Universidad de Chile

Facultad de Ciencias Físicas y Matemáticas.

Departamento de Ingeniería Civil.

HomeWork 1
CI7411 – Soil Dynamics

Students: Sebastián Cerda

Pablo Jara
Diego Silva Contreras
Instructors: Felipe Ochoa Cornejo
Cesar Pastén
Teaching Assistants: Mauricio Cárcamo
Alonso Cubillos
Diego Hernández
Delivery date: 04/09/2017
Section 1

Read and summarize the following paper: Poulos, S. J. (1981). The steady state of deformation.
Journal of Geotechnical and Geoenvironmental Engineering, 107(ASCE 16241 Proceeding).
Steady state definition:
The steady state of deformation for any mass of particles is that state in which the mass is
continuously deforming at constant volume, constant normal effective stress, constant shear stress,
and constant velocity and it’s only achieved after all particle orientation has reached and after all
particle breakage.
During the steady state of deformation, the original structure of the specimen has been completely
destroyed and reworked into a new flow structure.
Examples
Steady state of deformation occurs during undrained tests on fully saturated, loose sands after
liquefaction has been achieved.
It also occurs in drained tests on sands at large strains. Despite the slow deformation rate of these
tests, when large deformations are reached the flow structure develops. If the strains are stopped,
the specimen is no longer in the steady state of deformation. However, the steady state can be
achieved again if large enough shear strains are again applied.
Steady state deformation occurs in clays in drained shear when the ‘’residual’’ shear strength is
reached at very large strains. It also occurs in undrained test in soft clays at very large strains.
The steady state of deformation can occur in any particulate mass and for any loading and drainage
conditions that can break down the original structure and put it into a new ‘’flow’’ structure. The
process of structural breakdown begins with the tiniest shear stress and it ends when all the memory
of the original structure is lost.
The stress-strain curve depends on the way in which the specimen was prepared. In contrast, the
flow structure and the corresponding shear stress, for a given mass particle are not dependent on
initial structure. Thus, for a given soil and a given velocity of deformation, there is a unique relationship
between void ratio, effective normal stress and shear stress during steady state deformation.
Flow Structure
For the purpose of observing the consequences of flow structure in sands, a special undrained test
was performed on a fine quartz sand called Banding Sand. A specimen was compacted at 𝑒𝑐 = 0.693
and consolidated at 𝜎̅𝑐 = 0.2 kgf/cm2. These parameters were based on the known steady state line
as measured by Castro. The results of the triaxial compression test are shown in appendix a, Fig. 1.

The test shows that the specimen exhibits contractive behavior at onset and then becomes dilative.
During the dilative portion, the path on the state diagram crossed the steady state line at point S’ in
Fig.1(a). Yet, the steady state deformation did not occur at that stage.
The explanation for this behavior is that the structure of the specimen at S’ was not that flow structure
which would permit steady state of deformation. The structure has been dilative and I was still
internally braced. As the shear stress was increased and dilation continued, the effective minor
principal stress increased until it reached point M in Fig.1(d). Then the effective stress was high
enough so that the skeleton started trying to decrease in volume. The specimen has become
contractive. Therefore, the pore pressure increased and the effective stress decreased
correspondingly. As the strains increased and the effective stress was dropping, the particles had
enough freedom during shear so that they could become re-oriented under the influence of shear
strains, thus causing the process of pore pressure buildup grain orientation to continue. Finally, when
the strains were large enough, a steady state was reached at point S, after which there was no further
change in shear stress, effective normal stress or pore pressure. The strain rate, was essentially
constant throughout this test. Thus, the steady state of deformation has been reached at a strain of
about 14% and this steady state persisted between 14% and 19% strain, at which point the test was
terminated. Again, at points S and S’, the void ratio, effective normal stress, and shear stress were
identical. Yet the steady state of deformation occurred at S but not at S’.
The only variable that might be different at points S’ and S in Fig. 1 is the structure of the soil. Thus,
the difference in structure at these two points must be the cause of the major difference in behavior
that is evident at the corresponding points on the stress-strain curve. In principle, the steady state of
deformation is not a reached until particle breakage, if any, is complete.
Steady State and Critical State
Critical state described by Schofield and Wroth is when a soil or other granular materials, continuously
distorted until they flow as a frictional fluid. In other words, critical state is an ideal state in wich the
particulate mass deforms continuously but has not preferred structure. Also, particle breakage
neglected. It’s submit that no particulate mass can achieve continuous deformation at constant
volume, at a constant state of effective stress, and at constant velocity without having a preferred
orientation of grains (flow structure).
According to Casagrande the steady state of deformation is achieved in sands when these continuous
expand reaching a critical density and the constant shearing stress.
The steady state is similar to the concept presented by Roscoe, Schofield and Wroth in the following
description of critical state:
In drained test, the critical voids ratio state can be defined when any increment of shear distortion will
not result in any change of void ratio. In a series of drained tests, the critical voids ratios lie on or near
a line. In undrained test, the sample remains at a constant voids ratio, but the effective stress p will
alter to bring the sample into an ultimate state such that the initial voids ratio becomes critical voids
ratio. In series of drained test, the critical voids ratios lie on or near a line. If the results of the undrained
and drained tests coincide in a unique line, then this wil be called the critical voids ratios line (CVR).
One difference between the steady state line and CVR line as envisioned in this quote is the explicit
requirement that the steady state does not exist unless deformation occurs at constant velocity. In
contrast, the implications of the descriptions of critical state in the aforementioned drained case is
that, if shear is stopped, after the critical state is reached, the specimen remains in the critical state.
Finally, Rowe describes constant volume shear as follows:
The ultimate state of continual remoulding at large strains for an actual assembly of particles is
envisaged as one in which half the particles are effectively resisting the applied shear and are just
about to break contact while the other half have failed and are making contact.
In this description, the velocity of deformation is not considered, nor is there any explicit statement of
the particles orientation.
Applications
One of the principal advantages of the concept is that it permits rational organization of stress-strain
data. Also, additional practical applications are: 1) Measurements of pore pressure can be use to
analyze the failure, 2) Slope stability analysis, 3) Helpful to understand stress-strain curves for various
test conditions and 4) Clarify terms like ‘’normally consolidated’’ and ‘’lose’’ against ‘’over
consolidated’’ and ‘’dense’’.
Conclusions
Steady state of deformation requires a constant velocity of deformation and the steady state exists
only so long as deformation continues. The structure of the specimen during steady state deformation
is very special and quite different from both its initial structure and structure specimen before steady
state deformation.
The steady state can be achieved only after grain orientation and grain breakage, if any, are both
complete. Continuous shear after the residual strength is reached in drained tests on clays or sands
are examples of steady-state deformation.
Section 2
2.a. From a principal stresses perspective (Mohr Circle; (p’,q) analysis), describe in detail each
of the stages of the following tests:
2.a.1. Odometer and triaxial consolidation
During the test, the stress state of the soil is always known, this is due to the effective confinement
(𝜎𝑐′ ) applied, given by the difference between the chamber pressure and the back pressure, these
parameters are known. Consequently, in space 𝜏 − 𝜎 this is represented by a Mohr circle of radius 0
and whose center is given by (𝜎𝑐′ , 𝜎𝑐′ ). As the confinement increases, consolidation takes place and
the center moves along the axis of the abscissa. This behavior is represented in figure 1.

Fig. 1: Triaxial consolidation

2.a.2. Direct Shear

In this test, a failure plane is imposed unlike the triaxial test, for this reason, the principal stresses of
the sample during the test can’t be known. Although, at the beginning of the test, the principal stresses
are known, when the deformation is applied, these begin to rotate (in Fig. 2) and the measured
stresses are any stress state within the possible ones. Finally, on reaching the fault it is possible to
determine the stress state at this point.
 Fig. 2: Direct Shear representation. (Ref, 2)

2.a.3. Triaxial shearing in compression

This test begins at the end of the isotropic consolidation, here the applied deviatoric stress generates
differences in the main stresses which generates a cut, this is represented in 𝜏 − 𝜎 space as a Mohr
circle, which grows as it increases the deviatoric stress magnitude. (in Fig. 3)

Fig. 3: Triaxial compression in tau-sigma space

Given the nature of the test the tensional state is always known, because of this, it’s possible
represent the invariants of each Mohr circle generated in space p’-q, obtaining a line of slope 3/2 for
a 3D analysis and Slope 1 for a 2D analysis (in Fig 4).
 Fig. 4: Triaxial drained compression in p - q space

2.a.3. Triaxial shearing in extension

Unlike the triaxial compression test, a negative deviatoric stress is applied which extend the
specimen. The tensional behavior is analogous to the previous case, however, although the radius
of the circle increases, the center of this is shifted toward to the coordinate origin (in Fig. 5).

Fig. 5: Triaxial extension in tau-sigma space

Because of the above, the invariants of each circle fall on a line in p-q space with equal slopes in
modulo but with opposite sign to the case in compression (in Fig. 6)

Fig. 6: Triaxial extension in p - q space

2.a.4. Triaxial shearing with p’ constant

To achieve the constant condition of p', when applying the deviatoric stress it is necessary to decrease
the confinement by the same magnitude in order to maintain the value of p'. Given this, we have that
the trajectory of stresses in the plane p'-q corresponds to a vertical line that starts from the initial
confinement. Meanwhile in the plane τ-σ we have concentric Mohr circles that are increasing its radius
(Fig. 7 and Fig. 8).

Fig. 7: Constant p' in sigma - tau space

 Fig. 8: Constant p' in p - q space

2.a.5. Relate each of the stress analysis involved in the test to a field condition
- Odometer and triaxial Consolidation: a situation that resembles this condition is in the construction
of a tailingshed where given the speed of construction the soil drains the water and consolidates.
- Direct shear: the main condition imposed by this test is the fault plane. This situation can be found
in geological faults or clay lenses within granular soils.
- Triaxial Shearing in compression: this configuration is typical for soils under surface foundations,
where the confinement is given by the geostatic thrust and the diverter by the load on the foundation.
- Triaxial Shearing in extension: this condition can be found in the boot failure mode on piles. In the
interface that defines the fault cone has a traction of the ground.

2.b. Regarding the Mohr Circle, what is the origin of the poles? Provide a detailed description and
prodecures of calculations regarding this important concept
The pole is that point of the Mohr Circle, that if you draw a parallel line to the plane of which you want
to know the stress and extend it until it cuts the circle, then the Cartesian coordinates of the point of
intersection are precisely the tensions of the desired plane.

To obtain the pole from any stresses state, the following steps must be done:

1) Plot the two points that define the stress state (Points A and B).
2) Draw the Mohr circle associated with the stresses state.
3) Draw a parallel line to the plane of action of either of the two points. (The example is drawn with
respect to point A).
4) The point where the line drawn intersects to the Mohr Circle is the Pole.
5) Finally, the stress state can be known for any plane given an angle (alpha) with respect to one of
the initial planes.
This procedure is simple graphical way to obtain stresses on any plane, main stresses, and inclination
of principal planes. It’s shown in figures 9 and 10.

Fig. 9: Origin of the pole. (Ref. 5)

Fig. 10: Origin of the pole. (Ref. 5)

2.c. Propose a problem that involves principal stresses and the Mohr Circle. The problem
must include the solution. Please point out why the problem is relevant, and what do you
seek to measure with the question you are proposing.

Sandra, a passionate student of soil mechanics, visited her uncles that lives in Algarrobo in the
summer. One day, walking along the coast, she was struck by buildings built on the side of the hill,
without thinking any more, she proposed to find the principal stress of the ground three meters deep.
In his search for information, she entered the building and the concierge told him that the slope floor
was a soft rock that could be taken as sand with an internal friction angle of 35 °.

Without being able to solve the problem, Sandra returns to class and tells her professor that she has
not been able to solve a problem with a load distributed in a slope with an angle 𝛿 with respect to the
surface, on a slope with angle 𝛼 with respect to the horizontal, the teacher gives her a hint so she can
continue with the resolution of her problem.

Help Sandra estimate the normal stress at that depth and indicate the horizontal and vertical stresses
and find the principal stress with their respective directions.

Hint: Assume that load distributed and slope are infinity, the following expressions may be useful:
𝜕𝜎𝑧′ 𝜕𝜏𝑧′𝑥′ 𝜕𝜎𝑥′ 𝜕𝜏𝑥′𝑧′
+ = 𝛾 ⋅ 𝑐𝑜𝑠(𝛼) ; + = −𝛾 ⋅ 𝑠𝑖𝑛(𝛼)
𝜕𝑧′ 𝜕𝑥′ 𝜕𝑥′ 𝜕𝑧′

𝛼 30°
𝛿 10°
P 80 (KN/m2 )
𝛾𝑓𝑖𝑒𝑙𝑑 20(KN/m3 )

Fig. 11: Problem representation

Solution:
When decomposing the distributed load into the defined axes is as follows:

δ 𝑃𝑥′ = 𝑃 ⋅ 𝑠𝑖𝑛(𝛿) ; 𝑃𝑧′ = 𝑃 ⋅ 𝑐𝑜𝑠(𝛿)

Fig. 12: Discomposing force in principal axes.

𝜕𝜎𝑥′ 𝜕𝜏𝑧′𝑥′
Because the distributed load and the slope are infinite, is obtained that = = 0, so the state
𝜕𝑥′ 𝜕𝑥′
of stress is given by:
𝜎𝑧′ = 𝛾 ⋅ 𝑐𝑜𝑠(𝛼) ⋅ 𝑧 ′ + 𝑃 ⋅ 𝑐𝑜𝑠(𝛿)
𝜏𝑥′𝑧′ = −𝛾 ⋅ 𝑠𝑖𝑛(𝛼) ⋅ 𝑧 ′ + 𝑃 ⋅ 𝑠𝑖𝑛(𝛿)
(𝜎𝑧′ , 𝜏𝑥′𝑧′ ) = (130.75, −16,11)

The state of stress for this point can be presented by the following differential element:

Fig. 13: Differential element.

To find the other tensional states will be used the method of the Poles, for it is necessary to know the
circle of Mohr associated with the state of stress that has, it’s given that 𝜙 = 35° and how is on a
sand, can be considered null cohesion. Therefore, can be known the Mohr circle since there is only
one circle that is tangent to the Critical State Line and passes through the state of stress found.
 Fig. 14: Circle of Mohr given for the state of stress.

Note: The methodology to solve this problem is graphical, it is important to accurately perform the
figures to arrive at values close to the real ones.
If we are guided by the differential element drawn above, we know that the planes that act on the
tensional states of the element are parallel to the axes of the (𝜎, 𝜏) space, therefore, if we draw a line
with slope 𝑡𝑎𝑛(𝛼) from the pole will intersect at the known tensional state. If the inverse procedure is
done and the line is drawn from (𝜎𝑧′ , 𝜏𝑥′𝑧′ ) with slope 𝑡𝑎𝑛(𝛼). You can find the pole.
From the pole parallel lines are drawn to both planes and the states of stresses of the differential
element are found.
 Fig. 15: States of Stresses found in Circle of Mohr, using Pole Method.

If straight lines are drawn from the pole to the principal states it is possible to obtain the planes in
which they act and their directions, which are presented graphically in this way:

Fig. 16: Differential element for principal states

The values for the states of stresses are as follows:
𝑃𝑜𝑙𝑒 = (403,122)
(𝜎𝑧 , 𝜏𝑥𝑧 ) = (201, 122)
(𝜎𝑥 , 𝜏𝑧𝑥 ) = (403, −129)

(𝜎1, 0) = (130,0)

(𝜎3, 0) = (478,0)
403 − 130
𝜃 = 𝑎𝑡𝑎𝑛 ( ) = 66°
122
3. What are the relations between the parameters of the general failure envelope in the Mohr-
Coulomb Space (σ – τ) and (p’ -q’) space? State the expressions for both 2D and 3D.
The Mohr circle is represented by its invariants, which are defined from the principal stresses:

𝜎1 − 𝜎3 𝜎1 + 𝜎3
𝑅= ; 𝑂′ =
2 2
(1)
Critical state line is given by:
𝜏 = 𝜎𝑛 ∙ 𝑡𝑎𝑛(𝜙) + 𝑐
(2)
Ultimate stress state can be represented from the invariants by the following relations:
𝜎1 + 𝜎3 𝜎1 − 𝜎3
𝜎𝑛 = 𝑂′ + 𝑅 ⋅ 𝑠𝑒𝑛(𝜙) = + ⋅ 𝑠𝑒𝑛(𝜙)
2 2
𝜎1 − 𝜎3
𝜏 = 𝑅 ⋅ 𝑐𝑜𝑠(𝜙) = ⋅ 𝑐𝑜𝑠(𝜙)
2
Replacing the relations of the stress state in the equation of the critical state line is obtained:
𝜎1 − 𝜎3 𝜎1 + 𝜎3 𝜎1 − 𝜎3
⋅ 𝑐𝑜𝑠(𝜙) = ( + ⋅ 𝑠𝑒𝑛(𝜙)) ⋅ 𝑡𝑎𝑛(𝜙) + 𝑐
2 2 2
Grouping the terms according to 𝜎1 𝑦 𝜎3 is the following equation:
𝜎1 ⋅ (𝑐𝑜𝑠(𝜙) + 𝑡𝑎𝑛(𝜙) ⋅ 𝑠𝑒𝑛(𝜙) − 𝑡𝑎𝑛(𝜙)) = 𝜎3 ⋅ (𝑐𝑜𝑠(𝜙) + 𝑡𝑎𝑛(𝜙) ⋅ 𝑐𝑜𝑠(𝜙) + 𝑡𝑎𝑛(𝜙)) + 2 ⋅ 𝑐
1 + 𝑠𝑒𝑛(𝜙) 𝑐𝑜𝑠(𝜙)
𝜎1 = 𝜎3 ⋅ +2⋅𝑐
1 − 𝑠𝑒𝑛(𝜙) 1 − 𝑠𝑒𝑛(𝜙)
(3)

𝑐𝑜𝑠(𝜙) 1+𝑠𝑒𝑛(𝜙)
Since 𝑐𝑜𝑠 2 (𝜙) = 1 − 𝑠𝑒𝑛2 (𝜙) it’s got to the relationship =√
1−𝑠𝑒𝑛(𝜙) 1−𝑠𝑒𝑛(𝜙)
1+𝑠𝑒𝑛(𝜙)
To simplify the calculation is defined 𝑁𝜙 = , so the equation would be written:
1−𝑠𝑒𝑛(𝜙)

𝜎1 = 𝜎3 ⋅ 𝑁𝜙 + 2𝑐 ⋅ √𝑁𝜙

(4)
P-Q Space for 3D
For this space, it is assumed that 𝜎2 = 𝜎3 , the relations between the principal states and the p-q
space are given by:
𝜎1 + 2𝜎3 𝜎1 − 𝜎3
𝑝= ; 𝑞=
3 2
(5)
The critical state line is defined generically as:
𝑞 =𝑚⋅𝑝 +𝑛
 (6)
Replacing eq. (5) in (6), it’s obtained:
𝜎1 − 𝜎3 𝜎1 + 2𝜎3
=𝑚⋅ +𝑛
2 3
Oorganizing the terms in function of 𝜎1 𝑦 𝜎3 it remains:
4𝑚 + 3 6𝑛
𝜎1 = ⋅𝜎 +
3 − 2𝑚 3 3 − 2𝑚
(7)
If we compare Eq. (4) with eq. (7) leads to the following relations:
4𝑚 + 3
𝑁𝜙 =
3 − 2𝑚
(8)
6𝑛
2𝑐 ⋅ √𝑁𝜙 =
3 − 2𝑚
(9)
If m is cleared in eq. (8) and replaced 𝑁𝜙 is obtained:

3 𝑁𝜙 − 1 3 ⋅ 𝑠𝑒𝑛(𝜙)
𝑚= ⋅ =
2 2 + 𝑁𝜙 3 − 𝑠𝑒𝑛(𝜙)

(10)
In the same way, n is cleared in Eq. (9)
3 − 2𝑚 3𝑐 ⋅ 𝑐𝑜𝑠(𝜙)
𝑛 = 2𝑐 √𝑁𝜙 ⋅ =
3 3 − 𝑠𝑒𝑛(𝜙)
(11)
Finally, the critical state line in the space P-Q is given by:
3 ⋅ 𝑠𝑒𝑛(𝜙) 3𝑐 ⋅ 𝑐𝑜𝑠(𝜙)
𝑞= 𝑝+
3 − 𝑠𝑒𝑛(𝜙) 3 − 𝑠𝑒𝑛(𝜙)
(12)
P-Q space for 2D
For this space, it is assumed that 𝜎2 = 0, Because 𝜎3 radial, the relations between the principal states
and p-q are given by:
𝜎1 + 𝜎3 𝜎1 − 𝜎3
𝑝= ; 𝑞=
2 2
(13)
Replacing eq. (13) in (6), it’s obtained:
𝜎1 − 𝜎3 𝜎1 + 𝜎3
=𝑚⋅ +𝑛
2 2
Organizing the terms in function of 𝜎1 𝑦 𝜎3 it remains:
 1+𝑚 2𝑛
𝜎1 = ⋅ 𝜎3 +
1−𝑚 1−𝑚
(14)
If we compare Eq. (4) with eq. (14) leads to the following relations:
1+𝑚
𝑁𝜙 =
1−𝑚
(15)
2𝑛
2𝑐 ⋅ √𝑁𝜙 =
1−𝑚
(16)
If m is cleared in eq. (15) and replaced 𝑁𝜙 is obtained:
𝑁𝜙 − 1
𝑚= = 𝑠𝑒𝑛(𝜙)
𝑁𝜙 + 1

(17)
In the same way, n is cleared in Eq. (16)
𝑐 ⋅ 𝑐𝑜𝑠(𝜙)
𝑛 = 𝑐 √𝑁𝜙 ⋅ (1 − 𝑚) =
1 − 𝑠𝑒𝑛(𝜙)
(18)
Finally, the critical state line in the space P-Q is given by:
𝑐 ⋅ 𝑐𝑜𝑠(𝜙)
𝑞 = 𝑠𝑒𝑛(𝜙) 𝑝 +
1 − 𝑠𝑒𝑛(𝜙)
(19)
4. Read appendix B from the book: “Jefferies, M., & Been, K. (2015). Soil liquefaction: a critical
state approach. CRC press”, which describes details and procedure of the triaxial test.
Summarize the main information in, at least, one page. Images and plots should be included
as an appendix, in addition to the text.

The triaxial test allows to obtain the critical state line according to the characteristics of the sample.
In this sense, there are several factors that can change the results, for example, the void ratio, the
saturation of the sample, the accuracy of the measure system (essentially, the stresses and pore
pressures) and the preparation of the sample.

Sample Preparation

The specimen preparation affects the cyclic strength of sands (because, it changes the fabric of sail
and the stress path to failure). Then, it affects the behavior of the soil during the test, however, does
not affect the critical state line, because the initial structure was destroyed and the sample reached a
statically random critical state condition. Now, the main methods of sample preparation are presented.

1. Moist tamping method

The specimen is prepared at a moisture content of about 5%. This value imply that capillary forces
induced an increase of the effective stress, permits that the sample shape remains.
The sample is prepared in 6 or more equal layers of equal density.
It’s required calculate the void ratio of the sample. Therefore, the following equation is useful:
𝛾𝑤
1 + 𝑒 = 𝐺𝑠 ( ) (1)
𝛾𝑑

Then, we must obtain the dry weight of the sample, corresponding with the target void ratio of the
test.
Finally, the specimen is prepared as follows:
a) Pre-weigh six portions with equal weight of dry sand
b) Mix the material with enough distilled water to reached a moisture content of 5-6% and allow to
cure for several hours in a humid room
c) Tamp the firs layer into the membrane lined mould. It’s necessary adjust the tamper such that the
layer is not over compacted.
d) Scarify the top of the tamper layer to obtain a planar surface before the nest layer deposition.
e) Repeat the layer deposition and tamping process until six layers are formed.
2. Wet pluviation method

This method tries to reproduce the natural structure of soil on the sample. In this case, tries to
reproduce the natural sand deposition. Nevertheless, it’s difficult controlled the void ratio of a
pluviated sample. Also, this method does results on some fines loss from the sand, about 50%. The
main steps of method are:

a) Calculate the total dry weight of sand for the target void ratio
b) Weight out a single oven dried sample of the correct amount and place it in a long-necked flask
c) Add de-aired water to fill the flask and then, apply a vacuum to the top of the flask to ensure
saturation of the sample
d) Leave sample to cure for several hours
e) With a thumb over the neck of the flask, invert the flack and insert the neck into the water in the
mould to approximately 25 mm above the bottom of the mould.
f) Remove thumb
g) Allow the sand pluviated like this while moving the neck of the flask slowly and continuously in a
circular motion. The neck should be kept a constant height of about 25 mm above the top of the
forming sample.
h) Remove the flask when all the sand has pluviated put.

Finally, it’s important check the final fines content and density after the test.

3. Slurry deposition method

This method was proposed to overcome the problem of particle segregation in poorly graded or silty
sand samples1. The fines must be separated from the sand and then, the coarse and fine fractions
are mixed with water and boiled to de-air mixtures. Finally, the sample can be densified by vibration
or tapping.

4. Dry Pluviation Method

This method is used to achieve a uniform density in clean sands. However, this method does not
work in sands with plastic fines or to prepare very loose samples.
It’s required a close control on the rate deposition and the drop height, thus, a range of the density
index of about 30% – 70% can be achieved. For this, the method needs more sophisticated
equipment than the previous methods. Basically, the method consists of dropping the sample through
a sieve, in such a way to control the rate of deposit of the particles, obtaining a homogenous sample.

1
In the original paper, “Sand Sample Preparation – The Slurry Deposition Method” (1988), the authors say
that the method is for well graded and silty sands.
Sample saturation

The fully saturation is essential to assume a constant volume condition during the test.
The first step is flashed de-aired water thought the sample to displace the air. Next, the water pressure
is increased, results in another reduction of the air volume. The degree of saturation is checked by a
“B-test”, whose equation is the following:
∆𝑢
𝐵= (2)
∆𝜎3

When a cell pressure increment is applied, then the pore pressure increment should be equal or
similar if the sample is saturated. A B-value of about 0.97 (or more) in recommended to indicate full
saturation. Nevertheless, a fines content over 5% can difficult the saturation.
If the above procedure is not enough to achieved a B value over 0.97, it’s useful a carbon dioxide
treatment. Prior to saturation, a low volume and pressure CO2 source is connected to the lower platen
water line. The CO2 permits displacing the air and is more soluble than the air, thus, the CO2
treatment ensure a better saturation. Nevertheless, a failure to vent the CO2, will result in a collapse
of the sample.
Other way to obtain saturation is to conduct the flushing process under a vacuum, but requires that
the water containers are all attached to the same vacuum line. But, this method over consolidated
the sample, however, if the test consolidation pressures are higher than 100 kPa, this does not affect
the test.

Void ratio determination

During sample saturation and consolidation, may appear large errors in the void ratio determination,
getting a different result for de critical state line.
While the sample is saturated, the water flow into the sample minus the flow out, is not equal that the
void volume of the sample as air in the water lines and porous stones is also displaced by water.
Then, another way to measure de void ratio, is measure de axial deformation of the sample using the
displacement of the piston, however, calculate the volumetric change in this way required to assume
that the volumetric strain is isotropic, but the experience shown that this is not valid.
The solution is to measure the sample volume after saturation, but it’s needed to use the double
vacuum saturation method, because permit that the saturation process to be outside of the triaxial
cell.
But, the best way to deal with volume change during saturation is to measure the sample volume at
the end of the test, the frozen method. This method requires frozen the sample at the end of the test,
and then, it’s possible remove the sample and obtain the water content. Note that the volumetric
expansion of water on freezing is not an issue, because only the water content is used, not the sample
volume.
When pressure during consolidation is applied, the rubber membrane is deformed and pushed into
the pore spaces between the grains. Then, the sample volume change, therefore, the void ratio must
be corrected. The membrane penetration is quantified in terms of a normalized membrane
penetration:
 ∆𝑉𝑚
𝜀𝑚 =
𝑝′ (3)
𝐴𝑠 log ( 1′ )
𝑝2
And the void ratio correction can be calculated as:

∆𝑉𝑚 𝑝1′
∆𝑒 = (1 + 𝑒0 ) = (1 + 𝑒0 ) 𝜀𝑚 𝐴𝑠 log ( ′ ) (4)
𝑉0 𝑝2

Where:
𝜀𝑚 = normalized membrane penetration
∆𝑉𝑚 = volume change due to membrane penetration
𝐴𝑠 = sample area covered by the membrane
𝑝1′ , 𝑝2′ = net pressure acting across the membrane before and after the volume change
𝑉0 = initial volume
∆𝑒 = void ratio correction

Nevertheless, the membrane penetration correction depends on sand type and testing equipment,
therefore, there are many methods to obtain the void ratio after consolidation.

Additional test details

The triaxial test can be to strain rate controlled or load controlled. However, many tests shown that
the critical state line determined from load or strain rate controlled, does not exhibit a perceptible
difference. Then, it’s more pragmatic used the strain rate controlled testing, also, is easier and less
susceptible to dynamics effects during liquefaction.
Pore pressure measurement is important to determine the steady state line. The pressure transducer
must be close to the sample and its necessary used modern transducers.
Load measurement must use an internal load cell to avoid the friction in the piston, particularly for
small values of deviatoric stress and for loose samples. An area correction to the deviatoric stress its
essential at large strains, because, change the value of the deviatoric stress considerably.
Membrane penetration effect change the measurement of pore pressure. The solution is injecting
water back into the sample to maintain constant volume conditions. Then, the necessary water must
be calculated from the penetration equation of the membrane.
References

1. Been, K., & Jefferies, M. (2015). Soil Liquefaction: A Critical State Approach. Florida: CRC Press.

2. Franch, J. (31 de Enero de 2013). Estudios Geotécnicos. Obtenido de

http://www.estudiosgeotecnicos.info/index.php/ensayos-de-laboratorio-resistencia-de-los-
suelos-2-ensayo-de-corte-directo/

3. Harr, M. E. (1966). Foundations of Theoretical Soil Mechanics. New York: McGraw-Hill.

4. Poulus, S. J. (1981). The steady state of deformation. Virginia: Journal of geotechnical and
geoenvironmental Engineering, 107 (ASCE 16421 Proceeding).

5. Ochoa, F., & Pasten, C. (2016). Apuntes de Clases Dinámica de Suelos 2016