909

REVIEW
Thirty-Ninth Canadian Geotechnical Colloquium: Unsaturated
soil mechanics — bridging the gap between research and
practice
Greg A. Siemens

Abstract: The majority of geoengineering applications occur in the unsaturated (vadose) zone, which is the near-surface region
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23

forming the connection between meteorological phenomena above and saturated ground below. The key characteristic of the
unsaturated zone is that water is in tension or, put another way, pore-water pressure is negative. Moisture content, as well as
most material properties, vary spatially and temporally in the unsaturated zone and coupled processes are common. In geoen-
gineering applications in the vadose zone, unsaturated soils may be present during part or all of their design lives. The question
is how or when to consider the unsaturated soils’ principles in an analysis or design. Although most geoengineering applications
have an unsaturated component, use of unsaturated soil mechanics in practice lingers behind the prolific number of publica-
tions due the uncertain benefit of accounting for unsaturated effects, complexity, and conservativeness among other reasons.
The focus of this colloquium is to continue bridging the gap by illustrating unsaturated soils’ principles using application-driven
examples in the areas of capillarity as well as flow, strength, and deformation phenomena. As principles of unsaturated soils
become more understood and demand increases for incorporating climate change effects in design, use of unsaturated soils’
principles in practice will continue to increase.

Key words: unsaturated soils, capillarity, seepage, strength, expansive soils.

For personal use only.

Résumé : La majorité des applications de géoingénierie se produisent dans la zone insaturée (« vadose »), qui se situe dans la
région proche de la surface, faisant ainsi le lien entre les phénomènes météorologiques ci-dessus et le sol saturé ci-dessous. La
caractéristique clé de la zone insaturée est que l’eau est en tension ou, en d’autres termes, la pression de l’eau interstitielle est
négative. La teneur en humidité, ainsi que la plupart des propriétés du matériau, varient dans l’espace et dans le temps dans la
zone non saturée et les processus couplés sont courants. Dans les applications de géoingénierie dans la zone vadose, les sols non
saturés peuvent être présents pendant tout ou partie de ses durées de vie. La question est de savoir comment ou quand tenir
compte des principes des sols non saturés dans une analyse ou un plan. Bien que la plupart des applications de géoingénierie
comportent une composante insaturée, l’utilisation de la mécanique des sols insaturés persiste derrière le nombre important de
publications en raison des avantages incertains de la prise en compte des effets insaturés, de la complexité et de la prudence
parmi d’autres raisons. L’objectif de ce colloque est de continuer à combler le fossé en illustrant les principes des sols non saturés
en utilisant des exemples axés sur l’application dans les domaines de la capillarité ainsi que des phénomènes d’écoulement, de
résistance et de déformation. Au fur et à mesure que les principes des sols insaturés deviennent mieux compris et que la
demande augmente pour incorporer les effets du changement climatique dans la conception, l’utilisation des principes des sols
non saturés dans la pratique continuera d’augmenter. [Traduit par la Rédaction]

Mots-clés : sols insaturés, capillarité, suintement, force, sols expansifs.

Introduction waste consolidation (Bussière 2007; Qi et al. 2017; Simms 2017),

The majority of geoengineering applications occur in the unsat-
urated, or vadose zone, which can be up to hundreds of metres in
depth depending on the ground profile and climate. Unsaturated
soil applications include foundations, excavations, and buried in-
frastructure (Fig. 1a; Costa et al. 2003; Machmer 2012; Jung et al.
2016), infiltration and landslide triggering (Fig. 1b; Iverson 2000;
Blatz et al. 2004; Tohari et al. 2007; Cascini et al. 2010; Rahimi et al.
2010; Zhang et al. 2011; Robinson et al. 2017), compacted materials
(Fig. 1c; Siekmeier 2007), cover systems (Fig. 1d; Wilson et al. 1994;
O’Kane et al. 1998; Aubertin et al. 2009; Dobchuk et al. 2013;
Huang et al. 2015; Knidiri et al. 2017), evaporation-aided mine
sampling and load tests (Fig. 1e; Konrad 1990; Vanapalli and
Mohamed 2007; Costa et al. 2003), as well as cryogenic suction
processes and contaminant migration (Fig. 1f; Konrad and
Morgenstern 1980; Lenhard and Parker 1987; Azmatch et al. 2012).
The unsaturated component of these examples can be broadly
grouped into flow, strength, and deformation phenomena. The
key characteristic of the unsaturated zone is that water is in ten-
sion or, alternatively, pore-water pressure is negative. Soil mois-
ture content, in turn, varies with negative pore pressure or
suction. A consequence of variable moisture content is that soil
properties, which are relatively constant in the saturated zone,
vary spatially and temporally in the unsaturated zone. Unsatu-

Received 31 December 2016. Accepted 22 October 2017.

G.A. Siemens.* GeoEngineering Centre at Queen’s–RMC, Department of Civil Engineering, Royal Military College of Canada, Kingston, ON K7K 7B4,
Canada.
Email for correspondence: siemens@rmc.ca.
*G.A. Siemens currently serves as an Associate Editor; peer review and editorial decisions regarding this manuscript were handled by P. Simms.
Copyright remains with the author(s) or their institution(s). Permission for reuse (free in most cases) can be obtained from RightsLink.

Can. Geotech. J. 55: 909–927 (2018) dx.doi.org/10.1139/cgj-2016-0709 Published at www.nrcresearchpress.com/cgj on 15 November 2017.
 910 Can. Geotech. J. Vol. 55, 2018

Fig. 1. Geotechnical applications that include an unsaturated component: (a) foundations, tunnels, and excavations; (b) infiltration and
landslide triggering; (c) compacted materials for construction; (d) cover systems; (e) sampling, pile, and plate load tests; and (f) cryogenic
suction and contaminant migration. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.

rated processes are often coupled, meaning they span across one
or more of the flow–strength–deformation groups. For example,
infiltration into a slope (Fig. 1b) is a flow process, which is coupled
with the unsaturated strength of the soil and can lead to landslide
triggering. Unsaturated soils, found in the near-surface region,
are continually interacting with the surrounding environment as
they form the connection between weather systems above and
saturated ground below. Thus, in each of the applications in Fig. 1,
the unsaturated component may be present during part or all of
its design life.

Published by NRC Research Press

 Siemens
Fig. 2. Capillarity: (a) capillarity in zero gravity (image captured from Canada Space Agency video https://www.youtube.com/watch?v=o8TssbmY-
GM, accessed 29 November 2016; reproduced with the permission of Chris Hadfield Inc.), (b) capillary rise in tubes, and (c) contact angle on flat and
sloped surfaces. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
The question is how or when to consider the unsaturated com-
ponent in an analysis or design. Unsaturated soils researchers
continue to move the state of the art ahead with numerous unsat-
urated soils conferences every year. Beginning with the first issue
(Hamilton 1963) and continuing with several Canadian Geotech-
nical Colloquiums papers (Fredlund 1979; Barbour 1998; Bussière
2007; Simms 2017), the Canadian Geotechnical Journal is recog-
nized as one of the key journals for disseminating unsaturated
soils research. In practice, use of unsaturated soils’ principles de-
velops incrementally with local expertise or in applications that
have unsaturated effects at their core (e.g., cover systems or deep
geological repositories for spent nuclear fuel). Thus, despite nu-
merous geoengineering applications that include an unsaturated
component, a gap exists between research and practice. This gap
is due to a number of reasons that include the perceived compli-
cated nature of unsaturated soil mechanics. Another perception is
that, in all cases, ignoring unsaturated effects provides additional
safety in the performance of a geo-structure. In some cases, if the
unsaturated components are ignored in design, an additional
safety factor may be available to the designer, which in my opin-
ion is a reasonable approach. For example, ignoring the unsatu-
911
rated strength component for a foundation above the water table
will underestimate the strength and stiffness of the founding soil.
However, in other cases, ignoring the unsaturated effects can lead
to inefficiencies that affect the expected performance or lead to a
false conclusion of a failure mechanism. For example, ignoring
unsaturated soils effects in modelling transient triggering of a
rainfall-induced landslide. In this paper, unsaturated soils’ prin-
ciples are presented to illustrate opportunities to broaden use of
unsaturated soil mechanics. Unsaturated soils’ principles are il-
lustrated through application-driven examples in the areas of cap-
illarity as well as flow, strength, and deformation phenomena.
Capillarity
Capillarity is a surface tension phenomenon whereby a pres-
sure change occurs across an air–water interface (Lu and Likos
2004; Fredlund et al. 2012). The shape of the air–water interface
and the corresponding change in pressure across the interface is
described by the classic Young–Laplace equation (Maxwell and
Strutt 1911). Figure 2 includes three examples of capillarity: in zero
gravity, in capillary tubes, and on flat surfaces. Canadian astro-
Published by NRC Research Press
 912
Fig. 3. Hydrostatic ground profile highlighting soil type effect on capillary rise. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
naut Chris Hadfield is shown in Fig. 2a on board the International
Space Station with his hand coated in water. Water preferentially
coats his hand as surface tension forces dominate in zero-gravity
conditions. Water also forms menisci between his fingers and the
direction of the surface tension forces is indicated in the figure.
The angle at which surface tension forces act with respect to his
fingers is determined by the contact angle shown in Fig. 2c, which
displays a drop of water on horizontal and sloped surfaces. On a
horizontal surface, the interaction between air–water and the sur-
face is apparent by the contact angle, ␣. In this case, the contact
angle is <90°, which indicates water is the wetting fluid and air is
the nonwetting fluid. The solid surface prefers to be in contact
with water rather than air. The contact angle is also variable, as
illustrated in Fig. 2c for the sloped surface. In the sloped case,
there is a clear leading contact angle, ␣1, and trailing contact
angle, ␣2. Contact angle is a property that is unique to each fluid–
fluid–solid combination. In some cases, fluids can be wetting
while in other cases fluids are wetting. For example, mercury
forms beads on a surface indicating that air is the wetting fluid for
an air–mercury–solid combination.
A classic example of capillarity is given in Fig. 2b, which shows
capillary tubes of different sizes placed into a container of water.
Capillary forces act to draw water up into the capillary tubes to
different heights. The height that water is drawn up the capillary
tube is a function of the radius of the tube as well as the surface
tension, contact angle, and density of water. Performing vertical
force equilibrium by resolving the weight of the water and surface
tension forces, and solving for the capillary rise height, h, gives
the equation in Fig. 2b. The smaller the radius of the tube, the
greater the height of capillary rise. An important implication of
the curved meniscus at the top of each water column is the pres-
sure difference between the air and water in the tube. The pres-
sure difference between the air and water phases also leads to the
definition of matric suction (air pressure – water pressure or ua –
uw). The force arrows pointing upward and the concave shape of
the meniscus indicate that air pressure is greater than the water
pressure. This is a physical representation of the Young–Laplace
Can. Geotech. J. Vol. 55, 2018
equation (Maxwell and Strutt 1911), which describes the pressure
difference across curved surfaces.
In unsaturated ground, water is under the same surface tension
forces as in the capillary tubes with surface tension forces pulling
soil grains together. For a visual example, Bozkurt et al. (2017)
recently measured surface tension forces between two glass
spheres connected by a water bridge. Figure 3 is a hydrostatic
ground profile showing pore pressure and degree of saturation
versus elevation. The profile is at hydrostatic equilibrium and no
flow is occurring. At the groundwater table, pore pressure is zero
and varies hydrostatically above and below. Below the groundwa-
ter table, the ground is saturated and degree of saturation is 100%.
Moving up from the phreatic surface, three distinct zones are the
capillary zone, transition zone, and residual zone. In the capillary
zone, degree of saturation remains at, or close to, 100%. The soil–
air–water representation shows occluded bubbles, which is reflec-
tive of a continuous water phase and discontinuous gas phase.
From the capillary tube analogy, the height of the capillary zone is
reflective of the radius of the largest pores in the soil. Moving
upward into the transition zone, degree of saturation decreases
with elevation. In this zone, the air phase becomes continuous
across the element. Moving upward in the transition zone, degree
of saturation continues to decrease until the residual zone is
reached. The vertical distance from the groundwater table to the
residual zone (indicated in Fig. 3) is soil-type dependent and
ranges from a few millimetres for gravels to hundreds of metres
for clays. From the capillary tube analogy (Fig. 2), this vertical
distance is related to a continuous set of pores that have a radius
less than that associated with the vertical distance above the
groundwater table (equation in Fig. 2b). Within the residual zone,
Sr = 0% is approached asymptotically. The air phase is continuous
while the water phase has become discontinuous and only water
only coats the soil grains (because water is the wetting fluid and
air is the nonwetting fluid).
Published by NRC Research Press
 Siemens
Fig. 4. Typical (a) storage function and (b) unsaturated hydraulic
conductivity function. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
Storage function
The storage function is the fundamental relationship for un-
saturated soils (Brooks and Corey 1964; van Genuchten 1980;
Fredlund and Xing 1994; Barbour 1998; Lu and Likos, 2004;
Fredlund et al. 2012). Also termed the soil water characteristic
curve or water retention curve amongst other names, the storage
function is the relationship between water content and suction as
shown schematically in Fig. 4a. The storage function provides
valuable and important information about the soil. Most often
associated with soils, storage functions have also been reported
for geosynthetics (Bouazza et al. 2006; Bathurst et al. 2007, 2009;
Siemens and Bathurst 2010; Beddoe et al. 2011; Siemens et al. 2012)
and are used to develop frozen soil relationships (Azmatch et al.
2012). The storage function is the first, and often only, function
that is experimentally measured in unsaturated soils. From the
storage function (van Genuchten 1980; Fredlund and Xing 1994),
many other unsaturated soil relationships can be estimated
including unsaturated hydraulic conductivity (Mualem 1976;
Fredlund et al. 1994; Burdine 1953; Leong and Rahardjo 1997) and
unsaturated strength (Vanapalli et al. 1996).
The storage function in Fig. 4a is essentially equivalent to the
vertical profile of degree of saturation for the hydrostatic pore
pressure profile in Fig. 3. Analogous zones and soil–air–water rep-
resentations appear in Fig. 4a to reflect this similarity. Suction, or
matric suction (ua – uw) to be more precise, is expressed as the
difference between the air pressure (ua) and water pressure (uw).
This is consistent with the pressure difference across the air–
water interface explained by the Young–Laplace equation and ex-
pressed physically as the curved surface within the capillary tubes
in Fig. 2c. For the typical case of atmospheric air pressure, suction
is equivalent to the absolute value of the negative pore-water
913
pressure. Similar to the vertical profile (Fig. 3), the magnitude of
suction at which the residual water content is reached varies by
orders of magnitude. Because of this difference, the x-axis is ad-
justed to ensure the entire storage function can be viewed on a
single graph. Coarse-grained soils have suction plotted on an
arithmetic axis while fine-grained soils are plotted on logarithmic
axis. Similar to grain-size curves, logarithmic axes are used for
convenience to be able to visualize the entire relationship.
An important aspect of the storage function is interpretation of
material properties including the air-entry value (AEV) and unsat-
urated hydraulic conductivity function from experimental data.
This is especially true for deformable soils, which will experience
significant volumetric strains during drying and cause changes to
the soil’s pore-size distribution (Romero and Simms 2008). Histor-
ically, storage functions are plotted using a range of moisture
content variables including degree of saturation, volumetric wa-
ter content or gravimetric water content. These are mostly due to
unsaturated soil mechanics taking on aspects of hydrology or
agriculture rather than a geo-perspective. AEV is defined as the
suction at which air enters the soil. Soils that experience swelling–
shrinkage with changes in suction are particularly susceptible to
potential errors in AEV interpretation. Using gravimetric or volu-
metric water content can cause significant errors in determining
AEV or the unsaturated conductivity curve as changes in moisture
content could reflect loss of water rather than air entering the
soil. To properly interpret AEV and the unsaturated hydraulic
conductivity curve, moisture content must be plotted in terms of
degree of saturation (Fredlund et al. 2012). In deformable soils
(Fredlund et al. 2011; Wijaya et al. 2015; Saleh-Mbemba et al. 2016),
measurement of the shrinkage curve, in addition to the storage
function, is recommended to ensure proper interpretation of ma-
terial parameters including AEV and the unsaturated hydraulic
conductivity curve.
Flow phenomena
Unsaturated flow phenomena commonly occur in the near
surface zone with interactions between above-ground weather
systems. Examples include infiltration (Fig. 1b), cover systems
(Fig. 1d), water drawn towards a freezing front due to cryogenic
suction (Fig. 1f), and multi-phase flow applications such as contam-
inant migration (Fig. 1f). In each of these applications, the soil
serves as a conveyant for flow while also undergoing changes in
storage during transient events. The ability of soil to convey flow
applies to movement of water in the near-surface area, water
balance calculations, and ground-climate interactions.
The fundamental principle that controls flow phenomena in
unsaturated ground is that hydraulic conductivity varies with suc-
tion as shown in Fig. 4b. Saturated hydraulic conductivity alone is
one of the most variable material properties, fluctuating more
than 10 orders of magnitude. In unsaturated ground, the hydrau-
lic conductivity of an individual soil also differs by orders of mag-
nitude due to changes in moisture content. The shape of the
unsaturated conductivity function (Fig. 4b) resembles the storage
function (Fig. 4a) with three distinct zones. In the capillary zone at
suctions less than the AEV, the saturated conductivity is essen-
tially retained. In the transition zone, hydraulic conductivity de-
creases as pores desaturate and the water phase becomes more
disconnected while the air phase becomes continuous. In the
residual zone, the water phase becomes discontinuous and hy-
draulic conductivity decreases to essentially nil. The range of dif-
ference from saturated hydraulic conductivity to the minimum is
often greater than four orders of magnitude and can be greater
depending on the soil type.
That hydraulic conductivity varies by orders of magnitude
brings into question the validity of the flow laws for saturated
ground. Researchers found that D’arcy’s law (D’Arcy 1856) and
Bernoulli’s law apply; however, the process for solving the flow
Published by NRC Research Press
 914
Fig. 5. Vertical flow illustrative example: (a) two gravimetric water content profiles, (b) storage function with gravimetric water content
values added to interpret suction, and (c) total head profiles with flow direction indicated. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
equation becomes more complicated. In saturated ground, the
differential equation for flow reduces to the Laplace differential
equation, which can be solved graphically. For unsaturated flow,
the assumptions of isotropic and homogeneous flow are invali-
dated. Therefore, a numerical solution is often used, which may
take the form of finite element or finite difference methods. With
the advent of numerous groundwater flow software, steady-state
unsaturated flow problems can be solved relatively quickly using
numerical solutions.
Unsaturated flowphenomena impact all the applications shown
in Fig. 1, making the understanding of these phenomena vital for
application. Ever more powerful computers and software make
solving steady-state unsaturated flow problems possible with rel-
ative ease; however, care must be taken to appreciate the under-
lying assumptions and inputs. Unsaturated flow phenomena will
be presented in this section to illustrate where appropriate as-
sumptions can be made to obtain a suitable answer and where
detailed soil information is necessary. The three applications are
steady-state vertical flow, steady-state seepage in a dam, and infil-
tration in homogeneous and layered ground. The applications
will illustrate the consequences of making poor assumptions in
an analysis and the implications of those consequences.
Steady-state flow phenomena
Illustrative example No. 1: Determination of vertical flow
direction in unsaturated ground
The first example illustrates the driving gradient for flow in
unsaturated ground. Figure 5a plots two gravimetric moisture
content profiles and the question for both profiles is whether flow
is upward or downward. In both profiles (Fig. 5a), gravimetric
water content increases nonlinearly with depth and a constant
value is obtained at 5 m depth. An initial temptation is to inter-
pret the direction of flow as following the moisture content gra-
dient or that both profiles show downward flow due to gravity.
However, D’arcy’s law and Bernoulli’s law still hold in unsatu-
rated ground and, therefore, water flows due to an energy (i.e.,
total head) gradient. The equation for total head (Fig. 5c) includes
both elevation head and pressure head. Thus the determination of
direction of flow requires knowledge of the storage function. Fig-
Can. Geotech. J. Vol. 55, 2018
ure 5b plots the storage function in terms of gravimetric water
content versus both matric suction and matric suction head.
To properly determine the direction of flow, Bernoulli’s equa-
tion (Fig. 5c) is employed. A datum is set at 5 m depth to assess the
elevation head of each moisture content measurement. Then the
storage function (Fig. 5b) is used to calculate the pressure head for
each gravimetric water content measurement. The moisture con-
tent measurements are plotted on the storage function (ignoring
hysteresis) and the matric suction head for each measurement is
indicated in Fig. 5b. Finally, the total head profile can be calcu-
lated and plotted, as shown in Fig. 5c. Despite seemingly compa-
rable moisture content profiles, opposing flow gradients (one up
and one down) are indicated by the total head gradients. This
example illustrates the principle that unsaturated flow still fol-
lows Bernoulli’s equation, with flow from high total head to low
total head.
Illustrative example No. 2: Unconfined flow through an earth
dam
The second example provides an illustration of a storage func-
tion that is assumed to be sufficient to obtain a realistic solution
to a flow problem. Terzaghi (1943) used seepage through a dam as
an example of where unsaturated principles can be applied. A
similar example, given in Fig. 6, is an earthen dam with 2H:1V
slopes (where H refers to horizontal and V to vertical) and 2 m of
freeboard to retain a 10 m high water reservoir. A 9 m wide toe
drain is located on the downstream side of the dam. The solutions
provided in Fig. 6 include a graphical solution, a finite element
method solution that makes an incorrect assumption, and a finite
element method solution that employs a sensible unsaturated
hydraulic conductivity assumption to obtain an appropriate solu-
tion.
The graphical solution is given in Fig. 6a and a summary of the
flow calculations is shown in the figure. To draw the flownet, an
assumption about the water table is made. The water table is
assumed to be a no-flow boundary and the unsaturated compo-
nent of flow is zero. This is indicated in Fig. 6a as K = 0 in the
Published by NRC Research Press
 Siemens
Fig. 6. Steady-state flow comparison for (a) traditional flownet, (b) FEM assuming saturated conductivity in unsaturated zone, and (c) FEM
with assumed unsaturated conductivity function. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
unsaturated zone and K = Ksat in the saturated zone. A flow of
144 L/day per metre of dam was calculated using the flownet solu-
tion and a Ksat = 10−6 m/s.
A second solution is shown in Fig. 6b, which illustrates an in-
correct, or “black box”, approach to finite element modelling.
With some effort, flow calculations can be performed using soft-
ware; however, proper care is required even for straightforward
problems. In the case displayed in Fig. 6b, the saturated hydraulic
conductivity has been applied to the entire domain, which is in-
915
dicated in the figure as K = Ksat above and below the water table.
This assumption could be made to be “conservative” or by mis-
take. Comparing the flow calculations for Figs. 6a and 6b indicates
they are significantly different. Assuming that the entire domain
has saturated conductivity increases the total flow to 227 L/day/m
with 86 L/day/m being the unsaturated component. The flow be-
low the water table (141 L/day/m) is comparable to the flow net
(144 L/day/m). Therefore, the issue is with the unsaturated compo-
nent, which is too high.
Published by NRC Research Press
 916
A third solution, which sensibly accounts for the unsaturated
component of flow, is shown in Fig. 6c. In this case, the unsatu-
rated hydraulic conductivity curve was assumed using the “stock”
storage function from within the software. The only input re-
quired is a saturated hydraulic conductivity and the soil type. The
solution, shown in Fig. 6c, shows that the total head contours are
similar to Fig. 6b. However, the flow calculations are significantly
different. Flow below the water table is principally equivalent
(141 versus 142 L/day/m) while the unsaturated component is just
13 L/day/m compared with 86 L/day/m in the black box solution.
The most representative solution for estimating unconfined
flow within a homogeneous earthen dam can be made using a
finite element model (FEM) solution with an assumed unsaturated
conductivity function (Fig. 6c). The flownet solution ignored the
unsaturated component of flow and underestimated flow by 10%.
The black box solution, which applied saturated hydraulic con-
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
ductivity to the entire domain, overestimated flow by 40% and
the unsaturated component of flow by 500%. For cases in which
the saturated conductivity is known, assuming the unsaturated
conductivity function is normally acceptable for steady-state
problems. In these applications, the impact of the unsaturated
conductivity function serves to decrease conductivity above the
water table. With some practice, steady-state problems can be
solved relatively quickly. Sensitivity analyses can also be per-
formed to get a sense of the impact of uncertainties in the geom-
etry, material properties, and boundary conditions.
Transient flow phenomena
Transient unsaturated flow phenomena are inherently more
For personal use only.
complicated compared with steady-state and require a higher
level of material property measurements. For example, in cover
systems (Fig. 1d, Wilson et al. 1994; O’Kane et al. 1998; Aubertin
et al. 2009; Dobchuk et al. 2013; Huang et al. 2015; Knidiri et al.
2017) transient analysis allows for consideration of soil–weather
interactions, which are coupled problems that include heat trans-
fer, vapour flow, and liquid flow, to calculate evaporation from
the soil surface (Bittelli et al. 2008; Lehmann et al. 2008; Or et al.
2013). In this section, transient flow phenomena are illustrated
using unsaturated transparent soil. An unsaturated transparent
soil experiment is shown in Fig. 7a, which is a series of digital
photos of infiltration in a layered profile. Transparent soil is
formed by matching the refractive indices of a soil and pore fluid
and several combinations of transparent soil are found in the
literature (Iskander et al. 2015). Saturated transparent soil allows
for direct observation from within the soil mass of deformations,
strains, and soil–structure interaction rather than at the bound-
ary of an experiment. Unsaturated transparent soil takes this con-
cept in another direction (Peters et al. 2011). Observations in
unsaturated experiments are often limited by the number of mea-
surement devices that can be located within an experimental ap-
paratus. A nominal number of discrete measurement points are
normally included to measure pore pressure, suction, and (or)
moisture content. Unsaturated transparent soil, combined with
digital image analysis, allows for measurement of degree of satu-
ration to the millimetre resolution, which is orders of magnitude
more data than can be obtained from discrete measurements.
To illustrate transient moisture migration in a heterogeneous
profile, a transparent infiltration experiment is shown in Fig. 7a
with a fine layer between two coarse layers. A 10 cm high pond was
placed at the surface and the wetting front descends through the
profile. Digital images were collected every 5 s in this experiment
and a select number is shown in the figure. The wetting front is
clearly visible and is indicated by the black arrow. As the wetting
front reaches the fine layer, ponding occurs and the soil above the
fine layer changes to a darker colour associated with an increase
in saturation. In the final photo, ponding is complete and the
wetting front continues to descend through the fine layer. These
images visually illustrate how moisture migration is affected by a
Can. Geotech. J. Vol. 55, 2018
relatively thin finer layer. The finer layer serves as a bottleneck to
the system and causes saturation above as well as development of
positive pore pressure.
Transparent soil allows for clear identification of the wetting
front and measurement of the degree of saturation within the
profile at high spatial and temporal resolution. Each digital image
is processed by normalizing the image intensity between an im-
age of a dry profile (white) and a saturated profile (black) to calcu-
late the normalized intensity, IN, at the pixel scale. The calibration
curve plotted in Fig. 7b is used to convert normalized intensity to
degree of saturation. On the plot are calibration points as well as
a predicted curve. At IN = 0, the soil is dry and IN = 1, degree of
saturation is 100%. At 90%, discrete air bubbles are visible while at
lower saturations the soil visually lightens continually until it is
dry. Details on the digital image process can be found in Peters
et al. (2011) and Sills et al. (2016). Transparent soil used at Royal
Military College is formed from fused quartz and comes in two
types, termed coarse and fine gradations (Fig. 7c). Both gradations
are uniform sands with D10 equal to 0.75 and 0.13 mm, respec-
tively. The saturated conductivity of the two gradations differs by
less than one order of magnitude.
Impact of heterogeneities on infiltration
To quantitatively illustrate the impact of minor heterogeneities
on unsaturated processes, detailed results from two infiltration
experiments are plotted in Fig. 8 including degree of saturation
profiles during the two experiments, wetting front location ver-
sus time, and pore pressure profiles from the beginning and end
of the experiments. The only difference between the two experi-
ments is the 120 mm fine layer placed at 510 mm depth below the
surface. The D10 of the finer gradation is just 0.62 mm less than the
D10 of the coarse gradation. The water table is located a 180 mm
elevation and a 10 cm pond is placed on the surface at the begin-
ning of the experiment. The uniform profile experiment shows
the dry degree of saturation profile to start and the wetting front
starting to descend at time = 0 s. As the wetting front descends, the
soil increases Sr behind the wetting front and the experiment is
completed at 380 s. The wetting front versus time (Fig. 8c) shows
the wetting front descending linearly from the surface down to
the water table. The initial and final pore pressure results show
the initial vertical suction profile, which then becomes a linear
decrease in pore pressure from the 10 cm pond at the surface to
0 kPa at the water table.
Infiltration in a layered profile shows the impact of a fine layer
with D10 differing just 0.62 mm between the two gradations. Prior
to the wetting front reaching the fine layer, the experimental
results are identical in terms of the change in saturation and the
wetting front descent rate. As the wetting front reaches the fine
layer, the wetting front slows down significantly (Fig. 8c). The
wetting front continues to descend at a slower constant rate be-
low the top of the fine layer until the water table is reached at
the end of the test. The wetting front encountering the fine layer
causes a ponding above the top of the fine layer and a hydrostatic
pore pressure profile occurs (Fig. 8d). The pore pressure dissipates
across the fine layer and is essentially 0 kPa in the lower coarse layer.
Small heterogeneities affect the mobility of the wetting front,
infiltration rate, and pore pressure response during infiltration.
In the field, small differences in grain size leading to a layered
profile often occur naturally or during construction. Comparing
the degree of saturation profiles at the end of infiltration (last
profiles in Figs. 8a and 8b) shows they are identical above the fine
layer in the uniform and layered experiments. The wetting front
also descends at the same rate until the sharp change when the
fine layer is reached. Below the fine layer, the wetting front moves
four times slower compared with the uniform profile. From a flow
perspective, the fine layer serves as a bottleneck that reduces the
transmissive capacity of the system. The pore pressure response at
the end of the test is also significantly different. Pore pressure in
Published by NRC Research Press
 Siemens 917

Fig. 7. Transparent soil for unsaturated applications showing (a) series of photos from an infiltration test in a layered system (Siemens et al.
2014), (b) calibration curve for image analysis (modified from Sills et al. 2016), and (c) grain-size distributions for coarse and fine transparent
soils (after Siemens et al. 2014). [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.

Published by NRC Research Press

 918 Can. Geotech. J. Vol. 55, 2018

Fig. 8. Impact of heterogeneities on infiltration: (a) saturation profiles from open infiltration in coarse transparent soil, (b) saturation profiles
from open infiltration in layered soil profile, (c) elevation of wetting front versus time, and (d) equilibrium pore pressure profile (after
Siemens et al. 2014). [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.

Published by NRC Research Press

 Siemens
Fig. 9. Unsaturated strength phenomena illustrating variation in
apparent cohesion with degree of saturation. [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
the uniform coarse profile decreases from the 10 cm pond at the
surface to 0 kPa at the water table (Fig. 8d). In the layered experi-
ment, a hydrostatic profile occurs above the fine layer with the
initial suction dissipated below the fine layer. These experiments
illustrate the impact of a thin and minor heterogeneity on infil-
tration. The fine gradation layer is just 120 mm thick and the
For personal use only.
D10 is only 0.62 mm different. However, the fine layer caused
important differences to the pore pressure response and water
balance calculations.
In the steady-state examples given above, estimations of the
unsaturated functions were adequate to model the results. How-
ever, in transient applications, the unsaturated functions have a
more significant role in the results. The two experiments plotted
in Fig. 8 were modelled using the finite element method (Siemens
et al. 2014) and the results are plotted in Figs. 8c and 8d. The
storage functions and unsaturated conductivity functions were
measured for coarse and fine gradations. The results match the
model values throughout the experiments; however, both the
experiments and numerical simulations require experience to ob-
tain reliable results. If infiltration rate is the most important con-
sideration, another option is to use the Green–Ampt model (Green
and Ampt 1911), which gives the infiltration rate and wetting front
location (Siemens et al. 2013). With proper calibration of the
Green–Ampt parameters and knowledge of the conductivity of
the transmissive zone, these models are options depending on the
requirements of the application.
Flow phenomena summary
Flow phenomena impact most geotechnical applications that
include an unsaturated component. Applications in unsaturated
ground (Fig. 1) are inherently connected to the weather systems
above ground and therefore will interact hydraulically. Flow
mechanisms in unsaturated ground are impacted by the hydrau-
lic conductivity function, which varies with suction. This section
illustrated that the driving energy for flow follows Bernoulli’s law,
showed how reasonable assumptions can be made for steady-state
applications, and demonstrated the influence that small hetero-
geneities have on transient unsaturated flow.
Strength phenomena
Geotechnical applications that include an unsaturated strength
component include foundations and buried infrastructure (Fig. 1a),
landslides (Fig. 1b), compacted construction materials (Fig. 1c), and
pile and plate load tests (Fig. 1e). For these types of applications
located in the vadose zone, a component of their performance is
919
Fig. 10. (a) Wall painting from tomb of Djehutihotep in
southeastern Egypt (after Newberry 1895) and (b) interface friction
experiments plotted as load versus displacement. [Colour online.]
dependent on unsaturated soil principles. Within the unsaturated
zone, strength is a function of suction. Commonly found evidence
of unsaturated strength principles is sand castles constructed
from a coarse-grained (also termed cohesionless) soils as depicted
in Fig. 9. If a coarse-grained soil is dry (Sr = 0) or saturated (Sr = 1),
the maximum slope the soil is stable at is the angle of repose or
angle of internal friction. However, if the soil is unsaturated it is
stable at steeper slopes, which allows for creation of impressive
structures such as the one shown in Fig. 9. In an unsaturated state,
surface tension forces act at the granular level to increase the
normal stress between particles. In this framework, the friction
angle of the soil remains constant while the apparent cohesion
intercept is affected by suction. Figure 9 shows a Mohr–Coulomb
strength envelope for saturated and unsaturated soils. For satu-
rated and dry soils (Sr = 1 and 0, respectively), apparent cohesion is
nil and the strength envelope is dependent only on the friction
angle of the soil. For unsaturated soils, an apparent cohesion term
elevates the strength envelope vertically. The amount of apparent
cohesion is related to the storage function. The apparent cohesion
intercept increases through the capillary zone and transition zone
and decreases in the residual zone (Vanapalli et al. 1996; Lu and
Likos 2004; Fredlund et al. 2012).
All soils in the vadose zone have an unsaturated component of
strength associated with apparent cohesion. In practice, the un-
saturated component is infrequently counted on in design due to
its inherent transient nature. In this section, the focus is placed on
examples in which ignoring unsaturated strength would have led
to the wrong answer or development of an application. An inter-
Published by NRC Research Press
 920 Can. Geotech. J. Vol. 55, 2018

esting historical example of unsaturated soil strength and the Fig. 11. Effect of suction on plate load experiments in sand (after
effect of suction on plate load experiments are examined. Vanapalli and Mohamed 2007) showing (a) storage function and (b)
applied stress versus settlement. [Colour online.]
Historical example
An historical application of unsaturated soil mechanics is illus-
trated in Fig. 10a, which is a wall painting from a tomb in south-
eastern Egypt (Newberry 1895). The wall painting shows a statue
on a sled being dragged over the sandy desert. The application of
unsaturated soil mechanics is shown by the person adding water
to the ground directly in front of the sled. Once thought to be a
ceremonial act, this person is actually using unsaturated princi-
ples to minimize interface friction between the sand and the bot-
tom of the sled.
To test this hypothesis, Fall et al. (2014) performed a set of
experiments to examine the influence of moisture content on
interface friction. The experiment (plotted in Fig. 10b as load ver-
sus displacement) consisted of placing a mass on top of a 11 cm ×
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23

7.5 cm sled and pulling it across a sandy material prepared at

specified water contents. The results (Fig. 10b) show that an initial
stiff response transitions to one where load remains constant
with displacement. The volumetric moisture content of each ex-
periment is noted in Fig. 10b. The authors observed that when the
sand was dry or at high moisture content the sled sank into the
ground and plowed sand out of the way. Between these two ex-
treme moisture contents, the sled sank less into the soil and less
plowing occurred. Figure 10b indicates that when the soil is dry or
saturated the effort to pull the sled across the sand is high com-
pared with a minimal load for VWC = 5%. At VWC = 5%, for this
particular sandy material, the strength of the material is maxi-
For personal use only.

mized and the load to pull the sled is minimized.

The unsaturated soil principle illustrated in Fig. 9 is displayed
with this historical application. Figure 9 shows that apparent co-
hesion is a function of saturation. At Sr = 1 or 0, strength reduces
to the saturated value, which from an interface-friction perspec-
tive, led to plowing of the sled through the sand and an increase in
pulling load. At VWC = 5%, the soil strength was maximized, the
sled slid on top of the sand, and the pulling load was minimized.
The Egyptians wanted to maximize speed and mobility for trans-
port of statues and construction materials. Therefore, minimizing
the interface friction between the sled and the ground allowed for
more efficient transport. In a dry desert environment, the near-
surface sand would be at a moisture content less than this optimal
value. Thus the person on the front of the sled wetting up the
surficial sand would temporarily increase the sand strength and
minimize pulling effort. From Fig. 10b a 30% reduction in pulling
load was found for the optimized moisture content compared
with both dry and wetter sand. Thus a 30% efficiency was gained
by taking advantage of unsaturated principles.
Plate load tests
The effect of suction on unsaturated soil strength can be impor-
tant in back-analysis of failures as well as interpretation of in situ
tests. Understanding the impact of suction on the unsaturated
strength envelope can significantly impact interpretation of the
failure mechanism. To understand the significance of suction on
surface loads, Vanapalli and Mohamed (2007) performed a series
of laboratory experiments where they applied vertical pressure to
the surface of a soil formed at specified moisture contents. The
experiments were performed in a 90 cm × 90 cm × 75 cm (width ×
length × height) reinforced box and the loading surface was
10 cm × 10 cm. The soil is characterized as a sandy gravel with the
storage function plotted in Fig. 11a. The soil has an AEV of 3 kPa
(⬃30 cm) and its residual saturation is obtained at a suction of just
8 kPa (⬃80 cm). The friction angle of the soil, ␾  = 39°.
The question is, what impact could 6 kPa suction have on the
failure load? The four plate load experiments were performed at
suctions of 0, 2, 4, and 6 kPa with results plotted in Fig. 11b. The
results show that the applied stress – settlement curves differ due
to the suction. In each test a peak stress is obtained followed by
some decrease for the lower suction tests. A stiffer response to the
load is notable for the unsaturated experiments. A less stiff re-
sponse and a lower peak stress is apparent for the test performed
at 0 kPa suction (i.e., saturated). The peak value increases with
increasing suction over the range of tests performed. At just 6 kPa
suction, the peak load capacity increases seven-fold compared
with the saturated test.
Practical implications can be drawn from these tests. The only
difference in the plate load experiments was the moisture content
of the soil. Reducing the moisture content increases suction and,
in turn, increases the apparent cohesion intercept of the failure
envelope (Fig. 9). The implication for interpreting plate load tests
in the field is the need to measure moisture content in the zone of
influence. Any variation in moisture content will have a corre-
sponding effect on the strength of the soil. In the experiments
presented in Fig. 11, the degree of saturation varied from 58% to
100% and had a seven-fold effect on the peak capacity of the load
test.
Strength phenomena summary
Unsaturated soils experience an increase in apparent cohesion
associated with capillary forces acting at the granular scale. How-
ever, owing to the inherent transient nature of moisture migra-
tion within the unsaturated zone, apparent cohesion is also a
transient. In the analysis of failures, an in situ testing knowledge
of unsaturated strength principles is recommended to correctly
identify the governing mechanisms.

Published by NRC Research Press

 Siemens
Deformation phenomena
Deformation phenomena associated with unsaturated soils of-
ten coincide with a change in the environmental conditions that
alter the moisture regime. The key characteristic with respect to
deformation phenomena is that unsaturated soils can experience
volumetric strains due to fluctuations in both stress and moisture
content (i.e., suction). Volumetric deformations associated with
stress are perhaps more intuitive compared with deformations
associated with moisture fluctuations. Classic examples of volu-
metric deformations associated with moisture changes are expan-
sive soils and collapsible soils. Expansive soils respond to wetting
and drying cycles with significant volumetric swelling and shrink-
age as displayed in Fig. 12 due to their clay mineralogy. Collapsible
soils are recently deposited or altered soils, which have a high
void ratio, high sensitivity, and low interparticle bonding. In con-
trast to expansive soils, collapsible soils respond to wetting by
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
subsiding or hydroconsolidating with significant settlement
(Clemence and Finbarr 1981; Basma and Tuncer 1992; Rogers et al.
1994; Houston et al. 2001). Problems associated with collapsible
soils include identifying and characterizing them during a site
investigation, predicting the magnitude of wetting, quantitative
predictions of collapse strains, and selecting design or mitigation
alternatives (Houston et al. 2001). In this section, the main princi-
ples of unsaturated soil deformation phenomena associated with
wetting–drying of expansive clay soils are illustrated and a con-
ceptual framework for swelling potential is presented.
The most well-known soils that encounter deformations associ-
ated with moisture fluctuations are expansive soils. The most
For personal use only.
susceptible structures to expansive soil effects are lightly loaded
infrastructure such as shallow foundations (Fig. 12a; Domaschuk
1986). The American Society of Civil Engineers report that expan-
sive soils cause owners more financial loss than the combined
effect of hurricanes, earthquakes, and tornadoes. For reference,
Hurricane Sandy caused approximately USD$65.6 billion of dam-
age (NOAA 2018). In the basement schematic (Fig. 12a), a combina-
tion of events induces both swelling and shrinkage, which leads to
damaging differential displacements. Prior to construction, a veg-
etative layer is often removed. The vegetative layer provides an
elevated suction environment through the root systems, which
uptakes moisture from the ground. Once the vegetative layer is
removed and the area covered, the soil beneath the house experi-
ences a suction decrease as the house cuts off direct access to the
ground surface. The soil swells and the house heaves in response
to the decrease in suction. Adjacent to the house, a shrinkage
scenario ensues as a large tree is planted and the root systems
descend deep below the ground in the area next to the founda-
tion. A growing tree continually requires more moisture from the
ground, which comes from its roots and leads to an increase in
suction around the perimeter foundation. Owing to the increas-
ing suction, the soil shrinks in response. The overall effect of the
swelling under the centre of the house and the shrinkage on the
periphery is differential settlements and cracking of the base-
ment. The damaging cost to owners is worsened as they often
occur to personal dwellings and few options may be apparent,
which leads to extreme actions. The photograph in Fig. 12b shows
an example of extreme actions with basement walls being re-
moved from a house, which were then replaced along with the
foundation (Siemens 2007).
To illustrate the magnitude of volumetric strains that can occur
in expansive soils, digital images from a shrinkage experiment
and a swelling experiment are given in Figs. 12d and 12e, respec-
tively. In the shrinkage experiment (Fig. 12d), the test begins with
an initially saturated sample. The specimen is initially located
within a test ring to provide confinement, which is later removed
once separation between the ring and the specimen occurs. The
specimen dries in the low humidity laboratory environment and
shrinks. After 8 days, the specimen shrinks by 75% of its initial
921
volume. The suction – volumetric strain path for the shrinkage
experiment is illustrated in Fig. 12c, which is a constitutive model
for volumetric strains in unsaturated soil for both matric suction
and net normal stress. Initially saturated, the specimen is at 0 kPa
matric suction. As evaporation of the pore water occurs, matric
suction increases and the soil shrinks in response. In the swelling
experiment (Fig. 12e), an unsaturated expansive soil is given access
to water through filter paper strips in contact with a water reservoir
(Lim and Siemens 2013). The initially unsaturated sample takes on
water and swells in response. After 10 days, the soil swells to
–59.4% volumetric strain. Figure 12c plots the volumetric strain –
matric suction path, which is followed during the swelling exper-
iment. The specimen is initially unsaturated and at elevated
suction. As the specimen is given access to water, matric suction
decreases and the sample swells in response.
Unifying concept of swelling potential
Owing to the cost of damage to infrastructure constructed in
expansive soils as well as the importance of such soils for use as
environmental barriers (i.e., landfill liners), much research has
sought to characterize the swelling potential of these soils. Swell-
ing potential is measured in the laboratory (ASTM 2003, standard
D4546) and also estimated from index properties (Komine and
Ogata 1994; Prakash and Sridharan 2004; Cui et al. 2012; Ito and
Azam 2013). Concepts for analysis and prediction of infrastructure
constructed in expansive soil are based on moisture content
changes, suction or are empirically based (Fredlund 1983; Briaud
et al. 2003; Houston et al. 2011; Vanapalli and Lu 2012; Puppala
et al. 2014). Owing to the number of continually expanding re-
sources and publications associated with expansive soils, the re-
search and practice communities are still in need of a practical
methodology to assess the effect of swelling potential to infra-
structure.
Swell equilibrium limit concept
A unifying concept for swelling potential is illustrated in
Fig. 13a using the application of a retaining wall. The retaining
wall is constructed on swelling ground and expansive soil is also
used as backfill material. An extended infiltration event occurs to
induce wetting conditions to the system. During wetting, swelling
potential can be satisfied by a combination of volumetric expan-
sion and swelling-induced stresses, which depend on the bound-
ary conditions during wetting. Three cases are illustrated in
Fig. 13a and their stress–volume paths plotted in Fig. 13b in terms
of specific volume (V = e + 1) versus mean stress (p = (␴1 + ␴2 + ␴3)/3).
The three cases are unconfined swelling, swelling under the foun-
dation, and swelling adjacent to the retaining wall. For the first
case, swelling occurs in front of the retaining wall under uncon-
fined conditions, similar to the laboratory experiment in Fig. 12e.
In Fig. 13b, the stress–volume path begins at zero stress and moves
vertically upward until the swelling potential is satisfied. The sec-
ond case for swelling is under the foundation. In this case the
expansive soil is subjected to an elevated stress level prior to the
infiltration event, which serves to attenuate the magnitude of
swelling compared with the unconfined case. In Fig. 13b, the
stress–volume path begins at an elevated stress level and moves
vertically upward until the swelling potential is satisfied. The
third case for swelling is adjacent to the retaining wall. In this
case, volumetric confinement is provided by the retaining wall,
which restricts expansion. Thus swelling potential is satisfied by
swelling-induced stresses during wetting. From a stress–volume
perspective, the soil state begins at a location associated with the
stress level equal to the overburden stress and a volume equiva-
lent to the end of construction. During wetting, the stress–volume
path is sloped upward to the right as expansion is limited by the
existence of the stiff retaining wall. At equilibrium, all three soil
elements’ swelling potential have been satisfied and their end-
Published by NRC Research Press
 922 Can. Geotech. J. Vol. 55, 2018

Fig. 12. Unsaturated deformation phenomena including (a) combined effects of heave and shrinkage on a residential structure (after
Domaschuk 1986), (b) removal of basement walls from a residential structure affected by expansive soils, (c) constitutive model for volumetric
strains due to changes in suction and net normal stress, (d) shrinkage experiment, and (e) free swell experiment (after Lim and Siemens 2013).
[Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.

Published by NRC Research Press

 Siemens
Fig. 13. Swelling equilibrium limit (SEL) results: (a) retaining wall constructed in expansive soil illustrating swelling conditions and
(b) swelling paths and experiments for bentonite–sand buffer; SEL summary plots for (c) bentonite–sand buffer, (d) Bearpaw clay, and (e) Lake
Agassiz clay (after Lim and Siemens 2016). [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
points are on the “swelling equilibrium limit” (SEL, Siemens and
Blatz 2009).
Measurement of swelling equilibrium limit
The SEL is unique for each soil and is measured experimentally
with a series of swelling experiments such as those presented in
Fig. 13b. The experimental methodology (Siemens and Blatz 2007;
Lim and Siemens 2013; Powell et al. 2013) consists of bringing
identically prepared unsaturated specimens to defined initial con-
ditions and then wetting under idealized boundary conditions
(Fig. 13b). Results from three triaxial swelling experiments to de-
fine a SEL are plotted in Fig. 13b. In each test, the specimen is
923
brought to an initial stress condition and then given access to
water. During the swelling phase, controlled boundary conditions
define the stress–volume path followed during the test. In Fig. 13b,
each test was brought to an initial stress state of 250 kPa and then
the swelling phase was initiated. The three boundary conditions
applied during the tests were constant mean stress (CMS, vertical
path), constant stiffness (CS, upward sloping path), and constant
volume (CV, horizontal path). The constant mean stress – volume
path is equivalent to the soil swelling under the foundation in
Fig. 13a as the initial total stress is maintained during the swelling
phase. When subjected to a constant mean stress, swelling poten-
Published by NRC Research Press
 924
Fig. 14. SEL estimation for Regina clay: (a) SEL parameter “A” and (b) SEL parameter “B” versus liquid limit and (c) SEL estimation comparison
with laboratory data (after Lim and Siemens 2016). [Colour online.]
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.
tial is satisfied by volumetric expansion. The constant stiffness
stress – volume path is an idealized representation of swelling
adjacent to the retaining wall (Fig. 13b). In the experiment, the soil
swells against a linear-elastic spring boundary condition. The
third swelling test is constant volume, in which swelling potential
is satisfied entirely with swelling-induced stresses. The initial vol-
ume is maintained constant during swelling and a horizontal path
is followed during swelling. At the end of each test, swelling po-
tential is satisfied and the soil state lies on the SEL.
The unifying concept of swelling potential is illustrated for
three soils in Figs. 13c–13e, which include their SELs and unsatu-
rated isotropic compression lines. The SEL is defined by fitting a
logarithmic curve to the end of swelling test points as shown in
Figs. 13c–13e for the bentonite–sand buffer, Bearpaw clay, and
Lake Agassiz clay. The triaxial swelling experiments shown in
Fig. 13b have their end of swelling test states plotted in Fig. 13c for
reference. The isotropic compression line on which each of the
swelling tests was initiated is also shown in the SEL plots. The area
between the SEL and the isotropic compression line is the swell-
ing potential for each soil. The swelling potential is at a maximum
at zero stress and decreases with increasing stress. The isotropic
compression curve and SEL converge at higher stresses where the
confining stress overcomes the swelling potential of the expan-
sive soil.
Use of swell equilibrium limit
Practical use of the SEL concept relies on knowing the initial
stress and volume states of a soil and the boundary conditions
during swelling. For a house foundation or a retaining wall these
are defined for constant mean stress areas and strategies are being
developed for more complex cases (Lim 2014). Finite element mod-
Can. Geotech. J. Vol. 55, 2018
elling of foundations (Siemens and Blatz 2008) have shown the
potential for analysis tools. The question arises whether SELs can
only be defined in the laboratory or whether they can be esti-
mated from readily measured material properties as has been
used in other analyses (Komine and Ogata 2003; Prakash and
Sridharan 2004; Cui et al. 2012; Ito and Azam 2013). Lim and
Siemens (2016) showed that SELs can be estimated from liquid
limit, free swell potential, plasticity index, and initial specific
volume. The process involved fitting SEL equation parameters
(VSEL = A + Bln(p)) to the material properties for the soils as plotted
in Figs. 14a–14b for liquid limit. In the plots, SEL parameter “A”
increases with increasing liquid limit while SEL parameter “B”
decreases. Parameter A is the y-intercept of the SEL equation at
zero mean stress while B is related to the curvature. In both plots
(Figs. 14a and F14b) linear fits are provided although, in the future,
nonlinear fits may be justified when the database size is increased.
Using the predicted values from Figs. 14a–14b, the SEL for Regina
clay was estimated based on its liquid limit of 76% (Fredlund 1975).
The resulting SEL is plotted alongside experimental data in Fig. 14c
and the results are comparable.
Deformation phenomena summary
Expansive soils display most distinctly that volumetric defor-
mations in unsaturated soils are induced by changes in suction.
Due to this susceptibility, expansive soils cause significant dam-
age to infrastructure, which can lead to extreme actions by own-
ers (Fig. 12b). The SEL provides a conceptual framework to analyze
and predict swelling behaviour and the ability to predict SELs
from index properties gives more encouragement for its use. Fu-
ture research is directed at development of a practical analysis
tool.
Published by NRC Research Press
 Siemens
Summary
Unsaturated soils are found throughout nature in the vadose
zone and form the connection between above ground meteoro-
logical systems and saturated ground below. Geoengineering ap-
plications in the vadose zone include an unsaturated component
during some or all of their design lives (Fig. 1). Use of unsaturated
principles is increasing in practice owing to knowledge accessibil-
ity, appreciation of unsaturated effects on soil behavior, and in-
crease in computing capabilities to incorporate unsaturated
relationships. This paper serves to make unsaturated soil mechan-
ics more accessible and broaden their use in practice. More and
more designs are asked to incorporate climate change effects to
predict performance decades into the future. Incorporating cli-
mate change effects into design will only serve to increase the use
of unsaturated soil mechanics in practice.
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
Acknowledgements
The author is very grateful to the Canadian Foundation for
Geotechnique and the Colloquium Selection Committee for the
honour of giving the Colloquium Lecture and opportunity to pre-
pare this manuscript. Contributions from current and past collab-
orators and graduate students have made this work possible.
Collaborators that have contributed to research on unsaturated
soils from within the GeoEngineering Centre at Queen’s–RMC
include Richard Bathurst, Ryley Beddoe, Richard Brachman, Kerry
Rowe, and Andy Take, as well as graduate students Pedram Abootalebi,
John Kingswood, Stephen Laporte, Bee Fong Lim, Jesse McCaw, Chris
Oldroyd, Stephen Peters, Suzanne Powell, Lee-Ann Sills, and Bardia
For personal use only.
Tabiatnejad. Funding from the Natural Sciences and Engineering
Research Council of Canada, Nuclear Waste Management Organiza-
tion, Royal Military College of Canada, and Canada’s Department of
National Defence is gratefully acknowledged.
References
ASTM. 2003. Standard test methods for one-dimensional swell or settlement
potential of cohesive soils. ASTM standard D4546. American Society for Test-
ing and Materials (ASTM), West Conshoshocken, Pa. doi:10.1520/D4546-03.
Aubertin, M., Cifuentes, E., Apithy, S.A., Bussière, B., Molson, J., and
Chapuis, R.P. 2009. Analyses of water diversion along inclined covers with
capillary barrier effects. Canadian Geotechnical Journal, 46(10): 1146–1164.
doi:10.1139/T09-050.
Azmatch, T.F., Sego, D.C., Arenson, L.U., and Biggar, K.W. 2012. Using soil freez-
ing characteristic curve to estimate the hydraulic conductivity function of
partially frozen soils. Cold Regions Science and Technology, 83– 84: 103–109.
doi:10.1016/j.coldregions.2012.07.002.
Barbour, S.L. 1998. Nineteenth Canadian Geotechnical Colloquium: The soil-
water characteristic curve: a historical perspective. Canadian Geotechnical
Journal, 35(5): 873–894. doi:10.1139/t98-040.
Basma, A.A., and Tuncer, E.R. 1992. Evaluation and control of collapsible soils.
Journal of Geotechnical Engineering, 118(10): 1491–1504. doi:10.1061/(ASCE)
0733-9410(1992)118:10(1491).
Bathurst, R.J., Ho, A.F., and Siemens, G. 2007. A column apparatus for investiga-
tion of 1-D unsaturated-saturated response of sand-geotextile systems.
Geotechnical Testing Journal, 30(6): 433–441. doi:10.1520/GTJ100954.
Bathurst, R.J., Siemens, G., and Ho, A.F. 2009. Experimental investigation of
infiltration ponding in one-dimensional sand-geotextile columns. Geosyn-
thetics International, 16(3): 158–172. doi:10.1680/gein.2009.16.3.158.
Beddoe, R.A., Take, W.A., and Rowe, R.K. 2011. Water-retention behavior of geo-
synthetic clay liners. Journal of Geotechnical and Geoenvironmental Engi-
neering, ASCE, 137(11): 1028–1038. doi:10.1061/(ASCE)GT.1943-5606.0000526.
Bittelli, M., Ventura, F., Campbell, G.S., Snyder, R., Gallegati, F., and Pisa, P. 2008.
Coupling of heat, water vapor, and liquid water fluxes to compute evapora-
tion in bare soils. Journal of Hydrology, 362: 191–205. doi:10.1016/j.jhydrol.
2008.08.014.
Blatz, J.A., Ferreira, N.J., and Graham, J. 2004. Effects of near-surface environ-
mental conditions on instability of an unsaturated soil slope. Canadian
Geotechnical Journal, 41(6): 1111–1126. doi:10.1139/t04-058.
Bouazza, A., Zornberg, J.G., McCartney, J.S., and Nahlawi, H. 2006. Significance
of unsaturated behaviour of geotextiles in earthen structures. Australian
Geomechanics, 41(3): 133–142.
Bozkurt, M.G., Fratta, D., and Likos, W.J. 2017. Capillary forces between equally
sized moving glass beads: an experimental study. Canadian Geotechnical
Journal, 54(9): 1300–1309. doi:10.1139/cgj-2016-0213.
Briaud, J.L., Zhang, X., and Moon, S. 2003. Shrink test–water content method for
925
shrink and swell predictions. Journal of Geotechnical and Geoenvironmental
Engineering, 129(7): 590–600. doi:10.1061/(ASCE)1090-0241(2003)129:7(590).
Brooks, R.H., and Corey, A.T. 1964. Hydraulic properties of porous media. Hydrol-
ogy Papers, Colorado State University, 24 pp.
Burdine, N.T. 1953. Relative permeability calculations from pore-size distribu-
tion data. Journal of Petroleum Technology, 5: 71–77. doi:10.2118/225-G.
Bussière, B. 2007. Colloquium 2004: Hydrogeotechnical properties of hard rock
tailings from metal mines and emerging geoenvironmental disposal ap-
proaches. Canadian Geotechnical Journal, 44(9): 1019–1052. doi:10.1139/T07-
040.
Cascini, L., Cuomo, S., Pastor, M., and Sorbino, G. 2010. Modeling of rainfall-
induced shallow landslides of the flow-type. Journal of Geotechnical and
Geoenvironmental Engineering, 136(1): 85–98. doi:10.1061/(ASCE)GT.1943-
5606.0000182.
Clemence, S.P., and Finbarr, A.O. 1981. Design considerations for collapsible
soils. Journal of Soil Mechanics and Foundation Engineering, GT3, 107:
305–317.
Costa, Y.D., Cintra, J.C., and Zornberg, J.G. 2003. Influence of matric suction on
the results of plate load tests performed on a lateritic soil deposit. Geotech-
nical Testing Journal, ASTM, 26(2): 219–227. doi:10.1520/GTJ11326J.
Cui, S., Zhang, H., and Zhang, M. 2012. Swelling characteristics of compacted
GMZ bentonite–sand mixtures as a buffer/backfill material in China. Engi-
neering Geology, 141–142: 65–73. doi:10.1016/j.enggeo.2012.05.004.
D’Arcy, H. 1856. Les fontaines publiques de la Ville de Dijon. Victor Dalmont,
Dalmont, Paris.
Dobchuk, B., Nichol, C., Wilson, G.W., and Aubertin, M. 2013. Evaluation of a
single-layer desulphurized tailings cover. Canadian Geotechnical Journal,
50(7): 777–792. doi:10.1139/cgj-2012-0119.
Domaschuk, L. 1986. Is your house suffering? University of Manitoba, Depart-
ment of Civil Engineering, 18pp. [ISBN 0921267002.]
Fall, A., Weber, B., Pakpour, M., Lenoir, N., Shahidzadeh, J., Fiscina, C.,
Wagner, C., and Bonn, D. 2014. Sliding friction on wet and dry sand. Physical
Review Letters, 112(17): 175502. [4 pp.] doi:10.1103/PhysRevLett.112.175502.
PMID:24836256.
Fredlund, D.G. 1975. Engineering properties of expansive clays. In Proceedings of
the Seminar on Shallow Foundations on Expansive Clays, Regina, Saskatche-
wan, 27–28 October 1975, pp. 1–58.
Fredlund, D.G. 1979. Second Canadian Geotechnical Colloquium: Appropriate
concepts and technology for unsaturated soils. Canadian Geotechnical Jour-
nal, 16(1): 121–139. doi:10.1139/t79-011.
Fredlund, D.G. 1983. Prediction of ground movements in swelling clays. In Pro-
ceedings of the 31st Annual Soil Mechanics and Foundation Engineering
Conference, Earle Brown Centre, University of Minnesota, Minneapolis, Min-
nesota, 1 February 1983, pp. 1–48.
Fredlund, D.G., and Xing, A. 1994. Equations for the soil-water characteristic
curve. Canadian Geotechnical Journal, 31(4): 521–532. doi:10.1139/t94-061.
Fredlund, D.G., Xing, A., and Huang, S. 1994. Predicting the permeability func-
tion for unsaturated soils using the soil-water characteristic curve. Canadian
Geotechnical Journal, 31(4): 533–546. doi:10.1139/t94-062.
Fredlund, D.G., Stone, J., Stianson, J., and Sedgwick, A. 2011. Determination of
water storage and permeability functions for oil sands tailings. In Proceedings,
Tailings and Mine Waste 2011, Vancouver, B.C., 6–9 November 2011, 15 pp.
Fredlund, D.G., Rahardjo, H., and Fredlund, M.D. 2012. Unsaturated soil mechan-
ics in engineering practice. John Wiley & Sons, Hoboken, New Jersey, 926 pp.
Green, W.H., and Ampt, G.A. 1911. Studies on soil physics. Journal of Agricultural
Science, 4(1): 1–24. doi:10.1017/S0021859600001441.
Hamilton, J.J. 1963. Volume changes in undisturbed clay profiles in western
Canada. Canadian Geotechnical Journal, 1(1): 27–42. doi:10.1139/t63-004.
Houston, S.L., Houston, W.N., Zapata, C.E., and Lawrence, C. 2001. Geotechnical
engineering practice for collapsible soils. Geotechnical and Geological Engi-
neering, 19: 333–355. doi:10.1023/A:1013178226615.
Houston, S., Dye, H., Zapata, C., Walsh, K., and Houston, W. 2011. Study of
expansive soils and residential foundations on expansive soils in Arizona.
Journal of Performance of Constructed Facilities, 25(1): 31–44. doi:10.1061/
(ASCE)CF.1943-5509.0000077.
Huang, M., Barbour, S.L., and Carey, S.K. 2015. The impact of reclamation cover
depth on the performance of reclaimed shale overburden at an oil sands
mine in northern Alberta, Canada. Hydrological Processes, 29: 2840–2854.
doi:10.1002/hyp.10229.
Iskander, M., Bathurst, R., and Omidvar, M. 2015. Past, present, and future of
transparent soils. Geotechnical Testing Journal, 38(5): 557–573. doi:10.1520/
GTJ20150079.
Ito, M., and Azam, S. 2013. Engineering properties of a vertisolic expansive soil
deposit. Engineering Geology, 152(1): 10–16. doi:10.1016/j.enggeo.2012.10.004.
Iverson, R.M. 2000. Landslide triggering by rain infiltration. Water Resources
Research, 36(7): 1897–1910. doi:10.1029/2000WR900090.
Jung, J.K., O’Rourke, T.D., and Argyrou, C. 2016. Multi-directional force-
displacement response of underground pipe in sand. Canadian Geotechnical
Journal, 53(11): 1763–1781. doi:10.1139/cgj-2016-0059.
Knidiri, J., Bussière, B., Hakkou, R., Bossé, B., Maqsoud, A., and Benzaazoua, M.
2017. Hydrogeological behaviour of an inclined store-and-release cover exper-
imental cell made with phosphate mine wastes. Canadian Geotechnical Jour-
nal, 54(1): 102–116. doi:10.1139/cgj-2015-0530.
Published by NRC Research Press
 926
Komine, H., and Ogata, N. 1994. Experimental study on swelling characteristics
of compacted bentonite. Canadian Geotechnical Journal, 31(4): 478–490. doi:
10.1139/t94-057.
Komine, H., and Ogata, N. 2003. New equations for swelling characteristics of
bentonite-based buffer materials. Canadian Geotechnical Journal, 40(2):
460–475. doi:10.1139/t02-115.
Konrad, J. 1990. Sampling of saturated and unsaturated sands by freezing.
Geotechnical Testing Journal, 13(2): 88–96. doi:10.1520/GTJ10699J.
Konrad, J.-M., and Morgenstern, N.R. 1980. A mechanistic theory of ice lens
formation in fine-grained soils. Canadian Geotechnical Journal, 17(4): 473–
486. doi:10.1139/t80-056.
Lehmann, P., Assouline, S., and Or, D. 2008. Characteristic lengths affecting
evaporative drying of porous media. Physical Review E, Statistical, Nonlin-
ear, and Soft Matter Physics, 77: 056309. [16 pp.] doi:10.1103/PhysRevE.77.
056309. PMID:18643163.
Lenhard, R.J., and Parker, J.C. 1987. Measurement and prediction of saturation-
pressure relationships in three-phase porous media systems. Journal of Con-
taminant Hydrology, 1: 407–424. doi:10.1016/0169-7722(87)90017-9.
Leong, E.C., and Rahardjo, H. 1997. Permeability functions for unsaturated soils.
Journal of Geotechnical and Geoenvironmental Engineering, 123(12): 1118–
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
1126. doi:10.1061/(ASCE)1090-0241(1997)123:12(1118).
Lim, B.F. 2014. Advanced testing and a numerical tool for swell equilibrium limit
of expansive soils. Ph.D. thesis, Department of Civil Engineering, Royal Mil-
itary College of Canada, Kingston, Ont.
Lim, B.F., and Siemens, G.A. 2013. An unconfined swelling test for clayey soils
that incorporates digital image correlation. Geotechnical Testing Journal,
36(6): 823–833. doi:10.1520/GTJ20120220.
Lim, B.F., and Siemens, G.A. 2016. Unifying framework for modeling swelling
soil behaviour. Canadian Geotechnical Journal, 53(9): 1495–1509. doi:10.1139/
cgj-2015-0049.
Lu, N., and Likos, W.J. 2004. Unsaturated soil mechanics. Wiley, 584 pp.
Machmer, B. 2012. Understanding the behavior of a pile foundation in unsatu-
rated soils subjected to lateral loading. M.Sc. thesis, Civil Engineering, Clemson
University, 184 pp.
Maxwell, J.C., and Strutt, J.W. 1911. Capillary action. In Encyclopædia Britannica,
For personal use only.
Vol. 5, 11th ed., pp. 256–275. https://en.wikipedia.org/wiki/Young–Laplace_
equation [accessed 15 August 2017].
Mualem, Y. 1976. A new model for predicting the hydraulic conductivity of
unsaturated porous media. Water Resources Research, 12(3): 513–522. doi:10.
1029/WR012i003p00513.
Newberry, P.E. 1895. El Bersheh: The tomb of Tehuti-Hetep. Egypt Exploration
Fund, London, Vol. 1.
NOAA. 2018. Costliest mainland United States tropical cyclones 1900–2017, un-
adjusted for inflation. National Oceanic & Atmospheric Administration.
Available from http://www.aoml.noaa.gov/hrd/tcfaq/costliesttable.html [ac-
cessed 24 May 2018].
O’Kane, M., Wilson, G.W., and Barbour, S.L. 1998. Instrumentation and monitor-
ing of an engineered soil cover system for mine waste rock. Canadian
Geotechnical Journal, 35(5): 828–846. doi:10.1139/t98-051.
Or, D., Lehmann, P., Shahraeeni, E., and Shokri, N. 2013. Advances in soil evap-
oration physics—a review. Vadose Zone Journal, 12(4). 16 pp. doi:10.2136/
vzj2012.0163.
Peters, S.B., Siemens, G., and Take, W.A. 2011. Characterization of transparent
soil for unsaturated applications. ASTM Geotechnical Testing Journal, 34(5):
445–456. [Special issue on innovations in characterizing the mechanical and
hydrological properties of unsaturated soil.] doi:10.1520/GTJ103580.
Powell, J.S., Siemens, G.A., Take, W.A., and Remenda, V.H. 2013. Characterizing
the swelling potential of Bearpaw clayshale. Engineering Geology, 158: 89–
97. doi:10.1016/j.enggeo.2013.03.006.
Prakash, K., and Sridharan, A. 2004. Free swell ratio and clay mineralogy of
fine-grained soils. Geotechnical Testing Journal, 27(2): 220–225. doi:10.1520/
GTJ10860.
Puppala, A.J., Manosuthikij, T., and Chittoori, B.C.S. 2014. Swell and shrinkage
strain prediction models for expansive clays. Engineering Geology, 168: 1–8.
doi:10.1016/j.enggeo.2013.10.017.
Qi, S., Simms, P., Vanapalli, S., and Soleimani, S. 2017. Piecewise-linear formu-
lation of coupled large-strain consolidation and unsaturated flow. II. Testing
and performance. Journal of Geotechnical and Geoenvironmental Engineering,
143(7): 04017019. doi:10.1061/(ASCE)GT.1943-5606.0001658.
Rahimi, A., Rahardjo, H., and Leong, E.-C. 2010. Effect of hydraulic properties of
soil on rainfall-induced slope failure. Engineering Geology, 114: 135–143.
doi:10.1016/j.enggeo.2010.04.010.
Robinson, J.D., Vahedifard, F., and AghaKouchak, A. 2017. Rainfall-triggered
slope instabilities under a changing climate: comparative study using histor-
ical and projected precipitation extremes. Canadian Geotechnical Journal,
54(1): 117–127. doi:10.1139/cgj-2015-0602.
Rogers, C.D.F., Dijkstra, T.A., and Smalley, I.J. 1994. Hydroconsolidation and
subsidence of loess: studies from China, Russia, North America and Europe:
In memory of Jan Sajgalik. Engineering Geology, 37(2): 83–113. doi:10.1016/
0013-7952(94)90045-0.
Romero, E., and Simms, P. 2008. Microstructure investigation in unsaturated
soils: a review with special attention to contribution of mercury intrusion
Can. Geotech. J. Vol. 55, 2018
porosimetry and environmental scanning electron microscopy. Geotechnical
and Geological Engineering, 26(6): 705–727. doi:10.1007/s10706-008-9204-5.
Saleh-Mbemba, F., Aubertin, M., Mbonimpa, M., and Li, L. 2016. Experimental
characterization of the shrinkage and water retention behaviour of tailings
from hard rock mines. Geotechnical and Geological Engineering, 34(1): 251–
266. doi:10.1007/s10706-015-9942-0.
Siekmeier, J. 2007. Implementation of unsaturated soil mechanics during pave-
ment construction QA. Geo-Strata - Geo Institute of ASCE, 2011, 15(1): 36, 38,
40, 43.
Siemens, G.A. 2007. Replacement of a 30-year-old shallow foundation on swell-
ing soil: a case study. In The 2nd Canadian Young Geotechnical Engineers and
Geoscientists Conference (cYGEGC), Ganonoque, Ontario, 25–27 October
2007, CD Proceedings.
Siemens, G.A., and Bathurst, R.J. 2010. Numerical parametric investigation of
infiltration in one-dimensional sand-geotextile columns. Geotextiles and
Geomembranes, 28: 460–476. doi:10.1016/j.geotexmem.2009.12.009.
Siemens, G.A., and Blatz, J.A. 2007. Triaxial apparatus for applying liquid infil-
tration with controlled boundary conditions and internal suction measure-
ment. Journal of Geotechnical and Geoenvironmental Engineering, 133(6):
748–752. doi:10.1061/(ASCE)1090-0241(2007)133:6(748).
Siemens, G.A., and Blatz, J.A. 2008. Numerical modelling of shallow foundations
on swelling clay soil using the swelling equilibrium limit. In The 1st European
Conference on Unsaturated Soils (E-UNSAT), Durham, United Kingdom, 2–5 July
2008, pp. 855–860.
Siemens, G.A., and Blatz, J.A. 2009. Evaluation of the influence of boundary
confinement on the behaviour of unsaturated swelling clay soils. Canadian
Geotechnical Journal, 46(3): 339–356. doi:10.1139/T08-125.
Siemens, G., Take, W.A., Rowe, R.K.R., and Brachman, R.W.I. 2012. Numerical
investigation of transient hydration of unsaturated geosynthetic clay liners.
Geosynthetics International, 19(3): 232–251. doi:10.1680/gein.12.00011.
Siemens, G.A., Peters, S.B., and Take, W.A. 2013. Comparison of confined and
unconfined infiltration in transparent porous media. Water Resources Re-
search, 49: 851–863. doi:10.1002/wrcr.20101.
Siemens, G.A., Take, W.A., and Peters, S.B. 2014. Physical and numerical model-
ing of infiltration including consideration of the pore-air phase. Canadian
Geotechnical Journal, 51(12): 1475–1487. [Thematic issue on unsaturated
soils.] doi:10.1139/cgj-2013-0447.
Sills, L.-A.K., Mumford, K.G., and Siemens, G.A. 2016. Quantification of fluid
saturations in transparent porous media. Vadose Zone Journal. [In press.]
(October 2016, #VZJ-2016-06-0052-ORA). doi:10.2136/vzj2016.06.0052.
Simms, P. 2017. 2013 Colloquium of the Canadian Geotechnical Society: Geotech-
nical and geoenvironmental behaviour of high-density tailings. Canadian
Geotechnical Journal, 54(4): 455–468. doi:10.1139/cgj-2015-0533.
Terzaghi, K. 1943. Theoretical soil mechanics. John Wiley & Sons, 528 pp.
Tohari, A., Nishigaki, M., and Komatsu, M. 2007. Laboratory rainfall-induced
slope failure with moisture content measurement. Journal of Geotechnical
and Geoenvironmental Engineering, 133(5): 575–587. doi:10.1061/(ASCE)1090-
0241(2007)133:5(575).
van Genuchten, M.Th. 1980. A closed-form equation for predicting the hydraulic
conductivity of unsaturated soils. Soil Science Society of America Journal, 44:
892–898. doi:10.2136/sssaj1980.03615995004400050002x.
Vanapalli, S.K., and Lu, L. 2012. A state-of-the art review of 1-D heave prediction
methods for expansive soils. International Journal of Geotechnical Engineer-
ing, 6: 15–41. doi:10.3328/IJGE.2012.06.01.15-41.
Vanapalli, S.K., and Mohamed, F.M.O. 2007. Bearing capacity of model footings
in unsaturated soils. In Experimental unsaturated soil mechanics. Edited by
T. Schanz. Springer Proceedings in Physics, Vol. 112, pp. 483–493.
Vanapalli, S.K., Fredlund, D.G., Pufahl, D.E., and Clifton, A.W. 1996. Model for
the prediction of shear strength with respect to soil suction. Canadian
Geotechnical Journal, 33(3): 379–392. doi:10.1139/t96-060.
Wijaya, M., Leong, E.C., and Rahardjo, H. 2015. Effect of shrinkage on air-entry
value of soils. Soils and Foundations, 55(1): 166–180. doi:10.1016/j.sandf.2014.
12.013.
Wilson, G.W., Fredlund, D.G., and Barbour, S.L. 1994. Coupled soil-atmosphere
modelling for soil evaporation. Canadian Geotechnical Journal, 31(2): 151–
161. doi:10.1139/t94-021.
Zhang, L.L., Zhang, J., Zhang, L.M., and Tang, W.H. 2011. Stability analysis of
rainfall-induced slope failure: a review. Proceedings of the Institution of Civil
Engineers Geotechnical Engineering, 164(GE5): 299–316. doi:10.1680/geng.
2011.164.5.299.
List of symbols
A, B SEL equation parameters
AEV air-entry value
c== apparent cohesion
D10 10% passing
e void ratio
f(Sr) function of degree of saturation
g gravity
h capillary rise height
he elevation head
Published by NRC Research Press
 Siemens 927

hp pressure head uw water pressure

ht total head V specific volume (= e + 1)
IN normalized pixel intensity V0 initial specific volume
K hydraulic conductivity VSEL SEL specific volume
Ksat saturated hydraulic conductivity Vv volume of voids
LL liquid limit VWC volumetric water content
l length w gravimetric water content
P horizontal load ␣ contact angle
p mean stress ␣1, ␣2 leading and trailing contact angles
q applied stress ␥w unit weight of water
r capillary tube radius ␧v, ⑀v volumetric strain
⌬ displacement
S settlement
␳w density of water
SEL swelling equilibrium limit
␾  friction angle
Sr degree of saturation
␴1, ␴2, ␴3 principal stresses
Ts surface tension
␴mean mean stress
ua air pressure
ua – uw matric suction
Can. Geotech. J. Downloaded from cdnsciencepub.com by 38.25.15.181 on 04/09/23
For personal use only.

Published by NRC Research Press