¼ 0.088.

(c) After 10 years,
log t ¼ (log 81,600
log 1) ¼ 4.91, giving
" ¼ 0.034. Total strain is
" ¼ 0.094.

Are these answers reasonable? They indicate that after 1 year 32% of the strain is from
secondary consolidation, and after 10 years, 35%. These estimates appear to be excessive,
and do not take into account a probable decline in C" with time.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 459

17.6.3 Characteristics of Secondary Consolidation

An arguable ‘‘top 10’’ features that may help to explain secondary consolidation
are as follows:

1. There is no relation between rate of secondary consolidation and distance to

a drainage face. However, this does not preclude a relationship to layer
thickness.
2. According to the model shown in Fig. 17.2, primary and secondary
consolidation should begin simultaneously at a drainage face, but primary
consolidation will dominate until nearly all excess pore water pressure has
dissipated.
3. It usually is assumed that C remains constant with time, but field and labo-
ratory data indicate that it decreases with time (Mesri and Godlewski, 1977).
4. Data on different soils indicate that the ratio of C/Cc is approximately
constant, in the range 0.03–0.06. The more compressible the soil, the higher
the rate of secondary consolidation.
5. Secondary consolidation is not ideally plastic behavior because it involves
volume change.
6. Secondary consolidation ejects water and therefore must induce a hydraulic
gradient, however small, or the water would not move.
7. The slow rate of secondary consolidation may reflect a higher viscosity
of water loosely held in the electrical double layer. That being the case,
eqs. (17.6) ff. may still apply but with a much higher value of k.
8. The logarithmic relationship for C indicates that the modulus increases
linearly with time. This may be one aspect of ‘‘aging.’’
9. Secondary consolidation increases the apparent preconsolidation pressure by
shifing the e-log P curve to the right. From studies of Pleistocene clays,
Bjerrum(1954), of the Norwegian Geotechnical Institute, showed a relation to
the logarithm of time. This aspect of aging may help explain the scarcity of
normally consolidated soils.
10. Secondary consolidation settlement of structures still can be trouble-
some decades or even centuries after construction, and in thick clay strata
eventually may exceed settlement from primary consolidation. A saving
feature is that many structures are built with only a 50- to 100-year life
expectancy.

17.6.4 Application of the FORE to Secondary Consolidation

Some of the difficulties of predicting secondary consolidation may be managed
with FORE. Data after the end of primary consolidation are included in
Table 17.3, and the corresponding linear relationship for Ds ¼ 0.0686 in.
(1.74 mm) is shown in Fig. 17.12.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

460 Geotechnical Engineering

Table 17.3 Time in minutes Dial, inch Compression in inches Trial D in inches Log (D
d)
Spreadsheet for
64 3872 0.0606 0.0686
2.097
secondary
100 3860 0.0618 0.0686
2.167
consolidation for
225 3848 0.0630 0.0686
2.252
data in Table 17.2
400 3821 0.0657 0.0686
2.538
1440 3793 0.0685 0.0686
4.000
R2 ¼ 0.9984
a¼
0.0014
b¼
1.9902

Table 17.4 Soil E, tons/ft2 MN/m2

Some representative
Sand
values for soil
Loose 50–200 5–20
modulus E, after
Medium 200–500 20–50
Kulhawy (1991)
Dense 500–1000 50–100
Clay
Soft 25–150 2.5–15
Medium to stiff 150–500 15–50
Very stiff to hard 500–2000 50–200

Figure 17.12
FORE prediction
of secondary
consolidation end
value for data of
Fig. 17.7.

In Table 17.2 the 24-hour reading of 0.0685 in. is comparable with the projected
final predicted D ¼ 0.0686 in.

17.6.5 Transforming the Secondary Consolidation

Time Scale
As in the case of primary consolidation, the time involved in secondary
consolidation in the field is larger by orders of magnitude than in the laboratory,
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 461

so a time transformation is necessary in which drainage rate is not the

governing factor. A consolidometer test is a scale model, so the problem may
be examined using principles of similitude.

The scale factor is

H
n¼ ð17:28Þ
h
where H is the soil layer thickness in the field and h is the initial thickness of the
sample tested in the laboratory. The soil itself is not scale modeled because the
grain size is the same in the laboratory test as in the field, but it is more important
that other physical properties, including clay mineralogy, particle interference,
and amount, viscosity, density, and surface tension of the pore water, should be
the same in the laboratory as in the field. As the model is scaled downward the
hydraulic gradient is scaled upward, but that should influence only primary
consolidation.

A model can be developed on the basis of dimensionless ‘‘pi terms’’ that are equal
in the model and in the prototype. The Reynolds number is a pi term that is used
to differentiate between lamellar and turbulent fluid flow. As the Reynolds
number is defined on the basis of viscosity and, as shown in the next chapter,
viscous behavior appears to be involved in soil creep and secondary consolidation,
the number can be written for the field prototype (on the left) and for the model
(on the right):
vr m vm rm
¼ ð17:29Þ
 m
In this expression  and m represent fluid density, v and vm velocity, r and rm a
corresponding length term, and  and m fluid viscosity. Because fluid density 
and viscosity  are the same in the model and in the prototype, velocity and length
terms are the remaining variables:
vr ¼ vm rm

The length term can be any linear measure that distinguishes between the model
and the prototype, and therefore may be represented by the respective layer
thicknesses, h and H. Substituting H/t for v and H for r and their equivalents in
the model gives

H2 h2
¼
t tm
where t is time. Transposing,
t H2
¼ ¼ n2 ð17:30Þ
tm h2

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

462 Geotechnical Engineering

Figure 17.13
Laboratory
consolidation test
data of Fig. 17.7
transformed to the
field prototype
scale.

This is the same result as that obtained by Terzaghi based on drainage distance. It
suggests that field time for secondary consolidation equals model time multiplied
by the scale factor squared, which is the thickness ratio squared.

The assumption of equal Reynolds numbers implies that viscosity is a controlling

factor—not the viscosity of water but of the soil-water complex. If surface tension
were the controlling factor, which might be the case in an unsaturated soil, the
Weber number would be more appropriate and would give a time ratio of n3. If
compressibility were the control, as in the case of gases, the Cauchy number
(which is the square of the Mach number) would gives a time ratio of n. If no fluid
behavior were involved, n would equal 1, which observation shows is not the case.
Squaring the scale factor therefore appears to be the most likely scenario, even
though this will distort the scaling of surface tension and compressibility if the soil
contains any undisolved air.

Use of the same time transformation for both primary and secondary
consolidation simplifies the prediction procedure, and is the basis for
Fig. 17.13, with primary consolidation settlement predicted to be essentially
complete after 1 year, and secondary after 20 years.

17.6.6 A Rational Basis for the Time Transformation

Results from a dimensional analysis, like Einstein’s time-space relationship,
sometimes appear to come from another world. The basis for Terzaghi’s time
transformation is that it takes longer for fluid to flow through a long pipe
than through a short one, although without the mathematical derivation
one might intuitively (and incorrectly) assume a linear relationship. Might
there be a rational explanation for the time transformation for secondary
consolidation?

One possibility might relate to probability and a ‘‘weakest link first’’ hypothesis.
This is a concept that the weakest grain contacts will slip first, thereby transferring
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 463

stress to adjacent grain contacts so that they will be more disposed to slip next.
However, secondary consolidation is a viscous phenomenon, so each slip eats up
time. In a thin layer of soil, slipping may more readily become a joint effort than
in a thicker layer.

17.6.7 Applications of FORE for Surcharge Loading

Figure 17.13 predicts a total final settlement of 6 inches (150 mm), which is
unacceptable for most buildings where total settlement typically is limited to a
maximum of 1 inch (25 mm). The usual options are to go to an intermediate or
deep foundation system, discussed in later chapters, or to use a temporary
surcharge to induce settlement prior to construction. The compelling question
then is how long should a surcharge be left in place?

If a full surcharge is applied, the answer is as shown in Fig. 17.13. For example, if
1 inch (25 mm) is allowable, this can be subtracted from the total and the time read
from the graph. In Fig. 17.13 only 6 months will be required and will mainly
involve primary consolidation, which means that it could be speeded up by use of
vertical drains.

The effect of a surcharge load that is less than the anticipated foundation load can
be assessed from the e-log P curve. Even a small reduction in load can result in an
unsatisfactory performance because the desired void ratio may not be attainable,
or if it is, time-consuming secondary consolidation may be required to accomplish
the same settlement as will occur during primary consolidation under the
foundation load.

17.6.8 Adjustment for a Long Construction Time

A long construction time, as for an earth dam, can be represented by compressing
the time-settlement prediction curve vertically to represent one-half of the final
settlement. The starting time is when construction is estimated to be at 50 percent
completion. The curve then is duplicated and moved to a second starting time at
100 percent completion. The curves are plotted to a linear time scale and
connected by a common tangent, and a tangent is drawn from the first curve to
the origin. The tangents then simulate additions of load that are linear with time
during the construction period. An example is shown in Fig. 17.14.

17.7 MONITORING SETTLEMENT

17.7.1 Overview
As indicated in Fig. 17.14, the field performance has the last word, and as
settlement is monitored it may be necessary to revise expectations. A prediction
method by Asaoka (1998) involves settlement measurements made at regular

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

464 Geotechnical Engineering

Figure 17.14
Settlement
predictions and
measurements for
Ft. Randall Dam
adjusted for the
construction
schedule.

intervals. The amount of settlement at the end of each interval is plotted versus the
amount at the end of each preceding interval. The point at which settlement will
stop is defined by a line at 458, so an intersection with the extrapolated data line
defines the time and amount of expected total settlement. A difficulty with the
application of this method is because the two lines are nearly parallel, an
intersection is not precisely defined.

17.7.2 FORE in the Field

The Kansai artificial island airport in Osaka Bay, Japan, afforded an opportunity
to test FORE, particularly as the upper supporting strata were drained with
vertical sand drains while the lower strata were too deep for economical
installation.

The FORE treatment to predict final settlement, D, at the location of maximum

settlement is shown in Fig. 17.15. A steel plate was located at the contact between
the two soil layers and attached to a pipe extending above the ground surface to
differentiate between compression to the two layers. The FORE settlement
predictions for the two layers is shown in Fig. 17.16, and shows rapid compression
of the upper layer and a much slower compression of the lower layer, confirming
the benefit from the drains. The FORE analysis indicated that the upper layer that
contained the drains quickly went into secondary consolidation. Of the predicted
final total settlement, 40 percent will come from compression of the upper layer
and 60 percent by compression of the lower layer. As the lower layer is a stiffer
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 465

Figure 17.15
FORE fit to settlement measurements at the artificial island supporting the Kansai International
Airport, Osaka Bay, Japan (Handy, 2002). Progresssive loading during the first 5 years was not
included in the analysis. Settlement data are courtesy of Professor Koichi Akai, University of
Kyoto, Japan. (Reproduced with permission of the American Society of Civil Engineers.)

Figure 17.16
Data and projected total settlement from compression of two soil layers at the Kansai International
Airport, Osaka Bay, Japan (Handy, 2002). (Reproduced with permission of the American Society of
Civil Engineers.)

Pleistocene deposit, the amount of its compression was not fully anticipated in the
original design.

Additional fill will be required in order to maintain freeboard, but an additional

load will reinitiate settlement. The amount of fill necessary can be estimated from
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

466 Geotechnical Engineering

the ratio of the existing fill thickness to the total predicted settlement. This ratio is
almost exactly 2:1, indicating that 10 m (30 ft) of fill will be required to add 4 m
(15 ft) of freeboard to the low spots.

The Kansai airport is considered an outstanding feat of engineering at a very

difficult construction site, and a second airport is under construction.

Problems
17.1. The time factor T in Table 17.1 is 0.848 for 90% consolidation. Where does
this come from? Show the calculation.
17.2. Explain why H sometimes equals the thickness of a consolidating layer and
sometimes is one-half of the layer thickness.
17.3. A large mat foundation will support a supermarket. Explain how
incorporating a layer of sand between the foundation slab and the soil
might influence the rate of settlement and the amount of differential
settlement. What precautions are necessary to ensure that the sand layer
will be effective?
17.4. Determine Cv by the Taylor method from time-settlement data for the load
increment 0.5 to 1 ton/ft2 in Table 16.1.
17.5. (a) Repeat Problem 17.4 for load increments 2 to 4 and 4 to 8 tons/ft2, and
(b) plot Cv versus the average consolidating pressure. Explain any observed
trends.
17.6. (a) Make a sketch showing a method to accelerate primary consolidation
of clay soil under a highway embankment. (b) What two kinds of
measurements can be made to indicate the progress of primary
consolidation and show when it is completed?
17.7. (a) What can be done to accelerate secondary consolidation? (b) What
measurements should be made to show progress and indicate when
settlement is completed?
17.8. Repeat Problem 17.4 using the log-time method. How does the answer
compare with that from the square root of time method? Which method do
you recommend?
17.9. In a consolidation test, the void ratio of the soil decreases from 1.239 to
1.110 when the pressure is increased from 192 to 383 kPa (2 to 4 tons/ft2). If
the coefficient of permeability of the soil at this increment of pressure is
8.4  10 mm/s, determine the coefficient of consolidation, in square meters
(feet) per year.
17.10. A foundation 6.1 by 12.2 m (20 by 40 ft) in plan is to be constructed at a
site where the geological profile is as shown in Fig. 17.17. The foundation
load is 287 kPa (3 tons/ft2) and the unit weights of various soils are as
shown in the diagram. Consolidation tests of the compressible clay indicate
that it has a void ratio of 1.680 at 95.8 kPa (1 ton/ft2) and a compression
index of 0.69. The coefficient of consolidation is 8.3  10
4 cm2/s.
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 467

Figure 17.17
Geological profile
for problem 17.10.

Calculate the total settlement, and plot the rate of settlement at the center
of the foundation, in millimeters (inches) versus years.
17.11. Assume that the material beneath the compressible clay in Problem 17.10
is impervious rock instead of gravel. Calculate and plot the rate of
settlement.
17.12. The foundation described in Problem 17.10 is to be investigated by the
stress path method, with a soil sample from the middle of the compressible
layer directly under the center of the loaded area subjected to a loading
sequence simulating that which will occur in the field. Assume that
K0 ¼ 0.4, that is, horizontal stress is 0.4 times the vertical stress before and
after the foundation load is imposed. Define vertical and lateral stress
conditions for the test.
17.13. (a) Set up a spreadsheet such as Excel or QuatroPro and use FORE to
predict final primary and secondary consolidation amounts from the 16
to 32 ton/ft2 load increment in Table 16.1. (b) Determine the regression
coefficients and plot a graph of both contributors to settlement versus time
using a logarithmic scale for time. Is there any apparent overlapping
between primary and secondary consolidation? (c) What percentage of
final consolidation can be attributed to elastic compression, and to
primary and secondary consolidation?
17.14. (a) Plot compression amounts versus time for data in the 8 to 16 tons/ft2
load increment in Table 16.1. (b) Select a representative average sample
thickness during this time interval. (c) Apply a temperature correction,
168C in the field and 218C in the laboratory. (d) Calculate thickness and
time scale factors for a soil layer 2.5 m thick with drainage in one and in
Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)
Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

468 Geotechnical Engineering

two directions. (e) Replot the data to predict settlement in the field with
both of these drainage conditions.
17.15. Using 24-hour compression data in Table 16.1, predict a minimum void
ratio attainable by increasing pressure to a maximum that can be sustained
by the soil grains without breaking. Is this amount reasonable? How may
this compare with typical void ratios in shale?
17.16. Sketch a line through the data points representing secondary consolidation
in Fig. 17.10, select four or five representative amounts of settlement, and
perform a FORE analysis to predict final settlement.
17.17. A warehouse building on linear footings on clay is stable for 20 years after
construction, and then appears to settle sufficiently to affect the operation of
forklifts. (a) Can this be explained by consolidation theory, and if so, how?
(b) What other reasons might contribute to a sudden increase in settlement?
(c) Suggest another explanation for cracking, related to soil mineralogy and
moisture content, and suggest how the true cause can be identified.
17.18. The plan for Kansai International Airport was to use 18 m of fill plus an
extra 12 m to allow for settlement, making 30 m total fill. It appears that
this will result in as much as 16.4 m of settlement. Draw a sketch showing
how much additional fill will be required in the low areas to bring the
elevation up to the originally planned grade.
17.19. In the problem of Example 17.9 the owner requests that the surcharge
load be reduced by one-half. How will this affect the time-rate of
settlement?
17.20. As an illustration of the unlimited scope of geotechnical engineering, apply
FORE to men’s world track records for 100 m and project an ultimate
minimum time.

1912 10.6 seconds Donald Lippincott, U.S.

1921 10.4 Charley ‘‘World’s Fastest Human’’ Paddock, U.S.
1930 10.3 Eddie Tolan, U.S.
1936 10.2 Jesse Owens, U.S. (Berlin Olympics. Hitler left early.)
1956 10.1 Willie Williams, U.S.
1960 10.0 Armin Hary, W. Ger.
1968 9.9 James Hines, U.S.
1991 9.86 Carl Lewis, U.S.
1994 9.85 Leroy Burrell, U.S.
1996 9.84 Donovan Bailey, U.S.
1999 9.79 Maurice Green, U.S.
2001 9.78 Tim Montgomery, U.S.
2005 9.77 Asafa Powell, Jamaica

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.
 Time-Rate of Settlement

Time-Rate of Settlement 469

References and Furthur Reading

Akai, K. and Tanaka, Y. (1999). ‘‘Settlement Behavior of an Off-shore Airport KIA.’’
In Barends et al., ed., Geotechnical Engineering for Transportation Infrastructure.
Balkema, Rotterdam.
Bjerrum, L.(1954). ‘‘Geotechnical properties of Norwegian marine clays.’’ Geotechnique, 4.
49–69.
Casagrande, A. (1936) ‘‘The Determination of the Pre-consolidation Load and its
Practical Significance.’’ Proc. 1st Int. Conf. on Soil Mechanics and Foundation
Engineering 3, 60.
Crawford, C. B. (1964). ‘‘Interpretation of the Consolidation Test.’’ ASCE J. Soil Mech.
and Foundation Eng. Div. 90(5), 85–102.
Gibbs, H. J. and Bara, J. P. (1967). ‘‘Stability Problems of Collapsible Soils.’’ ASCE J. Soil
Mech. and Foundation Eng.Div. 93(SM4), 577–594.
Handy, R. L. (2002). ‘‘First-Order Rate Equations in Geotechnical Engineering.’’ ASCE
J. Geotech. Geoenviron. Eng. 128(5), 416–425.
Holtz, R. D. and Kovacs, W. D. (1981). Introduction to Geotechnical Engineering.
Prentice-Hall, Englewood Cliffs, N. J.
Janbu, N. (1969). ‘‘The Resistance Concept Applied to Deformations of Soils.’’ Proc.
7th Int. Conf. on Soil Mechanics and Foundation Engineering 1, 191–196.
Jumikis, A. R. (1962). Soil Mechanics. D. Van Nostrand Co., Inc., Princeton, N. J.
Kansai International Airport Co., Ltd. (2001). ‘‘Kansai International Airport (KIX) Brief
on Land Settlement,’’ http://www.kiac.co.jp/english/default.htm
Kulhawy, F. H. (1991). ‘‘Drilled Shaft Foundations.’’ In H. Y. Fang, Foundation
Engineering Handbook, 2nd ed. Van Nostrand Reinhold, New York.
Lambe, T. W. (1967). ‘‘Stress Path Method.’’ ASCE J. Soil Mech. and Foundation Eng. Div.
93(SM6), 309–331.
Lawson, G. (1997). World Record Breakers in Track and Field Athletics. Human Kinetics,
Champaign, Ill.
Li, K. S. (2003). ‘‘Discussion of ‘First-Order Rate Equations in Geotechnical Engineering.’ ’’
ASCE J. Geotech. and Geoenviron. Eng. 128(8), 778–780.
Mesri, G. and Godlewski, P. M. (1977). ‘‘Time and Stress-Compressibility
Interrelationship.’’ ASCE J. Geotech. Eng. Div. 103(Gt5), 417–430.
Mitchell, J. K. (1993). Fundamentals of Soil Behavior, 2nd ed. John Wiley & Sons,
New York.
Spangler, M. G. (1968). ‘‘Further Measurements of Foundation Settlements in the
Ft. Randall Dam Embankment.’’ Highway Research Record 223, 60–62.
Taylor, D. W. (1948). Fundamentals of Soil Mechanics. John Wiley & Sons, New York.
Terzaghi, K. (1943). Theoretical Soil Mechanics. John Wiley & Sons, New York.

Downloaded from Digital Engineering Library @ McGraw-Hill (www.digitalengineeringlibrary.com)

Copyright © 2007 The McGraw-Hill Companies. All rights reserved.
Any use is subject to the Terms of Use as given at the website.