GEO-04 - Confined Compression and Consolidation - GDS

-
GEOTECHNICS
LESSON 4
Confined compression and consolidation
Eng. Giorgia Dalla Santa

Course of GEOTECHNICS, A.Y. 2020-21 Lesson 4: Confined compression and consolidation 1
-
STRAINS IN SOILS
(Lagrangian approach)

Strain state – Lagrangian approach
-
When on the lateral sides of an infinitesimal volume of soil (with size x, y

and z) only normal forces act (i.e. normal stress without shear stress), the
volume reduce itself with a movement of the side normally to the stress
(only volume variation without distortion of the cube).
We indicate with x, y, z the displacements of each side parallel to the
reference axis x, y, z.
We can define longitudinal strains the ratios: z
x
x y z z
x  y  z  y x
x y z z
y
x z
These strains are positive x

if the volume reduces y
y
Strain state definition
-
When on the lateral sides of an infinitesimal volume of soil only shear forces
act (i.e. shear stress without normal stress), the body deforms changing its
form without change in volume.
We indicate with x, y, z the displacements of each side parallel to the
reference axis x, y, z, as it is shown in Figure.
We can define angular strains the ratios: z
x x x
 zx  arctan 
z z y x
zx
y
 xy 
x For the hypothesis
of small strains zx z
z
 yz  x
y
These strains are positive if they
induce a clockwise distortion. y
Plane strain condition
-
When the deformation along one direction is zero for the presence of lateral
constrains the conditions are known as plane strain condition.
The strains may be also defined as the gradient of displacements. The
definition are:
u
x 
x
v
y 
y
u v
 xy    
y x

Strain tensor
-
All the strain components for the strain tensor… Also in this case the tensor
is a double symmetrical.
As for the stress tensor, a principal reference system can be found. In this
reference system the distortional components are zero.
 1 1 
 x  xy  xz 
1
2 2  1 0 0
1 
ε    yx y  yz  ε   0  2 0 
2 2 
1   0 0  3 
 z 
1
 2 zx  zy
2 
The first invariant strain is the volumetric strain, defined as:
 v   x   y   z  1   2   3

-
ANALYTIC TOOLS FOR THE

MECHANICS OF SOILS
(continuum approach)

Field equations
-
The typical problem that geotechnical engineers have to solve is the

prediction (and control) of displacements due to a loads applied on the
ground surface.
The problem is similar to the calculation of inflection of a beam subjected to
a general loads.
The solution of this problem requires to:
1. know the previous stress distribution inside the soil mass, i.e. the stress
distribution due to the self-weight;
2. know the undisturbed physical state of the soil, i.e. the void ratio
distribution;
3. calculate the stress increment distribution due to load application;
4. calculate the strain increment distribution;
5. calculate the displacement as integral of strains.
In the steps 1-5 we have to consider the presence of water and its behavior
(coupled/uncoupled response, also called undrained/drained behaviour)

-
Field equations
Any evaluation of stress and strain distribution in the initial condition
(undisturbed state of soil) or subsequently to the load application has to be
obtained solving a mathematical problem (i.e. integrating some differential
equations):
The available analytical tools are:
1. Cauchy equations;
2. Equations of strain compatibility;
3. Stress-strain relationship;
4. Boundary conditions.
The rigorous solution of a particular problem requires that the equations of

equilibrium and compatibility are both satisfied for the given boundary
conditions and in relation with the most appropriate stress–strain
relationship describing the soil behaviour.

-
Strain
Equations ofvector
equilibrium
The Cauchy equations referred to an infinitesimal volume dx∙dy∙dz imposes
the condition:
Then principle of effective stress is introduced:
 x  yx  zx  x  yx  zx u
  0    0
x y z x y z x
 xy  y  zy  xy  y  zy u
  0    0
x y z x y z y
 xz  yz  z  xz  yz  ' z u
    0      sat  0
x y z x y z z
 xy   yx ,  xz   zx ,  yz   zy
-
EquationsStrain
of strain compatibility
vector
The equations of strain compatibility rule the spatial variation of strains in order to
respect the continuity of soil mass: no cracks or separation of the soil mass.
The general form of strain compatibility equations is:
 2 ij  2 hk  2 ih  2 jk
   0
xhxk xi x j xk x j xhxi
Varying the subscript in the above equations, we could obtains only 6 independent
conditions, known as the De Saint Venant equations.
The equilibrium equations together with the equations of strain compatibility

represent a system of equations that can be solved in accordance with the
boundary conditions only if we introduce the relationship between the stresses
and strains (stress-strain relationships or constitutive model of the continuum
medium).

-
Strain
An example: vector
elastic rod in traction
If we consider an elastic rod subjected to an axial traction, the calculation of stress
and strain distribution requires to write:
P
1. Equilibrium: F 0
x
u
2. Strain – displacement x 
relationship (strain compatibility): x
x
3. Constitutive model: x 
E
We have:
u
x = reference axis P   x A  E x A  EA
L = rod length x
A = rod section Differential equation that
u = displacement along x-axis  u
2
has to be integrated in
P = axial force applied to the rod EA F 0 accordance with boundary
at a generic section x x 2
conditions to determine
F = force for unit of volume (in displacements, strains and
this case … force for unit of length) stresses.
-
SOIL CONSTITUTIVE
MODEL

-
IsotropicStrain
Linearvector
Elastic model
The most simple model utilized in solid mechanics is the Linear Elastic model for an
isotropic material. The stress-strain relationships is:
 x   1   0 0 0   x 
     

 y 1  0 0 0   y 
 
  z  1    1 0 0 0   z 
ε  Cσ     

 xy  E  0 0 0 21    0 0  xy 

  yz   0 0 0 0 21    0  yz 
   0 21     zx 
 zx  0 0 0 0
C = Compliance Tensor 
E = Modulus of Longitudinal Elasticity or
Young’s Modulus
E
 = Poisson’s ratio
G = E/[2(1+)] = Shear Stiffness Modulus 1
K = E/[3(1-2)] = Volumetric or Bulk Modulus 
IsotropicStrain
Linearvector
Elastic model
-
The relationships could be also written in the inverse form:
σ  Dε
   
 1 0 0 0 
1  1 
   
x   1 0 0 0   
  1   1   
x

 y   
 1 0 0 0  y 
z  E 1    1   1     z 
      
 xy  1   1  2   0
 1 2 
0 0 0 0  xy 
 yz   21      yz 
   0 1  2   
 zx  
0 0 0
21   
0
 zx 
 1  2 
 0 0 0 0 0 
 21   
D = Stiffness Tensor determined as the inverse tensor of compliance= C-1

ReaIStrain vectorof soil
behaviour
-
In general the behavior of soil is:

1. not linear;
2. not elastic… the soil exhibits plastic (irreversible) strains also for very small
stress increments;
3. Density dependent (the behavior depends from the void ratio)
4. Stress dependent (the confinement stress controls the response)
A correct constitutive model ought to be composed by stress-strain relationships

written in terms of incremental stress and strain and taking into account effect of
density and confinement stress …
 σ' 
  C C  C σ', ,ε 
p
 σ' 
 

ReaIStrain vectorof soil
behaviour
-
The major two constitutive models that we can use in geotechnics are relative to
the soil behavior in a stress-strain condition:
• very far from failure or Serviceability Limit State (SLS)
• at failure or Ultimate Limit State (ULS).
In the first case we use a linear elastic model (Elasticity theory), in the second we
neglect the strains before failure and we consider a model of a rigid – perfectly
plastic material (Plasticity theory).
We usually analyze the  Rigid – Perfectly Linear Elastic –

Plastic Model Perfectly Plastic Model
problems SLS and ULS types
(ULS) (SLS)
in separate way and with
different theoretical analysis.
Some advanced constitutive
models implemented in
Finite Element Model are
able to analyze the overall
behavior at the same time. 
-
APPLICATION OF
ELASTICITY THEORY
TO THE CALCULATION OF THE
AT-REST HORIZONTAL STRESS
IN SOIL

-
Strain stress
At-rest horizontal vectorfor elastic soil
For a deposit composed by saturated, homogeneous, isotropic and linear elastic
soil formed by progressive deposition of layer …
assuming as x, y, z the Cartesian axis with z-axis oriented downward …
the z-axis is an axis of axial-symmetry, i.e. the vertical stress is the major principal
stress and on the horizontal plane the x and y-axes are principal axes
independently from their orientation (namely xz=zy=yx=0)
 x  y  z u
From equilibrium balance  0;  0;    sat  0
we have: x y z z
From the strain definition the u v w

  x  0;   y  0;  z  0
horizontal strains are 0. Only x y z
the vertical one is ≠ 0:
 xy   xz   zy  0
Substituting the strain definitions in the compatibility equations, the latter are all satisfied but
do not provide any tool for solving the problem.
The last equilibrium equation permits to calculate the vertical distribution of total and effective
stresses as already explained in Lesson 3.
-
Strain stress
At-rest horizontal vectorfor elastic soil
The constitutive model reduces to: 0  1      ' x 
  1   
 0     1     ' y 
   E    1    ' 
 z   z 
From which, imposing ’x = ’y we have:  
 x   y   z   z
1  1 
We define at-rest earth pressure coefficient K0 as the ratio between horizontal and vertical
effective stress in this condition:  x 
K0  
 z 1 
On the other hand, experimental measures of the horizontal stress in laboratory and in-situ
tests, it results:

K 0  1  sin f OCR m 
being f and OCR parameters defined in the next lessons

-
CONFINED COMPRESSIBILITY
AND CONSOLIDATION

-
Strain vector
Introduction
When a soil deposit is loaded, deformations will occur. The total vertical
movement at the surface resulting from the load, generally oriented
downward, is called settlement.
On the contrary, when a soil deposit is unloaded (excavation) the movement
of ground surface is upward and we speak of swelling.
In general, when a soil is loaded, it will compress because of:

1. compression of air and water in the voids, and/or
2. squeezing out of water and air from the voids.
At typical engineering loads soil grains and pore fluid can be considered
incompressible. Consequently the compressibility of soil is due to:
1. If the soil is saturated, changes in volume due to the flow of water.
2. If the soil is partially saturated, change in volume partly due to the
compression and flow of air from the voids.

-
Strain vector
Introduction
lf it is assumed that the soil is completely saturated the change in volume can
occur only if water is forced out of the voids.
As pore fluid squeezed out:
1. Soil grain rearrange themselves in a more stable and denser configuration
2. Decrease in volume and surface settlements are observed
How much rearrangement and compression?
Depend on void ratio and nature of the soil:
compression is higher in soils with higher void ratio (generally fine soils have
a compressibility 10-100 times that of granular soils)
How fast this occurs?
Depend on the permeability of soil
1. Compression of coarse soils occurs instantly because they have high
permeability
2. Compression of fine soils is very time dependent process: we call
consolidation the slow water squeezing process, coupled with
deformation occurrence.
One-dimension Strain vector
compression and consolidation
-
For the majority of practical settlement problems, it is sufficient to consider that both
seepage and strains take place in one direction only, this usually being vertical.
One‐dimensional consolidation specifically occurs when there is no lateral strain (an
example is given by the case sketched in Figure).
z=q
H0
z0 and x=y=0

In this condition we can assume:
1. The load q applied at surface becomes an increment of the total vertical stress
z.
2. Compressibility of sand is negligible respect that of clay.
3. The load surface is much larger than the thickness of compressible layer.
4. No lateral strains occur, only vertical ones.
Strain
Pore pressure vector
during consolidation
-
 At time t0 = 0
1= z u = 1

’= 0
H0
Increment of pore water

pressure = Excess pore
water pressure , u
1= z
z induces a vertical compression on the solid skeleton: since the solid grains and pore
fluid are incompressible, the soil can be compressed only if the fluid flow out through the
upper and lower porous layers.
But the water can not exit instantaneously, so, at t=0, it increases its pressure (u) exactly
of the same quantity applied at the surface = increment of total stress: u=z.
No increment of effective stress occurs (’z=0).
-
Strain
Pore pressure vector
during consolidation
 At time t ≠ 0
1= z
’+ u = 1
u ’
Excess pore water

pressure, u
1= z
Low permeability delays water drainage and the resulting compression. Gradually, the
excess pore water pressure decreases (before close to the boundary and then in the inner
part of sample): consequently the stress increment supported by water moves to the soil
skeleton and the effective stress increases up to the equalize the increment in total stress.
This process, called “consolidation”, can be reproduced in laboratory through the so‐called
oedometric test, or one-dimensional compression test or confined compression test.

Strain vector
Oedometer test
-
The one‐dimensional consolidation testing procedure was first suggested by

Terzaghi.
It is a vertical compression test without lateral expansion.
It consists of a progressively increasing load applied step by step on a
saturated cylindrical specimen, whose lateral sides cannot expand, leading to
vertical‐only deformations and water fluxes.
At each load, consolidation is achieved (the strain rate becomes almost zero
and the excess pore water pressure equal to zero).
This test is widely used to estimate compressibility and consolidation
parameters and to determine the stress history of a cohesive deposit.

-
Straintest
Oedometer vector
– The cell
D
Oedometric cell
Loading head
Water
bath Rigid
H confining
ring
Porous stone
Loading head Porous stone
The oedometric cell is constituted by:
Rigid confining ring with sample inside (Hmin= 19 mm, Dmin= 50 mm);
Porous stones above and below the sample;
Loading head in metal;
Lateral to the ring, the cell is filled with water in order to keep saturated the sample
during all the test;
Dial gauge is used for recording the time-evolution of the vertical settlement of the
loading head, i.e. the settlement of sample.
-
Strain
Oedometer test –vector
The procedure
Since the sample is set in the oedometric cell, the latter Loading system with dead weight
is installed in the loading system.
The loads are applied vertically thank to the dead-
weight and a lever.
The load sequence is: N1, N2, .. Ni with
- Ni+1/Ni = 2 during loading (load increase)
- Ni/Ni+1 = 4 during unloading Dial
gauge
Each load Ni is mantained for 24 h at least, measuring
the settlement H evolution with readings at increasing Dead
time steps t. N weight
The data (H, t) are used to draw the consolidation
curve and evaluate the relative parameters
(consolidation coefficients).
We indicate with Hi the settlement corresponding at
24h to the general load Ni . The data (Ni, Hi) are used to
draw the 1D compression curve and evaluate the
relative parameters (compression parameters).
Initial sample sizes, initial and final weight and dried
weight are measured to evaluate the initial void e0.
-
Strain
Oedometer test –vector
The procedure

Oedometer Strain vector
test – Data interpretation
-
Given that:
1. Water is incompressible
2. Soil grains compressibility is negligible
3. The specimen is saturated
The specimen’s volume decrease is equal to the drained water volume.
Saturated soils’ volume can change in a significant way only due to the particles
re‐arrangement with interstitial water loss and decreasing of void ratio.
Vertical strain = Volumetric strain because the cross-section is constant
H V V0  V1 Vs 1  e0   Vs 1  e1  e0  e1
z      H e
H 0 V0 V0 Vs 1  e0  1  e0 z  
H 0 1  e0
e  e0  (1  e0 ) z
If e0 is known, from the

measured H we calculate
the current void ratio e

-
Strain
Oedometer test – 1Dvector
compression curve
For the relation between e and z the data may be plotted alternatively in plane:
z-’z or z-log’z
e-’z or e-log’z
e-’z e-log’z
We define:
AB = First reload to
Void ratio e [-]

Void ratio e [-]
stress already
experimented by soil
BC = New or virgin
load
CD = Unload
DF = Reload F F
’z ’z
The passage from the natural scale to the semi-logarithmic scale permits to enlarge the curve at
small stresses, well observing the behavior in this range.
The curved lines (for instance curve BC) becomes almost straight segments in semi-log scale.

Oedometer testStrain vector
– Preconsolidation pressure
-
Considering the compression curves in semi-

log scale, we recognized the passage from
reload AB to the virgin load BC using a
Point of Horizontal geometrical construction:
maximum 1. Individuate the point of maximum curvature P;
curvature P Bisector 2. Drawn the horizontal and tangent lines in P;
3. Drawn the bisector between the horizontal
Tangent and tangent lines;
4. The intersection of bisector with the tangent
line to the virgin segment indicated the
preconsolidation pressure ’p.
’zo ’p
’p is the maximum effective vertical
 'p stress under which the soil could
Over-consolidation ratio is: OCR  consolidate in its loading history.
 'z0 Causes of over-consolidation may be:
• Over-consolidated soil: OCR > 1 erosion, ageing, natural cementation,
• Normal consolidated soil: OCR = 1 drying ...
• Under-consolidated soil: OCR < 1
Oedometer testStrain vector
– Preconsolidation pressure
-
Over-consolidation condition is related to the previous load history. In the case of a layer
formed for gradual deposition of material the soil is normal-consolidated with a pre-
consolidation pressure equal to the current effective vertical stress due to the weight of
above material (point B).
If for geological events the above material is subsequently eroded the effective vertical stress
decreases, while the pre-consolidation pressure remains constant. The OCR increases:
Normal
Deposition Erosion consolidated
Void ratio e
Over-consolidated
Vertical effective stress ’z

OedometerStrain vector
test – Compression index
-
 e
e Recompression index: Cr  
Cr   log  z '
e
Compression index: Cc  
 log  z '
 Cc

Generally a soil has Cc/Cr ≈ 10
Cr
Similar ratio are defined also for the z-
z’ (kPa) log’z (i.e. the recompression and the
compression ratios RR and RC) but are
 Reload less used.
 Preconsolidation pressure
 Virgin load
 Unload-reload loop
OedometerStrain vector
test – Compression index
-
e OCR=1 e ’z OCR>1

Cr ’z
Cr
Cc Cc
Cr Cr
’p’zo  z’ ’zo ’p  z’
Given the vertical strain definition we can calculate the settlement in different way,
depending from OCR and the vertical stress increment
H0  'z0   ' z If OCR=1
Cc log
e 1  e0  'z0
H  H 0 
1  e0 H 0   'p  'z0   ' z  If OCR>1
Cr log  Cc log and  'z0   ' z   'p

1  e0   'z0  'p 

-
Strain
Oedometer test – 1Dvector
compression curve
Considering the compression curves in natural
scale, we see:
e-’z
1. During unloading phase the strain is only
partially recovered, i.e. the strain is the sum
of two components, one elastic and one
 z ' plastic.
Void ratio e [-]
2. Deformability of soil, described by the slope

e of the curve, reduces with stress level.
We define the following parameters, all not

constant but stress-dependent:
Compressibility ratio:
’z
mv   z  z '
Oedometric or confined modulus:
M  1 / mv
Compressibility index:
av   e  z '
-
ONE-DIMENSIONAL
CONSOLIDATION THEORY

Strain vector
1D consolidation theory
-
At t=0 (supposing the load applied instantaneously) the pore water pressure increases
equal to the increment of total vertical stress. We indicate now as:
• uh = hydrostatic pore water pressure existing before loading
• ue = excess pore water pressure due to loading
At t=0 we have: ue = ui =  (blue line) constant in the thickness
At t>0 the excess pore pressure gradually decreases (dotted line)
The process of pore water pressure dissipation is analysed by the Terzaghi’s theory

ue at t=0  ui
Sand
ui
ue at
2d Clay general
uh time t
Sand Pore pressure, ui+u (t/m2)

-
Strain vector
Terzaghi’s theory studies the Primary consolidation of a saturated soil mass

subjected to a static constant load. It is based on the following assumptions:
1. Soil is laterally confined: consequently strains and water fluxes are in one direction
only (one‐dimensional consolidation).
2. The soil is homogeneous, isotropic and fully saturated.
3. Water and solid grains are incompressible.
4. Darcy’s law is valid throughout the consolidation process and the permeability
coefficient remains constant during the process.
5. Strains are small and the relationship void ratio - effective stress is linear with the
deformability parameter constant during the process.
6. No viscous strains occur during the pore water pressure dissipation.
Note that, we call Primary Consolidation the process of pore pressure dissipation
coupled to settlement occurrence.
There is also a Secondary Consolidation, due to viscous deformations of soil occurring at
constant effective stress. We call secondary because, in practical problems, we assume it
occurs only after the complete pore pressure dissipation (i.e. after the end of Primary
Consolidation), even if in some cases this assumption does not well reproduce the realty.

-
v x v y vz
  0
x y z
Balance of mass in a porous medium (ev=volumetric strain):
v x v y vz  v
  
x y z t
In a porous medium, for a one-dimensional processe occurring along z axis:
vz  z

z t
-
Darcy law for a seepage flow in vertical direction z:

h
v z  k z
z
Introducing the above equation into the mass balance:
 h  v
2
k 2 
z t
Substituting the value of h=(u/w-z) and cosidering that the variable is the excess
pore pressure ue (the derivative of hydrostatic water pressure is zero):
k  2ue  v
 
 w z 2
t
-
Using constitutive relationship for one-dimensional compression:
 z  mv ' z
k  2ue  ' z
  mv
 w z 2
t
Taking into account the priniciple of effective stress in z direction (’z=z-u):
=0
k 2ue   z ue 
   
mv w z 2
 t t 
Ipothesizing that vertical total stress is constant over time:

Strain vector
1D consolidation equation
-
We get the equation of the 1D consolidation, in this form:
Sand
 2ue ue z
cv 2  2d dz
dx
Clay
z t
Sand
k
being cv = coefficient of primary consolidation [L2t-1] cv 
mv   w
The equation relates 3 quantities:
1. excess pore water pressure ue
2. depth z inside the upper limit of layer (see the figure)
3. time t cured after the load application
It can be analytically integrated imposing the initial and boundary conditions.

1D consolidation equation
-
The solution of the equation depends on:

 drainage conditions;
 initial distribution of excess pore pressure [u0]
Boundary conditions
ue  ui for 0  z  2d e per t  0
ue  0 for z  0 and z  2d and for t  0

-
Strain vector
1D consolidation solution
In the case of draining layers above and below the clay layer (double drainage)
and a uniform distribution of the initial pore water pressure inside the clay
layer, the solution of Terzaghi equation is:
 
m
2ui  Mz 
ue ( z, t )    sin  exp  M 2
Tv
m 1 M  d 
Being:
- d = the drainage path (half of layer thickness in this case)
- m and M = two counters with M=p(2m+1)/2
- Tv = dimensionless time factor
cv t
Tv 
d2
We call isochrones the curves

describing the ue distribution
inside the layer at a fixed t (see the figure).

Strain
Degree vector
of consolidation
-
We define Degree of Consolidation the dimensionless quantity:
 
m
ui  ue ( z, t ) 2 Mz 
U z ( z, t )   1   sin  exp  M Tv
2
ui m 1 M  d 
Note that ui is ue at time t=0

Tv =0
Tv =0.05
Note that the Degree of
Consolidation depends on the
depth z
It may be reported in a
dimensionless plot independent
from the real case
Degree of consolidation Uz

-
Strainofvector
Mean degree consolidation
We define Mean degree of Consolidation the dimensionless quantity:
 
2d m
1 2
U
2d  U z ( z, t )dz  1   M 2
exp  M 2
Tv
0 m 1
This ratio varies in the range [0, 1] or [0, 100%] depending if expressed as
natural or percent number.
For the linear relation between strain and stress variation, it may be
demonstrated that:
s (t ) Settlement at time t
U 
sc Settlement at the end of consolidation
There are also two approximate equations for U:

p
Tv  U 2 for U  0.60
4
Tv  0.933 log(1  U )  0.085 for U  0.60
-
Strainofvector
Mean degree consolidation
The Mean degree of Consolidation is a function of time factor only. It may be
plotted in natural or semi-log scales. Here the theoretical trend from Terzaghi
solution.
Tv
U (%)
Tv
U (%)

Average Strain
degree vector
of consolidation
-
The theoretical trend of Average degree of Consolidation differs from the trend
observed in a real consolidation phase (for instance one determined in an
oedometric test) for the occurrence of secondary consolidation.
so so = initial settlement
Settlement
s100-so =
Experimental curve settlement for
primary
consolidation
Terzaghi curve
s100
s
log t s = settlement for

secondary consolidation
The time of the end of primary consolidation tp is found as the point of

intersection between the lines tangent to the final linear part of plot and to the
point of inflection.
Coefficients of Primary and Secondary
-
Strain vector
consolidation
From the experimental curve obtained from a consolidation phase in an
oedometer test, the coefficients of primary and secondary consolidation may be
obtained.
The coefficient of primary consolidation cv is obtained by the Casagrande or the
Taylor graphical constructions (see the lesson), which give values a little be
different (generally cv from Taylor method has a value 1,5-2 times the cv from
Casagrande).
The slope of the line tangent to the final linear part is defined as the coefficient
of secondary consolidation. It may be defined in term of variation of void ratio
(c) or vertical strain (c,):
e  z c
c  c ,  
 log t  log t 1  e0
For inorganic soils, it is: c  0.04  0.01  Cc

-
Methods to accelerate
consolidation

-
Preload
Preload is a load temporary imposed in an area for consolidating the
foundation soil before realizing the real structure. The technique is used
to:
- Case A: reduce the time of consolidation;
- Case B: reduce the settlements of the future structure .

Preload
-
CASE A: reduction of time of consolidation

Being Sc and tc the consolidation settlement and time for an embankment with a
general height equal to H1, we know that if we have to embankments with H2>H1
the consolidation time is the same (tc2 = tc1) while the settlements are Sc2>Sc1.
Height embankment
But, on the red curve, the settlement
H rilevato
Sc1 will be reached at a time t’ < tc1,
corresponding at a consolidation
degree equal to the ratio of settlements H2
Sc1 and Sc2: H1
t 'tc*tc1 tc1 tctc 2
S c1
U 1 tempo
time
Sc 2 Cedimento
Settlement
Hence, if we construct the embankment S(t) of emban. H1 high

H2 high, after a time t’ we reach the SSc1c
required settlement and we could
SScc2* S(t) of emban. H2 high
removed the upper part H2-H1.
Preload
-
CASE B: reduction of settlements of a structure

Suppose that the consolidation settlement of a structure is that corresponding to
the path A  B and this settlement is not admissible for the structure. We may
realize an embankment (preload) applying the same load on the soil, wait for the
settlement occurrence and then remove it (path A  B  A’). The settlements
developed by the future structure will correspond only to the reload component
(path A’  B).
v 'vo 'vf
Moreover, if the vertical stress
applied in the preload phase is A Primary
Cedimento di
greater then the final stress A’ settlement
consolidazione
B
(point D) we are able to develop, E Secondary
Cedimento
C D settlement
secondario
with the preload, also a part of
the secondary settlement ABC:
ABC:strain due con
percorso to the building
opera
(segment BC) that the structure ABD:
ABD:strain duecon
percorso to preload
precarico
DE:
DE:unloading
rimozioneofprecarico
preload
could develop in the future.
EC: reload due to the building
EC: costruzione opera
log'v
-
Vertical drains

Vertical drains
-
Vertical drains are high permeability elements which across the layer to be
consolidated and permit to reduce the consolidation time thank to:
- A reduction of drainage path from the semi-thickness of the layer to the semi-
interspace between two adjacent vertical drains;
- The water flowing along horizontal paths having higher permeability (kh>kv).
Vertical drains are often associated to preload or phase construction.
In many important projects

a monitoring system may
be useful to control the
evolution of settlements.
Typical distribution of a vertical

drain system and a monitoring
system (Rixner et al., 1986)
Sand drains
-
 The sand drains are columns of sand, with a

diameter in the range 7-80 cm, embedded in the
(cm/s)
Permeability
ground with different techniques (see the next
slides and the stone columns section).
 The material must be a clean granular material with
a low fine content (passing to #200 <5%), to have SILT
high permeability (as shown in the lateral plot, a
content of 5% of clay or silt could reduce the
permeability of 1-2 orders of magnitude). In the CLAY
table the optimal grain-size distribution is
suggested.
Passing at #200
Optimal Grain-size (%)
distribution  At the base of the embankment, a drainage blanket,
Passing constituted by clean sand, is generally realized
Seeve n.
Fraction (%)
before installing the drains (not used with levees). If
4 85 - 100
16 40 - 85 after the drain installation, due to the perforation
50 5 - 30 muds, the blanket is dirty, it must be substituted in
100 2 - 10 order to guarantee the drainage at the embankment
200 0-3 base.
Sand drains
-
The sand drains may be realized with various systems (see the stone columns
section for more details). In the table the typical diameters are listed.
Usual Interspace
Istallation Advantages Disavantages
diameter (m) (m)
Vibroinfision Great disturbance

Cheap (very used
with close end 0,40 - 0,80 1,50 - 6,00 (care in sensible
in the past)
casing soils)
No disturbance,
Screw auger 0,40 - 0,80 1,50 - 6,00 good for crossing Smear effect
consistent soils
No smear effect, High cost and need

Jetting 0,20 - 0,30 1,50 - 5,00
no disturbance of carefull controlls
Fast installation,
Sandwick 0,06 - 0,075 1,2 - 3,60 Smear effect
low disturbance

Prefabricated vertical drains
-
These drains have a band form, are supplied on roll, fixed to a small plate and
pushed inside the soil by a hollow rod (also called “mandrel”)
Drain end
Drain
Mandrel
folded and
passed in the
plate eyelet
Fixing
plate
1) Drain set up 2) Pushing in the 3) Mandrel

soil removing
The strip has a rectangular section, but in the
Cross-
calculation a circular section is assumed with an
section
equivalent well diameter dw given by the relation:
2a  b 
DRAIN FILTER dw 
(paper or
plastic strip)
(paper or
geosynthetic) p
MANDREL
Prefabricated vertical drains
-
Many types of prefabricated drains are produced in the world. In the table the most important and used.
The main feature of

pre-made drains is
the presence of
small pipes in the
strip that facilitate the
water flow along the
strip (high
permeability along
the band)
Radial consolidation theory
-
To evaluate the distribution of drains and the time required to complete the
consolidation, the consolidation equation was written in axial-symmetrical
condition (the pore water follows radial paths towards the drain/well). The
drainage length is linked to the drain interspace i and layout (triangular or
square pattern).
de = equivalent diameter = diameter of drained zone
• triangular layout de =1.06 i
• square layout de =1.13 i
Triangular pattern Square pattern
i i
i
i
de = equivalent diameter

-
Indicating with u the excess pore water pressure and r the distance from
the vertical axis of the drain, the radial consolidation equation (cylindrical
coordinates) becomes:
  2u 1 u  u kh
ch  2  
 t
ch 
 r r r  mv w
With: kh = permeability in horizontal direction

mv = compressibility ratio in vertical direction
ch = horizontal coefficient of consolidation
It must be integrated with:

1) Initial condition: for t = 0 and every r the excess pressure is u = uo
2) Boundary condition: for t > 0 the excess pressure along the drain is
zero:
u = 0 for r = rw (the lose of total head inside the well is negligible
because it has high permeability)  Barron solution

-
The local consolidation degree is independent by z and it is equal to:
U r   1 
u (r , t )
with: uo ( r )
U (r ) = 0 for u = uo at t =0
U (r ) = 1 for u = 0 at t = ∞
The mean consolidation degree:
 8T 
U h  r e U r dr  1  exp  h 
r
w
 F 
drain
where Th is a time factor defined as:
ch t
Th  2
de
and F is a function of the ratio n=re/rw=de/dw

-
n2 3n 2  1
F  F (n)  2  ln n  
n 1 4n 2
We can used also a simplified relation: F (n)  ln n  0,75
Mean consolidation degree Uz or Ur (%)
Mean excess pore pressure uz or ur (%)
Vertical flow
Radial flow
Time factor Th or Tv


GEO-04 - Confined Compression and Consolidation - GDS

Uploaded by

Copyright:

Available Formats

GEO-04 - Confined Compression and Consolidation - GDS

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

GEO-04 - Confined Compression and Consolidation - GDS

Uploaded by

Copyright:

Available Formats

-

Eng. Giorgia Dalla Santa

Eng. Giorgia Dalla Santa

When on the lateral sides of an infinitesimal volume of soil (with size x, y

These strains are positive x

Eng. Giorgia Dalla Santa

Eng. Giorgia Dalla Santa

ANALYTIC TOOLS FOR THE

Eng. Giorgia Dalla Santa

The typical problem that geotechnical engineers have to solve is the

Eng. Giorgia Dalla Santa

The rigorous solution of a particular problem requires that the equations of

Eng. Giorgia Dalla Santa

The equilibrium equations together with the equations of strain compatibility

Eng. Giorgia Dalla Santa

Eng. Giorgia Dalla Santa

The relationships could be also written in the inverse form:

D = Stiffness Tensor determined as the inverse tensor of compliance= C-1

In general the behavior of soil is:

A correct constitutive model ought to be composed by stress-strain relationships

Eng. Giorgia Dalla Santa

We usually analyze the  Rigid – Perfectly Linear Elastic –

Eng. Giorgia Dalla Santa

From the strain definition the u v w

Eng. Giorgia Dalla Santa

Eng. Giorgia Dalla Santa

In general, when a soil is loaded, it will compress because of:

Eng. Giorgia Dalla Santa

z0 and x=y=0

1= z u = 1

Increment of pore water

Excess pore water

Eng. Giorgia Dalla Santa

The one‐dimensional consolidation testing procedure was first suggested by

Eng. Giorgia Dalla Santa

Eng. Giorgia Dalla Santa

If e0 is known, from the

Eng. Giorgia Dalla Santa

Void ratio e [-]

Eng. Giorgia Dalla Santa

Considering the compression curves in semi-

Vertical effective stress ’z

e OCR=1 e ’z OCR>1

2. Deformability of soil, described by the slope

We define the following parameters, all not

Eng. Giorgia Dalla Santa

Sand Pore pressure, ui+u (t/m2)

Eng. Giorgia Dalla Santa

Terzaghi’s theory studies the Primary consolidation of a saturated soil mass

Eng. Giorgia Dalla Santa

Balance of mass in a porous medium (ev=volumetric strain):

In a porous medium, for a one-dimensional processe occurring along z axis:

Darcy law for a seepage flow in vertical direction z:

Using constitutive relationship for one-dimensional compression:

Ipothesizing that vertical total stress is constant over time:

We get the equation of the 1D consolidation, in this form:

Eng. Giorgia Dalla Santa

The solution of the equation depends on:

ue  0 for z  0 and z  2d and for t  0