Department of Construction Engineering Advanced Soil Mechanics

Chaoyang University of Technology -- Radial Consolidation --

UNIT 4
APPLICATION OF TERZAGHI'S THEORY OF CONSOLIDATION
TO PROBLEMS INVOLVING RADIAL FLOW

Prepared by Dr. Roy E. Olson on Spring 1989

Modified by Jiunnren Lai on Fall 2002

______

4.1 Introduction
A variety of cases develop, in geotechnical engineering, where it becomes necessary to
accelerate the rate of consolidation. Several examples are:

1. A highway crosses over a river on a bridge. For some distance on each side of the bridge,
the pavement is supported on fill, which is underlain by a soft, silty clay that is quite
compressible. Thus, the approach embankments are likely to settle by significant amounts,
whereas the abutments, which are typically pile supported, will remain at essentially constant
elevation. If the fill settles after the pavement is placed, then a dangerous bump can develop
at the point where the pavement passes from the fill area to the bridge area. The solution to
this problem is typically to ensure that fill settlement is essentially complete when the
pavement is placed. Unfortunately, the one-dimensional consolidation times may last for
decades.

2. Light structures are going to be placed at a site that has soft, compressible clays. A layer of
fill will be placed first and the structure will be supported on shallow foundations in the fill.
However, anticipated settlements are substantial, and are expected to be non-uniform across
the site. The engineer decides to place an extra thickness of fill, with a pressure exceeding
the pressure applied by the structure. After consolidation is essentially complete, the extra
fill will be removed and the structure constructed. Minimal settlements of the structure are
then anticipated. This process is usually called preloading. Unfortunately, predicted
consolidation times for the preload fill are several years, making this process uneconomical.

3. Oil storage tanks are to be constructed at a site underlain by soft, compressible, clays.
Calculations indicate that rapid filling of the tanks will result in bearing capacity failures.
However, if the soil consolidates during filling, then the factor of safety against a bearing
capacity failure is quite large. Calculations indicate that it will not be possible to fill the
tank completely for several years. The owner is unable to wait that long.

The solution to many problems of these types, is to insert vertical, highly permeable, elements in
the soil so the water can drain laterally to one of these elements and then vertically to escape
from the soil system. Years ago, the process was to form holes in the soil, by driving pipes or

50
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

by drilling, and to fill the holes with sand, forming sand drains. It is now more economical to
insert pre-formed plastic drainage elements, called wicks. Hereafter, we will use the term wick
to refer to any vertical drainage element installed in the soil.

The zone of influence of a single wick can often be approximated as a cylinder of compressible
soil with an impervious outer boundary and with a cylindrical wick, with a freely draining
surface, at the vertical axis. The consolidation pressure is applied to the top of the cylinder
under conditions that often involve negligible lateral deformation; thus, deformation may be
assumed one dimensional. Water flow is three dimensional but for many problems of practical
interest radial flow dominates.

The other assumptions are the same as those made in deriving Terzaghi's one-dimensional theory,
i.e., the soil is saturated and homogeneous, the pore water is incompressible, Darcy's law is valid,
there are no secondary effects, and average soil properties may be used without significant error.

In addition to the wick problem, it is convenient to consider the analogous case of a laboratory
consolidation test in which the deformation is one dimensional and vertical, but the flow of pore
water is radial.

4.2 Differential Equation for One Dimensional Deformation with Radial

Drainage
The derivation of the differential equation governing one dimensional deformation with radial
drainage (Rendulic, 1935) follows the same sequence of operations used to develop the equation
for one dimensional consolidation. The differential element of soil undergoing consolidation is
shown in Fig. 4.1. According to Darcy's law:

k r ∂u
qr = − rdθ dz (4.1)
γ w ∂r

dr
dθ

dz

Fig. 4.1 Differential Element of Soil

51
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

The net time rate of volume change of the element, qnet, is:

∂q r k  1 ∂u ∂ 2 u 
dq net = dr = − r  + rdθ dz (4.2)
∂r γ w  r ∂r ∂r 2 

where r dθ dr dz is the volume dV. From the analysis of Unit 2, the time rate of volume change
of the element is also:

dV ∂e
dq net = − (4.3)
1 + e ∂t

Equations 4.2 and 4.3 are equated and the coefficient of compressibility for one dimensional
deformation, av, is used to convert from a change in void ratio to a change in excess pore water
pressure. Then:

∂u  1 ∂u ∂ 2 u 
= cr  + 2 (4.4)
∂t  r ∂r ∂r 

where cr is the radial coefficient of consolidation and is defined as:

k r (1 + e)
cr = (4.5)
avγw

4.3 Laboratory Consolidation Tests

The discussion of laboratory tests is included here because the mathematics is simplified
compared to cases involving central drains, and thus serves as an introduction. Details of data
reduction for radial flow tests will be included in other notes.

A radial-flow laboratory consolidation test may be performed using a disc-shaped specimen, as

for the usual one-dimensional consolidation test, but with drainage either through an outer
porous ring or through a central cylindrical drain. In this section, the case of a porous outer
boundary will be analyzed.

The flat upper and lower boundaries, which are subjected to applied load, are impervious. The
horizontal, flat boundaries, may be loaded through rigid caps that force the vertical strains to be
uniform, the socalled equal strain case, or through flexible membranes, in which case the
pressure is uniform but the strains are nonuniform, the socalled free strain case.

4.3.1 Free Strain Consolidation Tests

For loading through a flexible membrane, the differential equation (Eq. 4.4) must be solved
applying the initial condition that:

52
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

u ( r ,0) = u i (4.6a)

and the boundary conditions:

u(R, t) = 0 (4.6b)

∂u (0, t )
=0 (4.6c)
∂r

As for the one dimensional problem, it is assumed that variables can be separated so that the
excess pore water pressure depends on the multiple of two functions, F(r) and G(t), which
depend only on radius and time respectively:

u = F(r) G(t) (4.7)

Equation 4.7 is inserted into Eq. 4.4 which is then rearranged to yield:

1 G'(t) F''(r) 1 F'(r)

(c ) G(t) = F(r) + r F(r) (4.8)
r

Apparently neither side of Eq. 4.8 is a function of either time or radius. Thus each side may be
set equal to a constant, which is chosen to be -A2, and solved. From Unit 2:

G(t) = C1 exp (-A2crt) (4.9)

The term containing radius may be written:

d2F(r) 1 dF(r)
+ + A2F(r) = 0 (4.10)
dr2 r dr

Equation 4.10 is known as Bessel's equation. Its solution may be written in the form:

F(r) = C2J0 (Ar) + C3Y0 (Ar) (4.11)

where J0 and Y0 represent Bessel's functions of zero order of the first and second kind
respectively. The excess pore water is then:

u = [C4J0(Ar)+C5Y0 (Ar)]exp(-A2crt) (4.12)

The second boundary condition (Eq. 4.6c) is satisfied if C5 is zero. Thus:

u = C4J0(Ar)exp(-A2crt) (4.13)

To satisfy the first boundary condition (Eq. 4.6b), αn must be a root of:

53
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

J0 (αn) = 0 (4.14)
where:
AR = αn (4.15)

There are an infinite number of roots corresponding to values of n from one to infinity.
Tabulations of these roots may be found in handbooks of mathematical functions (Abramowitz
and Stegun, 1964). Thus:

A = αn/R (4.16)

When Eq. 4.16 is inserted into Eq. 4.13, a linear equation is obtained that has an infinite number
of solutions, one for each value of n. Thus, a general solution is obtained by summing the
individual solutions:

∞
αn r α2 c t
u = ∑ CnJ 0 ( ) exp( − n 2r ) (4.17)
n =1 R R

It is convenient to define a radial time factor, Tr, as:

cr t
Tr = (4.18)
R2

and substitute it into Eq. 4.17.

The initial condition now requires that:

∞
αn r
ui = ∑ Cn J 0 ( ) (4.19)
n =1 R

Equation 4.19 is a Fourier-Bessel series, the coefficients of which are given by:
R
αn r
∫ ru J
i 0 (
R
)dr
2u i
Cn = 0
= (4.20)
R
 α r 
2
α n J1 (α n )
∫0 r J 0 ( Rn ) dr
Thus:

∞
2u i α r
u=∑ J 0 ( n ) exp( − α 2n Tr ) (4.21)
n =1 α n J 1 ( α n ) R

The degree of consolidation at radius r is:

54
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

∞
u 1
Ur = 1 − =∑ J 0 ( α n r , R ) exp( − α 2n Tr ) (4.22)
u i n =1 α n J 1 ( α n )

The average degree of consolidation is:

R R

∫ u i 2πrdr − ∫ u 2πrdr
U= 0
R
0
(4.23)
∫ u 2πrdr
0
i

Integration yields:
∞
1
U = 1 − 4∑ exp( −α 2n Tr ) (4.24)
n =1 α n
2

The numerical relationship between Tr and U is presented in Table 4.1 and is plotted in Fig. 4.2.
Isochrones corresponding to various values of U are plotted in Fig. 4.3.

A free strain laboratory consolidation test can be performed in the same general way that a
standard one dimensional test is performed except that volume change will probably be
determined by measuring the volume of fluid pumped into or out of the consolidation cell rather
than with a dial indicator, and a different fitting method is needed for the estimation of the
coefficient of consolidation.

Table 4.1 Relationship Between the Average Degree of Consolidation and Time Factor for
Radial Flow to an Outer Freely Draining Boundary

Average Degree of Free Strain Radial Equal Strain Radial

Consolidation, % Time Factor Time Factor
0 0 0
10 0.002 0.013
20 0.008 0.028
30 0.020 0.044
40 0.038 0.064
50 0.062 0.087
60 0.096 0.114
70 0.144 0.150
80 0.215 0.201
90 0.335 0.288
95 0.455 0.375
99 0.800 0.575

55
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

-10

Average Degree of Consolidation (%)

equal strain theory
-20

-30

-40

-50

-60
-70 free strain theory
-80

-90

-100
.001 .01 .1 1
Radial Time Factor (dimensionless)

Fig 4.2 T-U Relationships for Laboratory Consolidation Tests with Radial Outwards Drainage

1.0

0.9 30
Pore Water Pressure (dimensionless)

0.8 40

0.7 60 50
0.6 70
0.5

0.4 80

0.3

0.2 90

0.1 95

0.0
0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0
r/R
Fig. 4.3 "Free Strain" Isochrones for Radial Consolidation Test. The numbers on the lines are
average degrees of consolidation.

4.3.2 Equal Strain Consolidation Tests

In most laboratory tests, a rigid loading cap is used. Thus the upper boundary condition is not
one of uniform stress, as assumed for the free strain theory, but rather one of uniform vertical
strain. Under such conditions the soil next to the porous ring begins to consolidate at once and
removes support from the loading cap, causing a redistribution of the contact stress between the
rigid cap and the soil. The contact stress decreases near the ring and increases near the vertical
axis through the center of the specimen.

56
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

The total height and radius of the soil specimen are L and R respectively and r is the variable
radial coordinate. From Darcy's law:

∂Q k ∂u
=− r 2πrL (4.25)
∂t γ w ∂r

where Q is the total volume change and the other terms are as defined previously. For equal
vertical strains, and assuming no lateral deformations within the sample:

∂Q ∂ ∂L
= − ( πr 2 L ) = − πr 2 (4.26)
∂t ∂t ∂t

Equations 4.25 and 4.26 are equated to obtain:

1 ∂u γ
= − w = C1 (4.27)
r ∂r 2k r L

where C1 varies with time but is independent of radius. Equation 4.27 is integrated for any one
instant of time and the boundary condition u(R,t) = 0 is applied to obtain:

1
u= C1 ( r 2 − R 2 ) (4.28)
2

The average value of the excess pore water pressure, um, is defined as:

∫ 2πrudr
um = 0
R
(4.29)
∫ 2πrdr
0

Equation 4.28 is substituted into Eq. 4.29 and integrated to obtain:

4u m
C1 = − (4.30)
R2

Thus:

R2 − r2
u = 2u m ( ) (4.31)
R2
Similarly:

∂Q ∂ e  πr 2 L ∂e
=−  πr 2 L  = − (4.32)
∂t ∂t  1 + e  1 + e ∂t

57
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

Equations 4.25, 4.27, 4.30, and 4.32 are combined to obtain:

∂e 8u k (1 + e)
= − 2m r (4.33)
∂t R γw

The substitution is made that:

de de
av = − = (4.34)
d σ du

to obtain:

∂u 8u
= − 2m c r (4.35)
∂t R

where cr is defined in Eq. 4.5.

Integration and insertion of the condition um(t = 0) = q yields the solution:

u m = q exp( −8Tr ) (4.36)

The excess pore water pressure is:

r2
u = 2q(1 − ) exp( −8Tr ) (4.37)
R2

and the average degree of consolidation is:

U = 1 - exp(-8Tr) (4.38)

The time-settlement curve is calculated from Eq. 4.38 converting from time factor to time (Eq.
4.18) and from the average degree of consolidation to settlement (Eq. 2.44, Unit 2). The
relationship between Tr and U is plotted in Fig. 4.2 and numerical results are given in Table 4.1.

The initial radial distribution of excess pore water pressure is found from Eq. 4.37 to be
parabolic with u ranging from zero at the contact between the soil and the ring to 2q at the
central axis of the specimen. If the ring were smooth and rigid, the initial deformation would
have to be one dimensional and the initial excess pore water pressure would have to be uniform,
not parabolic. However, as soon as consolidation begins, the contact stress between the loading
cap and the soil must be gradually redistributed and approach the distribution calculated with the
equal strain theory. As consolidation proceeds, the contact stresses continue their redistribution
and approach a uniform value as time approaches infinity.

In comparing the free strain and equal strain theories it may be noted that when applied to a
laboratory consolidation test in which a rigid, impervious cap and a porous ring is used, the free

58
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

strain theory contains the proper initial excess pore water pressure but makes no provision for
the redistribution of stress, whereas the equal strain theory redistributes the stress but contains an
unlikely initial distribution of excess pore water pressure.

Methods to be used in fitting the theory to laboratory data were presented by Trautwein, Olson
and Thomas (1981) and will not be considered here.

4.4 Drains (Wells) with Free Strain

This analysis will apply to the case of a cylindrical column of homogeneous soil containing a
coaxial column of freely draining soil of the same compressibility. Water flow is assumed to be
radial toward the drain well. No flow of water occurs across the outer surface of the soil
column. Equations describing the consolidation of the soil are obtained by solving Eq. 4.4
using the following boundary conditions:

u(rw,t) = 0 (4.39a)

∂u ( re , t )
=0 (4.39b)
∂r

and the initial condition:

u ( r ,0) = u i , rw ≤ r ≤ re (4.39c)

where rw is the radius of the drainage well and re is the external radius of the consolidating
column of soil. The analysis is too lengthy to be presented. The solutions (Barron, 1944,
1948; Moran et al., 1958) are:
∞
u 2A n Bn
=∑ exp( − α 2n N 2 Tr ) (4.40a)
q n =1 α n ( N C n − D n )
2

where:
An = Jo(αn)Y1(αn)-Yo(αn)J1(αn) (4.40b)
αnr αnr
Bn = Jo( r )Yo(αn)-Yo( r )Jo(αn) (4.40c)
w w

Cn = [Jo(αnN)Yo(αn)-Yo(αnN)Jo(αn)]2 (4.40d)
Dn = [Jo(αn)Y1(αn)-Yo(αn)J1(αn)] 2 (4.40e)
cr t
Tr = (4.40f)
r2

and:

59
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

∞
4A 2n
U = 1− ∑ exp( −α 2n N 2 Tr ) (4.41)
n =1 α n ( N − 1)( N C n − D n )
2 2 2

and:

J0, J1 are Bessel functions of the first kind of order zero and one respectively.
Y0, Y1 are Bessel functions of the second kind of order zero and one respectively,
αn are roots of the equation:

J1(αn N)Y0(αn) - Y1(αn N)J0(αn) = 0 (4.42)

and:

N = re/rw (4.43)

The Tr-U curves for a range of values of N are presented in Fig. 4.4 (Moran et al., 1958). These
curves together with Eqs. 2.44 and 4.18 are used to determine the time-settlement curves.

-10
Average Degree of Consolidation (%)

-20

-30
N=40
-40

-50
N=10
-60

-70
N=5
-80

-90

-100
.01 .1 1 10
Radial Time Factor
Fig. 4.4 T-U Relationship for Radial Inwards Drainage Calculated using the Free Strain
Theory

60
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

4.5 Drains (Wells) with Equal Strain

The geometry is the same as that just utilized for the free strain theory. In this case, however,
the assumption is made that the surface is loaded with a rigid plate so that the surface settlement
must be uniform. The analysis is similar to that used to develop the equations for consolidation
of soil in a laboratory consolidation test with rigid, impervious caps and a pervious ring except
that the drainage now occurs toward the central drain well and there is no flow across the outer
boundary of the soil. The solutions are as follows (Barron, 1944, 1948; Moran et al., 1958):

q  2 r 1  r2   2
u=  N ln( ) - ( 2)rw2-1 exp(-F Tr) (4.44)
N2Fn  rw n

and:

2
U = 1 - exp(- F Tr) (4.45)
n

where:

N2 3N2-1
Fn = ln(N) - (4.46)
N2-1 4N2

and the other terms are as defined for the free strain theory.

The time settlement curve is again defined in using Eqs. 2.44 and 4.18. The relationships
between Tr and U (Eq. 4.45), for a range of values of N, are presented in Fig. 4.5.

-10
Average Degree of Consolidation (%)

-20

-30

-40

-50 40

-60 20
-70
10
-80
-90 5

-100
.01 .1 1 10
Radial Time Factor
Fig. 4.5 T-U Relationships for Radial Consolidation to a Central Cylindrical Drain Based on
the Equal Strain Theory. Numbers shown are values of N, the drain spacing factor.

61
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

According to the equal strain theory the initial excess pore water pressure varies from zero at the
edge of the drain well to a maximum at the outer edge of the zone of influence of the well. For
N = 5, the maximum pore water pressure is about 1.2q. This distribution is substantially
different from that encountered with the free strain theory and raises the question as to which
distribution is the more likely in the field.

The actual distribution of excess pore water pressure in the field must depend on such factors as
the relative compressibilities of the wick and the surrounding soil, on the fraction of the total
surface area occupied by the drains (wicks), and on the actual rigidity of the structure applying
the load. However, from a practical point of view, the errors associated with defining
numerical values for the coefficient of radial consolidation, are likely to be considerably more
than differences between the two theories. For hand solutions, the equal strain theory is very
much easier to apply and thus has been used for essentially all of the field designs
that have been reported in the literature. Numerical analyses typically use the free-strain
approach (see later discussion).

4.6 Sand Drains with Smear Zones

Installation of wicks in the field, by driving or by static force, must disturb the soil near the
drains. The disturbance is likely to vary with distance and probably with depth; it certainly
varies with the soil type and details of the placement method. The disturbed zone is likely to
exert considerable influence on the time rate of settlement, and perhaps even on total settlement.
Any attempt to take into account the disturbance in a rigorous mathematical manner would lead
to great analytical problems and result in equations that would probably be too complex for
practical application. Further, the "exact" equations would contain soil parameters and
geometric factors that could not be evaluated for practical work. Thus a compromise must be
made to obtain tractable solutions yet ones that do not yield answers too much at variance with
field observations. In the end, the available solutions have been used only to get a crude idea of
disturbance effects and have not, to my knowledge, been used for practical design.

Barron (1944, 1948) assumed that the zone of disturbance could be approximated as a
homogeneous cylindrical shell of soil surrounding the drain well. He assumed that this soil
would consolidate so rapidly that its consolidation time could be ignored. Thus, the model he
analyzed consists of a cylindrical freely draining well, of radius rw, surrounded by a cylindrical
shell of incompressible disturbed soil, of outer radius rs and permeability ks, and then a shell of
undisturbed homogeneous soil extending to radius re.

Free strain solutions for this case were developed by Barron (1944, 1948) and discussed by
Moran et al. (1958). They are too complex for use in normal engineering design and will not be
reviewed.

Application of the assumptions of the equal strain theory to the case of sand drains with smear
zones leads to the following solutions (Moran et al., 1958):

q r r2-rs2 kr N2-S2  2Tr

u= lnr - + k ( 2 )ln(S) exp(- ) (4.47)
ζ  s 2re 2 s N  ζ

62
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

in which q is the average surface pressure and:

 N2 N S2 krN2-S2 
ζ =  2 2 ln( S ) - 0.75 + 2 + k 2 ln(S) (4.48)
 N -S 4N s N 

rs
S= r (4.49)
w

Further:

2Tr
U = 1 - exp(- ) (4.50)
ζ

The time-settlement curve is easily calculated using Eqs. 2.44, 4.18, and 4.50. Analyses show
that any significant decrease in the permeability of the smear zone, compared to the radial
permeability of the undisturbed soil, will result in a drastic increase in the time required to
achieve any specified amount of consolidation unless, of course, the thickness of the smear zone
approaches zero.

4.7 Sand Drains with Smear Zones and Well Resistance

In previous analyses it was always assumed that the coefficient of permeability of the sand in the
drainage well exceeded that of the surrounding soil by such a large amount that no consideration
need be given to the possibility of excess pore water pressures existing in the well itself. It is
important to have a theory that takes into account the finite coefficient of permeability of the
drainage well, not because the theory will be used regularly in engineering design, but rather to
establish what coefficients of permeability the drainage wells must have to ensure adequate
drainage. Barron (1944, 1948) applied the assumptions of the equal strain theory to this
problem with the added assumption that the coefficient of permeability of the soil surrounding
the drain is zero in the vertical direction. He obtained the following solutions:

 ξ  r r2-rs2 kr N2-S2   2 ξT r
u (r,z,t) = q  ln(r ) - + (k ) 2 ln(S)+1-ξ exp(- ) (4.51)
ζ ζ
2re 2 s N
s  
where:

exp[β(z-2L)]+exp(-βz)
ζ= (4.52)
1+exp(-βz)

2(N2-S2)
β= (4.53)
(kw/kr)re2ζ

63
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

N2 N 3 S2 kr N2-S2 
ζ= ln( S ) - 4 + + k  2 ln(S) (4.54)
N2-S2 4N2 s  N 

and L is the thickness of the homogeneous compressible stratum containing the drains. The
symbol u(r,z,t) is used to denote specifically that the excess pore water pressure varies with
radius, depth, and time.

The average degree of consolidation is found by integrating the excess pore water pressures as
follows:

L
1
U = 1 - qL ⌠
⌡u(z,t)dz (4.55)
0
where:
2ξTr
u (z,t) = q exp(- )(56)
ζ

Equation 4.55 is integrated graphically.

To demonstrate the effects of well resistance, Eq. 4.55 was solved for the case of no smear zone
and N = 10 for a range of values of L/rw,kw, and Tr. Representative data are shown in Table
4.2. If the compressible layer is very thin, e.g., L/rw = 10,

Table 4.2. Influence of Well Resistance on the Time Rate of Consolidation when S=1 and N=10

L/rw= 10 50 100
Tr .1 1 6 .1 1 6 .1 1 6
kw/kr=102 9 60 99 2 17 46 1 8 23
103 12 71 100 6 48 96 4 27 71
104 12 72 100 11 69 100 9 60 99
105 12 72 100 12 72 100 12 71 100
inf. 12 72 100 12 72 100 12 72 100

satisfactory drainage is achieved if the drain well has a hydraulic conductivity of the order of 102
to 103 times that of the soil. As the compressible layer becomes thicker there is a larger
influence of well resistance but a ratio kw/kr of 105 would appear to preclude any significant
well resistance for the range of values of L/rw normally encountered in the field. If small-
diameter drains are used in deep deposits it appears that care should be taken to avoid significant
amounts of well resistance.

Since smear zones tend to reduce the inflow into the wells, the existence of smear zones makes
well resistance less of a problem.

64
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

If the field conditions correspond to the free strain assumptions, well resistance will be more of a
problem because of the higher rate of consolidation at low values of the time factor. However,
such well resistance would appear to be compensated for by the common use of the equal strain
analysis.

4.8 Time Dependant Loading

Wicks are often used in cases where it is important to have a major part of consolidation
completed during or shortly after construction. Thus, the construction time represents a major
part of the time during which settlements are to be calculated and the assumption of
instantaneous loading at time zero leads to significant errors.

The method of analysis used in Unit 3 can be applied to the case of time dependent loading of
sites provided with sand drains. The loading diagram is assumed to be that shown in Fig. 3.4
(Unit 3). The equations for the excess pore water pressures obtained using the equal strain
theory, Eqs. 4.44, 4.47, and 4.51, can all be put into the form:

u = Bq exp(-ATr) (4.57)

Application of the procedures of Unit 3 then leads to the solutions:

Bqc
ν≤1 u (r,t) = AT [1-exp(-ATr)] (4.58a)
rc

Bqc
ν>1 u (r,t) = AT [exp(ATrc) - 1]exp(-ATr) (4.58b)
rc

where:
qc = applied average surface pressure at the end of construction,
Trc = crtc/r2 and tc is the construction time (Fig. 3.4, Unit 3).

Similarly, the equations for the average degree of consolidation (Eqs. 4.45 and 4.50) can be put
into the form:

U = 1 - exp(-ATr) (4.59)

and applied to the time dependent loading to obtain:

1 1 
ν ≤ 1 U = T Tr - A[1 - exp(-ATr)] (4.60a)
rc 

1
ν>1 U = 1 - AT [exp(ATrc) - 1] exp(-ATr) (4.60b)
rc

65
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

Solutions are easily obtained by evaluating the factors A and B from the equations and charts
presented previously. These solutions, and the resulting T-U curves, were published by Olson
(1977).

4.9 Combined Vertical and Radial Flow

Cases in the field that involve radial flow also involve vertical flow. The differential equation
for combined vertical and radial flow, for instantaneous loading, is:

∂u ∂2u 1 ∂u ∂ 2 u
= cv 2 + cr ( + ) (4.61)
∂t ∂z r ∂r ∂r 2

Carrillo (1942) demonstrated that variables in Eq. 4.61 could be separated in such a way that
the degree of consolidation for purely vertical flow (Uv) and purely radial flow (Ur) could be
calculated as if there was only vertical or radial flow, respectively, and then the solutions
combined in a simple equation:

U = 1 - (1-Uv)(1-Ur) (4.62)

where U is now the degree of consolidation for combined vertical and radial flow. By
calculating values for Uv and Ur for effects of time dependent loading, we can now solve cases
with time dependent loading and combined vertical and radial flow.

4.10 References
Abramowitz, M. and I. A. Stegun (1964), Handbook of Mathematical Functions, National
Bureau of Standards, Applied Mathematics Series 55 1046 pp.

Barron, R. A. (1944), "The Influence of Drain Wells on the Consolidation of Fine-Grained

Soils," U.S.Army Crops of Engrs. Dist., Providence.

Barron, R. A. (1948), "Consolidation of Fine-Grained Soils by Drain Wells," Trans., ASCE, Vol.
113, pp. 718-754.

Carrillo, N. (1942), "Simple Two- and Three-Dimensional Cases in the Theory of Consolidation
of Soils", Jour. of Math. and Physics, Vol. 21, No. 1, pp. 1-5

McKinlay, D. G. (1961), "A Laboratory Study of Rates of Consolidation in Clays with Particular
Reference to Radial Porewater Drainage," Proc., Fifth Intern. Conf. on Soil Mech. and
Found. Engr., Vol. 13, pp. 98-111.

Moran, Proctor, Meuser, and Rutledge (1958), "Study of Deep Soil Stabilization by Vertical
Sand Drains," Publ. PB151692, U.S.Dept. of Commerce.

66
Department of Construction Engineering Advanced Soil Mechanics
Chaoyang University of Technology -- Radial Consolidation --

Olson, R. E. (1977), "Consolidation under Time-Dependent Loading," Jour., Geotechnical

Engineering Div., ASCE, Vol. 103, pp. 55-60.

Rendulic, L. (1935), "Der hydrodynamische spannungsausgleich in zentral entwasserten

tanzylindern," Wasser-wirtschaft, . 2, pp. 250-253.

Trautwein, S. J., R. E. Olson, and R. L. Thomas (1981), "Radial Flow Consolidation Testing,"
Proc., Tenth Intern. Conf. on Soil Mech. and Found. Engr., Stockholm, Vol. 1, pp. 811-
814.

67