76 Geotechnical Journal Vol. 6 No.

1 2014

SUBGRADE SOIL STABILISATION USING NATIVE VEGETATION

M. Pallewattha1, B. Indraratna2, B. Fatahi3,C. Rujikiatkamjorn4 and U. Pathirage5
1 PhD Candidate, School of Civil, Mining & Environmental Engineering, University of Wollongong, Australia.
E-mail: map804@uowmail.edu.au
2 Professor of Civil Engineering, School of Mining and Environmental Engineering
Research Director of Centre for Geomechanics and Railway Engineering, University of Wollongong, Wollongong City, Australia,
E-mail: indra@uow.edu.au
3 Senior Lecturer of Geotechnical Engineering at University of Technology, Sydney. Australia,
E-mail: behzad.fatahi@uts.edu.au
4Associate Professor, Centre for Geomechanics and Railway Engineering, University of Wollongong, Wollongong City, Australia
5Research Fellow, School of Civil, Mining & Environmental Engineering, University of Wollongong, Australia.
ABSTRACT: Current demand of infrastructure facilities along metropolitan areas has led to construction of earth structures,
major highways and railways on soft soils. Subsequently civil engineering is in a more challenging situation to discover more
cost effective, reliable and sustainable methods for ground improvement. The green corridor concept, or ground improvement
using native vegetation, can be considered as often a more effective method. Even though this is considered as a new idea,
the use of vegetation in hill slopes to prevent erosion and to give some stability has long been known, but without proper
engineering quantification of the stabilisation. Trees are capable of increasing the matric suction of the soil subgrade underneath
the substructure throughroot water uptake, in conjunction with the tree canopy evapotranspiration. At the same time, these trees
are capable of providing significant mechanical reinforcement through the anchoring effect provided by the root network plus
the additional cohesive increment due to hair roots generating osmotic suction.
Much of the previous research carried out to quantify the mechanical strength generated by tree roots has been mainly based on
empiricism. Empirical relations have been developed for given tree species grown under known soil conditions. The extrapolation
of such empirical relations from one tree-soil system to another can be misleading. At the same time, the effect of transpiration by
the tree canopy and its influence on the sustained suction equilibrium generated at the root zone for stabilising soft subgrade has
not been considered rationally in the design of rail corridors. To accommodate the above natural phenomena, a novel and rational
computational model has been developed to quantify the overall suction effect provided by the tree roots and its continual
link with the rate and magnitude of canopy evapo-transpiration. Root based suction of a tree improves the shear strength and
accelerates the pore water pressure dissipation. In addition it may alter the potential failure conditions of the soil-root system
from a saturated to an unsaturated domain. Therefore, it is necessary for the root based suction and the mechanical properties of
the root network to be analysed within a coupled multiphase framework. Accordingly, this paper will present a model developed
to capture the root based suction and the importance of the combined effect of the root reinforcement and suction.
1 INTRODUCTION
Ground improvement using native vegetation is an economic,
environmental friendly and expedient, method which can be used in
most geotechnical engineering projects after a thorough engineering
quantification of subsequent effects. In the process of Ground
improvement using native vegetation tree roots look after the most
important fragments of the process.
Tree roots provide mechanical strengthening to the soil by the
anchoring effect of the main roots and increase of cohesion due to the
hair roots in a similar fashion to geofibres(Fatahi 2007; Fatahi et al.
2012). Moreover, tree roots increase matric suction of soil by means
of the root water uptake in conjunction with the evapotranspiration
of the tree (Cameron 2001, Fatahiet al. 2014,Potter 2006). Therefore
the true effect of the vegetation on the soil stability is accountable
by the mechanical strengthening, suction improvement as well as the
influence of the suction on mechanical strengthening.
Previous research which has been conducted to quantify the effect
of mechanical strengthening of tree roots in slope stabilisation are
mainly focused on saturated soil conditions, as well as most of this
research has been based on empiricism(Docker 2008; Docker 2009).
When a mature tree is under the evapo-transpiration process, tree
roots are capable of developing 3MPa root suction at wilting point
(Fatahi et al. 2009) and this leads to the enhancement and preserving
of the vadose zone. Therefore, most of the time, the sub-soil structure
is in the unsaturated condition and the parameters calculated for the
saturated condition may not represent realistic behaviour. In that case
the suction generated on surrounding soil by tree roots is important.
Suction development by a tree directly depends on the tree species, soil
condition and other environmental parameters such as temperature,

wind speed, humidity since the depth of the root system is determined
generally by the above conditions. Therefore a comprehensive
understanding about the root system is of utmost importance for in
depth study of the effect of suction for soil stabilisation (Cameron
2000)
Because of the complexity and unpredictability of the root system
and their subsequent effects on the surrounding soil, engineers were
reluctant to use the bio engineering method of strengthening soil using
trees in the past. However it has been proven that bioengineering
techniques can be effectively used in hill slopes where there are risk
of land slides in Australia South East Asia and other tropical regions
(Fatahi 2007)
1.1 Root Systems
The functions of roots include anchorage, the absorption of water,
minerals and nutrients, synthesis of various essential compounds such
as growth regulators and the storage of food in root crops (Weaver
1958). At the same time tree roots give more strength to the adjacent
soil by means of mechanical strengthening and also by increasing
suction. When soil is mechanically strengthened or strengthened with
suction effects, the shape of the root system affects directly. It is a
known fact that different tree species have different root systems and
Soil profile, climate and water availability also affect the depth and
the spatial distribution of the root system. Figure 1(a) and (b) show
the two main root systems distinguished in trees.
Many root systems in trees in well-drained soil look like the shape in
figure 1(a). There is a combination of lateral and oblique roots and no
real tap root. Only a small percentage of tree root systems look like
the shape in figure 1(b). (Kramer 1995)

rd
Journal Vol. 6 No. 1 2014
3Geotechnical
Proff 18-02-2015

Geotechnical Journal Vol. 6 No. 1 2014 77

(a)

(b)
Figure 1 (a) Fibrous root system (b) Tap root system
Aeration
Aeration
1.2 Environmental Factors Affecting the Root Growth
1.2 Environmental Factors Affecting the Root Growth
Environmental factors which affect the growth of tree roots are soil
texture and structure; aeration; moisture; temperature; pH; salinity;
Environmental
affect
theas
growth
tree roots
are soil
the
presence offactors
toxic which
elements
such
lead, of
copper,
aluminium;
texture and structure;
pH; salinity;
competition
with otheraeration;
plants andmoisture;
presencetemperature;
of bacteria, fungi
and soil
the presence
of toxic
elements
inhabiting
animals.
(Weaver
1958)such as lead, copper, aluminium;
competition with other plants and presence of bacteria, fungi and
soil inhabiting
Soil
texture andanimals.
structure(Weaver 1958)
Soil texture and structure
Root growth is affected by properties of the soil, directly due to root
penetration
indirectly of
due
the directly
aerationdue
andtowater
Root growthrestrictions
is affected and
by properties
thetosoil,
root
content.
Figure
2 shows and
the different
roots
due
penetration
restrictions
indirectlyspatial
due todistribution
the aerationofand
water
to
different
compaction
in soils.spatial distribution of roots due
content.
Figure
2 shows levels
the different
to different compaction levels in soils.

To have
spatial
distribution
of rootofsystem
exchange
havegood
good
spatial
distribution
root proper
systemgas
proper
gas
in
the soilinhas
be confirmed.
Pore structure
the inside
soil hasthea
exchange
thetosoil
has to be confirmed.
Poreinside
structure
soil has
a great
influence
on inside
gas exchange
inside the poor
soil.
great
influence
on gas
exchange
the soil. Furthermore
Furthermore
oxygen
levels
theroot
soilgrowth
leads toaspoor
growth
oxygen
levelspoor
in the
soil leads
to in
poor
wellroot
as the
poor
as well aslevel
the in
poor
soil canfixation
limit the
nitrogen
thenitrogen
soil can level
limit in
thethe
nitrogen
of nitrogen
roots by
fixation of
roots by
legume. Weaver (1958)
legume.
Weaver
(1958)
Temperature
Temperature
Optimum temperature which is suitable for vegetation makes
Optimum spatial
temperature
whichof is
vegetation
maximum
distribution
rootsuitable
system. for
Figure
3 clearly makes
shows
maximum
distribution
of root
system.
Figure 3 clearly shows
the
effect ofspatial
temperature
on shape
of the
root system.
the effect of temperature on shape of the root system.

1000C

1500C 2000C

2500C

3000C

3500C

Figure 3. Influence of root zone temperature on root morphology

and shoot growth of potato seedlings (modified after Sattelmacher et
al. 1990)
Figure 2. Root system of young barley plants grown in the field in
-3
soils with different bulk densities (Left) 1.35 g cm -3; (right) 1.50 g
-3
-3
cm . (modified after Gilmen 1980)

Geotechnical Journal Vol. 6 No. 1 2014

78 Geotechnical Journal Vol. 6 No. 1 2014

Moisture
The Growth
Growth of
The
of the
the root
root system
system is
is directly
directly affected
affected by
by the
the amount
amount of
of
soil water
Excess or
of soil
soil water
water reduces
soil
water availability.
availability. Excess
or deficiency
deficiency of
reduces the
the
growth of
of root.
Excess water
water limits
root growth
growth due
due to
growth
root. Excess
limits root
to reduction
reduction of
of
oxygen
level
in
soil
and
deficient
water
levels
can
cause
oxygen level in soil and deficient water levels can cause reduction
reduction
as well as the cessation of root growth. Moreover the occurrence of
as well as the cessation of root growth. Moreover the occurrence of
the wilting level leads to the permanent death of the root system.
the wilting level leads to the permanent death of the root system.
Therefore the soil water potential is a dominant factor in root
Therefore the soil water potential is a dominant factor in root
distribution. Figure 4 shows the root depth variation due to the
distribution.
Figure
4 shows the root depth variation due to the
different rail fall
precipitations.
different rail fall precipitations.

Figure 4. The effect of the amount of rainfall on the depth of rooting

of winter wheat (very fine sandy loam - silty loam in the Great
Plains)(modified after Weaver 1958)
Root competition
Root competition
The size of the root system depends on the availability of adjacent
The size
of presence
the root system
dependsinonroot
the systems
availability
plants.
The
of competition
tendsoftoadjacent
reduce
plants. The presence of competition in root systems tends to reduce
the spatial retribution of roots. Although roots often seem to be
the spatial retribution of roots. Although roots often seem to be
intertwined in the soil, there is evidence of a mechanism that prevents
intertwined in the soil, there is evidence of a mechanism that
root from some kind of plants growing very close to one another.
prevents root from some kind of plants growing very close to one
Waisel
al. (2002)
another.etWaisel
et al. (2002)
1.3 Measurements of Root Systems
1.3 Measurements
of Root Systems
Since
there is a high variance
of root systems even in the same tree
species,
measuring
of
the
root
system
is more even
important
research
Since there is a high variance of
root systems
in theinsame
tree
projects
so that root
mayisbe
reasonably
by
species, measuring
of architecture
the root system
more
importantpredicted
in research
practitioners.
Most
wellarchitecture
known methods
rootpredicted
systems are
projects so that
root
may of
bemeasuring
reasonably
by
excavation,
profile
wall, of
glass
wall androot
radioactive
practitioners.auger,
Mostmonolith,
well known
methods
measuring
systems
tracers.
Other thanauger,
radioactive
traces,profile
these methods
involve
tedious
are excavation,
monolith,
wall, glass
wall
and
excavation
these
methods which
explained
in
radioactive process.
tracers. Some
Other ofthan
radioactive
traces,are
these
methods
Bhm
andexcavation
Waisel et al.
(2002)Some
are asof
follows;
involve(1980)
tedious
process.
these methods which
are explained in Bhm (1980) and Waisel et al. (2002) are as
Excavation
follows;
Excavation is good for trees and shrubs on stiff and dry sandy soils
than it is for grass or annual corps. A deep trench is dug with vertical
Excavation is good for trees and shrubs on stiff and dry sandy soils
sides in some distance from root and soil is removed applying
than it is for grass or annual corps. A deep trench is dug with
compressed
air in
parallel
the root from
since root
maximum
force
vertical sides
some todistance
root exerts
and soil
is removed
parallel
their growthairdirection.
figures
applyingtocompressed
parallel While
to thedoing
root so,
since
root should
exerts
be
drawn and
should
be taken
to interpret
the more
maximum
forcephotographs
parallel to their
growth
direction.
While doing
so,
reasonable
root
system.
figures should be drawn and photographs should be taken to
interpret the more reasonable root system.
Monolith
Monoliths of root system are to be taken away through the side of a
trench which is 1m long and deepened to the maximum root depth.
Metal sheets are driven into the root system through the trench wall

Monolith
are
used to extract the monoliths. Then soil is separated from the roots
by washing and photographs are taken.
Monoliths of root system are to be taken away through the side of a
trench which is 1m long and deepened to the maximum root depth.
Auger
Metal sheets are driven into the root system through the trench wall
arehand
usedauger
to extract
monoliths.
Thenmeans
soil isisseparated
from the
the
A
or anytheother
mechanical
used to obtain
roots by with
washing
photographs
taken.
samples
leastand
disturbance
to theare
root
system. Samples are broken
horizontally
and washing is done to separate root from soil.
Auger
A handwall
auger or any other mechanical means is used to obtain the
Profile
samples with least disturbance to the root system. Samples are
This
method
root mapping
is done
the trench
wall
,which
broken
horizontally
and washing
is through
done to separate
root
from
soil. is
dug to the required length and depth. No collection of roots is done
Profile
wall
and
if the
dry weights of the roots are required, another method has
to
be
incorporated.
This method root mapping is done through the trench wall ,which is
dug to the required length and depth. No collection of roots is done
Glass
and if wall
the dry weights of the roots are required, another method has
to be incorporated.
A glass wall is placed along the trench wall contacted with the root
Glass wall
system
and root mapping is done.
A glassactive
wall traces
is placed along the trench wall contacted with the root
Radio
system and root mapping is done.
A
radioactive
activetraces
tracer to be injected to the stem of the plant and soilRadio
root sample is taken to measure the radioactivity.
A radio active tracer to be injected to the stem of the plant and soilTrees
grownisintaken
different
areas with
different environmental factors
root sample
to measure
the radioactivity.
display quite different root systems, Nevertheless the root based
Trees grown
areas
with on
different
environmental
factors
suction
whichinisdifferent
able to be
applied
the adjacent
soil system
by
root systems,
Nevertheless
thethe
root
based
adisplay
maturequite
tree, different
is approximately
10 times
greater than
practical
suction which
is able to be vertical
applied on
the (Indraratna
adjacent soiletsystem
by a
vacuum
for prefabricated
drain
al. 2008).
mature
tree,
is
approximately
10
times
greater
than
the
practical
Therefore this paper aims to present the model developed to capture
vacuum
for prefabricated
vertical drain
(Indraratna et al. and
2008).
the
root based
suction in conjunction
with evapotranspiration
the
Therefore this
paper
aims to effect
present
developed
to suction.
capture
importance
of the
combined
of the
rootmodel
reinforcement
and
the root based suction in conjunction with evapotranspiration and
importance
of the combined
of root
reinforcement and
2theSUCTION
EFFECT
OF TREEeffect
ROOTS
ON SOIL.
suction.
Root water uptake of trees increases the metric suction of adjacent
soil due to reduction of the moisture content. Some tree varieties
2 Pinusradiata
SUCTION EFFECT
TREE
ROOTS
SOIL.
like
can absorbOF
water
from
soil upON
to its
own weight
per day (Teskey, and Sheriff 1995)and some of the mature trees can
Root water uptake of trees increases the metric suction of adjacent
generate suction in soil- root systems up to 3MPa (Fatahi 2007). The
soil due to reduction of the moisture content. Some tree varieties
main
factor that affects
the root
water
of transpiration
like Pinusradiata
can absorb
water
fromuptake
soil upistorate
its own
weight per
of
a
tree,
since
the
volume
of
water
consumed
by
plant trees
cells can
for
day (Teskey, and Sheriff 1995)and some of the mature
metabolism
is
negligible
compared
to
the
total
root
water
uptake
generate suction in soil- root systems up to 3MPa (Fatahi 2007). The
(Radclife
et that
al 1980).
rate water
of transpiration
mainly
depends on
main factor
affects The
the root
uptake is rate
of transpiration
environmental
parameters
and
tree
physiology.
of a tree, since the volume of water consumed by plant cells for
metabolism is negligible compared to the total root water uptake
2.1
Transpiration
of a tree
(Radclife
et al 1980).
The rate of transpiration mainly depends on
environmental parameters and tree physiology.
Transpiration of a tree is affected by the environmental factors
like
temperature
andwind speed, as well as soil moisture
2.1 humidity,
Transpiration
of a tree
condition (soil water potential) and tree physiology.
Transpiration of a tree is affected by the environmental factors like
humidity, istemperature
as and
soil ismoisture
Humidity
the amountandwind
of waterspeed,
vapourasin well
the air
usually
condition (soil
potential)
andpressure
tree physiology.
described
as water
vapour
density;
or relative humidity.
Temperature and wind speed also affect the relative humidity.
Humidity is the amount of water vapour in the air and is usually
According to Ficks law of diffusion, the rate of transpiration can be
described as vapour density; pressure or relative humidity.
stated
as directly
the affect
difference
vapourhumidity.
pressure
Temperature
and proportional
wind speed toalso
the in
relative
between
the
leaf
and
the
atmosphere
and
inversely
proportional
to the
According to Ficks law of diffusion, the rate of transpiration can
be
summation
of the resistance
to water
encountered
in air (Fatahiet
stated as directly
proportional
to theflow
difference
in vapour
pressure
al.
2014).the leaf and the atmosphere and inversely proportional to
between
the
summation
of the resistance
to water
in air
Temperature
regulates
the transpiration
to flow
high encountered
extend through
its
(Fatahiet
al.
2014).
effect on vapour pressure, if there is no any inhabitant acclimatization
in
the plant species.
leaftranspiration
exposed to to
fullhigh
sunlight
actually
Temperature
regulatesAthe
extendmay
through
its
reach
100C higher
than ambient
air. is
If stomata
remains
open
effect 50C
on tovapour
pressure,
if there
no any
inhabitant
transpiration
will
in aspecies.
leaf even
at 100%
relative
acclimatization
in occur
the plant
A leaf
exposed
to fullhumidity
sunlight
and
condensed
the leaf.
is a
may vapour
actuallyget
reach
50C to once
100C comes
higher out
thanfrom
ambient
air. IfThis
stomata
usual
phenomenon
in tropical
as steaming
jungle
remains
open transpiration
will jungles
occur incalled
a leaf even
at 100% relative
(Hopkins 1999)

This is a usual phenomenon in tropical jungles called as steaming

jungle
jungle (Hopkins
(Hopkins 1999)
1999)
Wind
speed
controls
through
the
of
humidity
andcontrols
vapour transpiration
get condensed
once comes
out from
the leaf.
Wind
speed
transpiration
through
the change
change
of resistance
resistance
and
effective
length
of
diffusion
path
of
water
vapour
(Nobel
rd
humidity
and
vapour
get
condensed
once
comes
out
from
the
leaf.
ThisProff
is a usual
phenomenon
in tropical
called
as steaming
humidity
and
vapour
condensed
comes
out from
the1991).
leaf.
effective
length
ofget
diffusion
path once
of jungles
water
vapour
(Nobel
1991).
3and
18-02-2015
In
higher
wind
speeds,
the
rate
of
transpiration
increases
due
to
This
is
a
usual
phenomenon
in
tropical
jungles
called
as
steaming
jungle
(Hopkins
1999)
This
is a usual
in of
tropical
jungles increases
called as due
steaming
In
higher
wind phenomenon
speeds, the rate
transpiration
to the
the
depletion
of
diffusion
jungle
(Hopkins
jungle
(Hopkins
1999)
depletion
of the
the 1999)
diffusion path
path and
and vice-versa.
vice-versa. Even
Even though
though the
the
Wind
speed controls
transpiration
through
the change
of
resistance
transpiration
of
by
speed
transpiration
transpiration
of tree
tree leaves
leaves are
are affected
affected
by wind
wind
speed of
transpiration
Wind
speed
controls
through
thevapour
change
resistance
and
effective
lengthtooftranspiration
diffusion
path
ofacclimatization
water
(Nobel
1991).
is
also
subjected
the
inhabitant
in
is also
subjected
toof the
inhabitant
acclimatization
in the
the plant
plant
and
effective
length
path
of
(Nobel
In
higher
wind
speeds,
the rate of
transpiration
increases
due 1991).
to the
and
effective
length
of diffusion
diffusion
path
of water
water vapour
vapour
(Nobel
1991).
species.
species.
higher
wind
speeds,
the
rate
ofand
transpiration
increases
duedue
to the
the
In
higher wind
wind
speeds,
therate
rateof
of
transpiration
increases
to
depletion
of
thespeeds,
diffusion
path
vice-versa.increases
Even
though
higher
the
transpiration
due
to
Soil
condition
affects
the
rate
of
transpiration
of
aathough
tree.
depletion
of
the
diffusion
path
and
Even
the
transpiration
tree
leaves
are
affected
by wind speed
the
ofalso
the
diffusion
path
and
vice-versa.
Even
though
the
depletion
of of
the
diffusion
path
and
vice-versa.
Even
though
Soildepletion
condition
also
affects
the
rate
ofvice-versa.
transpiration
oftranspiration
tree. Soil
Soil
water
inversely
the
root
water
uptake
(S).
transpiration
of tree
tree
are
by
wind
transpiration
is
alsopotential
subjected
toleaves
the affects
inhabitant
acclimatization
the plant
leaves
are affected
affected
byof
speed
transpiration
of
transpiration
water
potential
inversely
affects
the rate
rate
ofwind
rootspeed
waterin
uptake
(S).
Soil
suction
relates
the
soil
moisture
content
through
the
soil
is
also
subjected
toto
the
inhabitant
acclimatization
in the
the
plant
species.
also
subjected
to
the
acclimatization
in
plant
Soil
suction
relates
toinhabitant
the inhabitant
soil acclimatization
moisture
content
through
the
soil
is
also
subjected
to the
in the
plant
species.
water
characteristic
curve.
Therefore
water
uptake
reduction
factor
species.
species.
water characteristic curve. Therefore water uptake reduction factor
Soil
conditionofalso
also
affectscontent
the rate
rate
of
transpiration
of al.
tree.
Soil
is
moisture
of
the
soil
2010).
A
Soil
condition
affects
the
of
aaa tree.
Soil
is aa function
function
ofalso
moisture
content
of of
thetranspiration
soil (Fatahiet
(Fatahiet
al.
2010).
A
Soil
condition
affects
the
rate
transpiration
of
tree.
Soil
water
potential
inversely
affects
the of
rate
of for
rootroot
water
uptake
(S).
Soil
condition
also
affects
the
rate
of
transpiration
of
a
tree.
Soil
model
suggested
by
Feddes
et
al.
(1974)
water
uptake
water
affects
rate
of
root
water
uptake
(S).
model potential
suggestedinversely
by Feddes
et the
al.
(1974)
for
root
water
uptake
water
potential
inversely
affects
the
rate
of
root
water
uptake
(S).
Soil
suction
relates
to
the
soil
moisture
content
through
the
soil
water
potential
inversely
affects
the
rate
of
root
water
uptake
(S).
reduction
factor
in
to
soil
moisture
content
is
shown
in
Soil
suction
relates
thethe
soilsoil
moisture
content
through
the
water
reduction
factor
intorelation
relation
to
soil
moisture
content
is soil
shown
in
Soil
suction
relates
to
moisture
content
through
thefactor
soil
water
characteristic
curve.
watercontent
uptake through
reductionthe
Soil
suction
relates
to
the Therefore
soil moisture
soil
Figure
5.
Figure
5.
characteristic
curve.
Therefore
water
uptake
reduction
factor
is
a
water
characteristic
curve.
Therefore
water
uptake
reduction
factor
is a function
of moisture
of the
soiluptake
(Fatahiet
al. 2010).
A
water
characteristic
curve.content
Therefore
water
reduction
factor
function
of
moisture
content
of
the
soil
(Fatahiet
al.
2010).
A
model
is
function
of moisture
moisture
content
of (1974)
the soil
soil for
(Fatahiet
al. 2010).
2010).
A
model
suggested
by Feddes
et al.
root water
uptake
is
aa function
of
content
of
the
(Fatahiet
al.
A
suggested
by Feddes
et
al. (1974)
root water
uptake
reduction
model
suggested
Feddes
al.
(1974)
for
root
water
uptake
reduction
factor
inby
toet
moisture
is shown
in
model
suggested
byrelation
Feddes
et soil
al.for
(1974)
forcontent
root
water
uptake
factor
moisture
content
is shown
in Figure
5.
reduction
factor
in
relation
to
moisture
content
is
in
Figurein5.relation
reduction
factor to
in soil
relation
to soil
soil
moisture
content
is shown
shown
in
Figure
5.
Figure 5.

Rate
Rate
Rate
of
ofRoot
of
Root
Root
Rate
Water
Rate
Water
Water
ofofRoot
Uptake
Uptake
Root
Uptake
Water
Water
(S)
(S)(S)
Uptake
Uptake(S)
(S)

S
Smax
max
Smax
Smax
max

ww
Moisture
Moisture content
content ()
()

w
ww

dd

sat
an
an sat

d
dd

an sat
an
an sat
sat

Figure 5.The
5.The water
water
uptake content
moisture
content relationship
relationship Feddes
Feddes et
et
Moisture
()content
Figure
uptake
moisture
Moisture
content
()
Moisture
content
()
al. (1974)
(1974)
al.
the
content
at
point,
d is
is the
the
in this
this 5.The
Figure,water
in
Figure,
wwwis
isuptake
the moisture
moisture
content
Figure
moisture
content
relationship
et
is
the
moisture
content
at wilting
wilting
point, Feddes
dd

is
the
maximum
water
minimum
moisture
content
at
S
=
S
Figure
5.The
water
uptake

moisture
content
relationship
Feddes
et
max,
an
minimum
moisture
content at
SS == SSmax
, anan isrelationship
water
Figure 5.The
water uptake
moisture
content
Feddes
et
the
maximum
max,
al. (1974)at maximum root water uptake. There are numerous
content
content
at
maximum
root
water
uptake.
There
are
numerous
equations
at
maximum
root
water
uptake.
There
are
numerous
al.
(1974)
al. (1974) to quantify root water uptake reduction factor due to soil
equations
equations
quantify
water
uptakefactor
reduction
due to soil
to
quantifytoroot
water root
uptake
reduction
due tofactor
soil suction.
in this Figure, wis the moisture content at wilting point, d is the
suction.
suction.
is
the
moisture
content
at
wilting
point,
dd is
the
in
this
Figure,

the
maximum
water
minimum
moisture
atthe
S =
Smax,at
is content
theaffects
moisture
content
wilting
point,
is
the
in thisphysiology
Figure,
wwalso
an is
Tree
transpiration.
Factors
controlling
Tree
physiology
also
affects
the
transpiration.
Factors
controlling

is
the
maximum
water
minimum
moisture
content
at
S
=
S
max,
an
content
at
maximum
root
water
uptake.
There
are
numerous

is
the
maximum
water
minimum
moisture
content
at
S
=
S
Tree
physiology
also
affects
the
transpiration.
Factors
controlling
max,
an
transpiration
include
leaf
area,
number
of stomata
in aaare
leaf
and
other
transpiration
include
leaf
area,
number
of
stomata
in
leaf
and
content
at
maximum
root
water
uptake.
There
numerous
equations
quantify
root
water
uptake
soil
content
atto
maximum
root
water
uptake.
There
numerous
transpiration
include
area,
number
ofreduction
stomata
infactor
aare
leafdue
andatoother
other
biological
features
ofleaf
the
tree,.
The
total
transpiration
rate
of
plant
biological
features
of
the
tree,.
The
total
transpiration
rate
of
a
equations
to
quantify
root
water
uptake
reduction
factor
due
to
soil
suction.
equations
to
quantify
roottree,.
waterThe
uptake
reduction
factor
due
toplant
soil
biological
features
of
the
total
transpiration
rate
of
a
plant
can
be
calculated by
adding allallthethe
transpiration rates
of leaves
on the
be
suction.
suction.
can
be calculated
calculated by
by adding
adding all the transpiration
transpiration rates
rates of
of leaves
leaves on
on
plant
which
meansalso
the higher
the
leaf
area,
the
rateFactors
of transpiration
is
Tree plant
physiology
affects
the
transpiration.
controlling
the
which
means
the
higher
the
leaf
area,
the
rate
of
the
plant
which
means
the
higher
the
leaf
area,
the
rate
of
Treehigher.
physiology
also
affects
the
transpiration.
Factors
controlling
transpiration
include
leaf
area,
number
of
stomata
inthe
a leaf
and other
also
Based
onhigher.
Green
(1993),
the
transpiration
of
atranspiration
whole
plant
Tree
physiology
also
affects
the
transpiration.
Factors
controlling
is
also
Based
on
Green
(1993),
transpiration is
also higher.
Based on Green
(1993), the
transpiration
transpiration
include
area,
of
stomata
leaf
and
biological
features
ofleaf
the
tree,. number
Theusing;
total
plant
transpiration
include
leaf
area,
number
oftranspiration
stomata in
in aa rate
leaf of
anda other
other
can
be calculated
using;
of
plant
calculated
of aa whole
whole features
plant can
canofbe
bethe
calculated
using;
biological
tree,.
The
total
transpiration
rate
of
can
be
calculated
by
adding
all
the
transpiration
rates
of
leaves
on
biological features of the tree,. The total transpiration rate of aa plant
plant
can
calculated
adding
the
rates
the be
plant
which by
means
theall
the leaf area,
theleaves
rate on
of
can
be
calculated
by
adding
allhigher
the transpiration
transpiration
rates of
of
leaves
on
the
plant
means
the
higher
the
leaf
area,
the
rate
of
transpiration
is also
higher.
Based
on
Green
(1993),
the
transpiration
the
plant which
which
means
the
higher
the
leaf
area,
the
rate
of

(1)

transpiration
is
Based
on
the
(1)
))
of a whole plant
canhigher.
be calculated
transpiration
is also
also
higher.
Based(( using;
on Green
Green (1993),
(1993),
the transpiration
transpiration
of
of aa whole
whole plant
plant can
can be
be calculated
calculated using;
using;

(
((

)
))

(1)
(1)
(1)

is the net radiation flux density

total leaf area of the canopy,
absorbed
is
is
absorbed by
by each
each leaf,
leaf,
is the
the vapour
vapour pressure
pressure deficit
deficit of
of air,
air,
is
the
boundary
layer
resistance
of
each
leaf,
is
the
stomatal
the
boundary
layer
resistance
of each
leaf,
is
the
stomatal
is
the
fractional
expression
of
each
leaf
in
terms
of
the
Where,
f
i of each leaf,
resistance
is
the
of
saturation vapour
pressure
resistance
of the
each
leaf,canopy,
isexpression
the slope
slope
of
vapour
is
fractional
of
each
leaf
in
terms
of
Where,
ffii area
is the
net is
radiation
fluxpressure
density
total
leaf
of
the
is
the
fractional
expression
of saturation
each
leaf
inpsychrometric
terms
of the
the
Where,
curve
at
the
ambient
air
temperature,
the
curve
at
the
ambient
air
temperature,
is
the
psychrometric
is
the
net
radiation
flux
density
total
leaf
area
of
the
canopy,
is
the
net
radiation
flux
density
total
leaf
area
of
the
canopy,
absorbed
by
each
leaf,
is
the
vapour
pressure
deficit
of
air,
is
Geotechnical
Journal
Vol.
6
No.
1
2014
constant,
is
the
air
density
and
is
the
specific
heat
capacity 79
of
constant,
is layer
theleaf,
airresistance
density
andvapour
is pressure
theleaf,
specific
heat
of
absorbed
by
is
deficit
of
air,
is
absorbed
by each
each
leaf,
is the
the
vapour
pressure
deficit
ofcapacity
air,
is
the atboundary
of each
is the
stomatal
air
constant
pressure.
air
atboundary
constant
pressure.
the
layer
resistance
of
each
leaf,
is
the
stomatal
the
boundary
layer
resistance
of
each
leaf,
is
the
stomatal
resistance
of
each
leaf,
is
the
slope
of
saturation
vapour
pressure
Where, fi is the fractional expression of each leaf in terms of the total
resistance
of
each
leaf,
of
vapour
pressure
curvearea
at ofthe
ambient
temperature,
isflux
thedensity
psychrometric
resistance
ofthe
each
leaf, Rairis
isisthe
the
slope
of saturation
saturation
vapour
pressure
leaf
canopy,
theslope
net radiation
absorbed
n,i
curve
at
the
ambient
air
temperature,
is
the
psychrometric
constant,
is
the
air
density
and
is
the
specific
heat
capacity
of
curve
at
the
ambient
air
temperature,
is
the
psychrometric
by
each
leaf,
D
is
the
vapour
pressure
deficit
of
air,
r
is
the
boundary
2.2
Rate
Water
Uptake
a
2.2
Rate of
ofis
Root
Water
Uptake
constant,
isRoot
the air
air
density
and
is the
the specific
specifica,iheat
heat capacity of
of
constant,
the
is
air at resistance
constant
pressure.
layer
of eachdensity
leaf, rs,iand
is the stomatal
resistance ofcapacity
each leaf,
air
at
constant
pressure.
Rate
of
root
water
uptake
governs
the
suction
generation
on
soil
by
air
at
constant
pressure.
S
is the
slopewater
of saturation
vapourthe
pressure
at the on
ambient
Rate
of root
uptake governs
suctioncurve
generation
soil byairaa
tree.
is
by
the
tree.
tree. Root
Root water
water
uptake
is done
done only
onlyconstant,
by active
activearoots
roots
ofair
thedensity
tree. Tree
Tree
temperature,
is uptake
the psychrometric
is theof
and
roots
can
be
broadly
classified
into
two;
(a)
active
roots
(b)
2.2
Rate
of
Root
Water
Uptake
roots
can
be
broadly
classified
into
two;
(a)
active
roots
(b)
C
is
the
specific
heat
capacity
of
air
at
constant
pressure.
p
anchoring
roots.
Active
roots
are
the
smaller
roots
and
absorb
water
2.2
Rate
of
Root
Water
Uptake
2.2
Rate
of
Root
Water
Uptake
anchoring roots. Active roots are the smaller roots and absorb water
RateRate
of root
waterWater
uptake
governs
the suction
generation
on soil
by a
and
other
through
an
process.
Anchoring
roots
2.2
ofnutrients
Root
Uptake
and
other
nutrients
through
an osmotic
osmotic
process.
Anchoring
rootsbydo
do
Rate
of root
root
water
uptake
governs
the
suction
generation
ontree.
soil
tree.
Root
water
uptake
is governs
done
only
by
activeof
roots
of
the
Tree
Rate
of
water
uptake
the
suction
generation
on
soil
byinaa
the
main
anchoring
process
of
the
tree.
Rate
root
water
uptake
the
main
anchoring
process
of
the
tree.
Rate
of
root
water
uptake
in
tree.
Root
water
uptake
isclassified
done
onlyinto
by active
active
roots
of the
the according
tree.
Tree
roots
be
broadly
two;
(a) system
active
roots
Root
water
uptake
is
done
only
by
roots
of
tree.
atree.
given
time
varies
from
aagoverns
point
point
in
root
Rate
ofcan
root
water
uptake
suction
onaccording
soil Tree
by(b)a
aroots
given
timebe
varies
fromclassified
point to
tothe
point
in aa generation
root
system
can
broadly
into
two;
(a)
active
roots
(b)
anchoring
roots.
Active is
roots
theinto
smaller
roots
and
absorb
roots
can
be distribution
broadly
classified
two;
(a) Therefore
active
roots
(b)
to
the
spatial
ofare
the
active
roots.
aa water
clear
tree.
Root
water
uptake
done
only
by
active
roots
of the
tree.
Tree
to
the
spatial
distribution
the
active
roots.
Therefore
clear
anchoring
roots.
Active
rootsof
are
the
smaller
roots
and absorb
absorb
water
and
other
nutrients
through
anare
osmotic
process.
Anchoring
roots
do
anchoring
roots.
Active
roots
the
smaller
roots
and
water
interpretation
about
the
active
root
distribution
is
essential.
roots
can be
broadly
classified
into
two;
(a)
active
roots
(b) anchoring
interpretation
about
the
active
root
distribution
is
essential.
and
other
nutrients
through
an
osmotic
process.
Anchoring
roots
do
the
main
anchoring
process
of
the
tree.
Rate
of
root
water
uptake
in
and
other
nutrients
through
an
osmotic
process.
Anchoring
roots
do
roots.
Active
rootsrecent
are
the
smaller
roots
and absorb
water
and
other
Considering
the
researches
by
Docker
and
Hubble
(2001),
the
main
anchoring
process
of
the
tree.
Rate
of
root
water
uptake
in
aConsidering
given
varies
from
a
point
to
point
in
a
root
system
according
the
maintime
anchoring
process
of
the
tree.
Rate
of
root
water
uptake
in
the
recent
researches
by
Docker
and
Hubble
(2001),
nutrients
through
anfrom
osmotic
process.
Anchoring
roots
doaccording
the main
Dabson
and
Moffat
(1995),
(2007)
the
following
aatogiven
time
varies
aa point
to
in
aaproposed
root
system
the spatial
distribution
ofFatahi
the
active
a clear
given
time
varies
from
point
to point
point
inroots.
rootTherefore
system
according
Dabson
and
Moffat
(1995),
Fatahi
(2007)
proposed
the
following
anchoring
process
of
the tree.
Rate
of root
water Therefore
uptake
in aa given
schematic
diagram
for
shape
aa maximum
root
is
to
distribution
of
the
active
roots.
interpretation
about
theroot
active
distribution
is essential.
to the
the spatial
spatial
distribution
ofroot
thewith
active
roots.
Therefore
a clear
clear
schematic
diagram
for
root
shape
with
maximum
root density
density
is aa
time
varies
from
a
point
to
point
in
a
root
system
according
to the


at
depth
of

and
an
exponential
circle
with

interpretation
about
the
activeofroot
root
distribution
is essential.
essential. reduction
interpretation
about
active
is
at depth
distribution
reduction
circle
with 
the
and an exponential
Considering
the
recent
Docker and
Hubble
(2001),
spatial
distribution
of theresearches
active
roots.by
Therefore
a clear
interpretation
in
in
vertical
and
directions.
in root
root density
densitythe
in recent
verticalresearches
and radial
radial by
directions.
Considering
Docker
and Hubble
Hubble
(2001),
Dabson
Moffat
(1995),
Fatahi
(2007)
proposed
the following
Considering
the
recent
researches
by
Docker
and
(2001),
about
theand
active
root
distribution
is essential.
Dabson
Moffat
Fatahi
(2007)
proposed
the
following
schematicand
diagram
root shape
with
a maximum
root
is a
Dabson
and
Moffatfor(1995),
(1995),
Fatahi
(2007)
proposed
thedensity
following
schematic
for
root
shape
with
aaDocker
maximum
density
is
Considering
the
recent
researches
and root
Hubble
(2001),

at depth
of
and
an exponential
reduction
circle with diagram
schematic
diagram
for
root
shape
with
maximum
root
density
is aa
by
at
depth

and
exponential
reduction
circle
Dabson
and
Moffat
Fatahi
(2007)
proposed
the following
in rootwith
density
in
vertical
andof
radial


at(1995),
depth
of
directions.

and an
an
exponential
reduction
circle
with
in
in
and
radial
directions.
schematic
diagram
for root
a maximum root density is a
in root
root density
density
in vertical
vertical
andshape
radialwith
directions.
circle with r= r0(t) at depth of z=z0(t) and an exponential reduction in
root density in vertical and radial directions

Figure
Figure 6.
6. Schematic
Schematic diagram
diagram for
for soil-plant
soil-plant atmosphere
atmosphere system
system
(Fatahi
2007)
(Fatahi 2007)
Fatahi
Fatahi et
et al.
al. (2010)
(2010) developed
developed an
an equation
equation for
for root
root water
water uptake
uptake at
at
any
point
of
the
root
zone
based
on
the
potential
transpiration
of
Figure
6.
Schematic
diagram
for
soil-plant
atmosphere
any point of the root zone based on the potential transpirationsystem
of the
the
tree
aa reduction
factor
due
suction;
Figure
6.
Schematic
diagram
for
soil-plant atmosphere
atmosphere system
system
(Fatahi
Figure
6.
Schematic
diagram
soil-plant
tree and
and2007)
reduction
factor
due to
to for
suction;
(Fatahi
2007)
(Fatahi 2007) (
)
(
)
(
)
(2)
(
)
Fatahi et al. (2010)
( developed
)
(an )equation
( ) ( for) root water uptake
(2) at
Fatahi
et
al.
(2010)
developed
an
equation
for
root
water
uptake
at
Fatahi
et
al.
(2010)
developed
an
equation
for
root
water
uptake
at
any point
of the
rootdeveloped
zone basedanonequation
the potential
transpiration
of the
Fatahi
et al.
(2010)
for root
water uptake
at
any
the
zone
the
tree
and a of
reduction
due
to on
suction;
any point
point
of
the root
rootfactor
zone based
based
on
the potential
potential transpiration
transpiration of
of the
the
tree
factor
to
)) is computed
where,
according
to
figure 66 of
al.
tree and
and aa((reduction
reduction
factor due
due
to suction;
suction;
where,
is (computed
to
of Feddeset
Feddeset
) according
( the
) (potential
) (figure
(2) al.
) transpiration
(1974),
(
)
is
factor
related
to
referred
(1974), ( ) is ((factor related
to
))
(( the
)) ((potential
)) (( )) transpiration referred
(2)
(2)
to
to model
model developed
developed by
by Nimah
Nimah and
and Hanks
Hanks (1973)
(1973) where;
where;
(( according
( (is
) computed
where, f()
according
to figure
6 of)) Feddeset
al. (1974),
to figure
6 of Feddeset
al.
)is computed
(3)
(( ( ))) )is
where,
computed
according
to
figure
66referred
of
al.
(3)
F(T
)
is
factor
related
to
the
potential
transpiration
toreferred
model
(1974),
is
factor
related
to
the
potential
transpiration
where,
is
computed
according
to
figure
of Feddeset
Feddeset
al.
(
)
(
)

P
( ) related
( ) ( to the potential transpiration
)where;
by
(1974),
(( by))Nimah
is
referred
(and
)Nimah
to model developed
and
Hanks
(1973)transpiration
(1974),
is factor
factor
related
to(1973)
the
potential
referred
developed
Hanks
where;
to
to model
model developed
developed by
by Nimah
Nimah and
and Hanks
Hanks (1973)
(1973) where;
where;
(
)
( )
((
))
(3)
(( ))
)
( ) ( )(
(3)
(3)
))

(( )) (( )) ((

Geotechnical Journal Vol. 6 No. 1 2014

80 Geotechnical Journal Vol. 6 No. 1 2014

G() is root density effect which
species
and soil
condition
k4
density
effect and
which
( )is root
(Fatahi
al. 2014),
species etand
soil condition and
(Fatahi et al. 2014),
( )

(
(

is
is
is
is

directly dependent on tree

an
experimental
coefficient
directly
dependent
on tree
an experimental coefficient

( ))
( ))

(4)

In equations 2,3, 4 and 5 k1 ,k2, k3 and k4 are experimental parameters

and
(x,y,z,t) is2,3,
the root
density
at time
in a givenpoint
can be
In equations
4 and
5 
are which
experimental
 ,
 
  and 
calculated
5) is the root density at time in a given
(
parametersusing
and equation
point which can be calculated using equation 5
(

( )

(( )|
)|

(( )|
)|

(( )|
)|

(( )|
)|

(5)

Figure 6.shows the initial root water uptake rate developed by Fatahi
(2007) using the values for experimental coefficients related to a
poplar tree.

Figure 6. Initial distribution of root water uptake of a poplar tree (Fatahi 2007)
2.3 Suction Development in Soil through Root Water Uptake
2.3 Suction Development in Soil through Root Water Uptake
Fatahiet al. (2009) developed an unsaturated flow equation using
Fatahiet al. (2009) developed an unsaturated flow equation using
Darcys
with root
root
Darcys Law
Law and
and Reynolds
Reynolds differential
differential transport
transport theorem
theorem with
water
uptake
as
a
sink
term
and
it
is
as
follows;
water uptake as a sink term and it is as follows;
(

(6)

where k is hydraulic conductivity is soil suction, z is vertical coordinate and t is time. S(x,y,z,t) is root water uptake. Equation 6 can
be applied
homogeneous
as well as
where
k isfor
hydraulic
conductivity
heterogeneous
is soil suction,media.
z is vertical
co-ordinate and t is time. S(x,y,z,t) is root water uptake. Equation
Equation 6 can be used to calculate the instantaneous moisture
6content
can be according
applied for to
homogeneous
as well
as heterogeneous
media.
the root water
uptake
and the subsequent
suction values can be computed by relating to the appropriate soil
Equation 6 can be used to calculate the instantaneous moisture
water characteristic curve which represents the suction variation
content according to the root water uptake and the subsequent
with respect to the moisture content variation in particular soil
suction values can be computed by relating to the appropriate soil
Two dimensional
analysis the
hassuction
been conducted
by
water
characteristicfinite
curve element
which represents
variation with
Indraratna
et moisture
al. (2008)content
for given
soil parameters
respect
to the
variation
in particularusing
soil governing
equations and the developed model verification was done for several
Two
dimensional
finite
analysis
beencomputed
conducted
by
case studies.
Figures
7,8,9element
show the
suctionhas
values
using
Indraratna
et al.
(2008)
for given
usingofgoverning
the ABAQUS
finite
element
modelsoil
for parameters
the parameters
Fredlund
and Huangs
analysis.model verification was done for several
equations
and(2001)
the developed
case studies. Figures 7,8,9 show the suction values computed using
The increasing of suction with time leads to the shear strength
the ABAQUS finite element model for the parameters of Fredlund
improvement of soil and at the same time root water uptake due to
and
Huangs (2001)
transpiration
leads toanalysis.
the fast dissipation of pore water.
The increasing of suction with time leads to the shear strength
improvement of soil and at the same time root water uptake due to
transpiration leads to the fast dissipation of pore water.

Above model captures the suction improvement of soil due to

Above modelof captures
suction
improvement
soil due to
transpiration
a tree, butthe
it does
not capture
the rootofreinforcement
transpiration
a tree,
but it does
not capture
theimprovement
root reinforcement
effect
which isofmore
significant
in shear
strength
of soil
effect which is more significant in shear strength improvement of
by vegetation.
soil by vegetation.
3 MECHANICAL STRENGTHENING WITH ROOTS.
3
MECHANICAL STRENGTHENING WITH ROOTS.
Tree roots are capable of binding the soil matrix together. Therefore,
it gives the fibrous effect to the soil to have good erosion control
Tree roots are capable of binding the soil matrix together. Therefore,
and to prevent shallow failures within1.5-2.0m (Bache et al. 1989).
it gives the fibrous effect to the soil to have good erosion control
At the same time large tree roots are capable of providing anchoring
and to prevent shallow failures within1.5-2.0m (Bache et al. 1989).
effect
thetime
soil.large
According
to are
thecapable
past studies,
amount
of root
At the to
same
tree roots
of providing
anchoring
materials present in the shear plane is the governing parameter of the
effect to the soil. According to the past studies, amount of root
reinforcement contribution to the soil. The amount of root material
materials present in the shear plane is the governing parameter of
available
in the soil contribution
is defined as Root
Area
Ratio
the reinforcement
to the
soil.
The(RAR).
amount of root
material available in the soil is defined as Root Area Ratio (RAR).
(

Effect
Effect of
of the
the tree
tree roots
roots on
on shear
shear strength
strength increment
increment of
of soil
soil has
has been
been
Effect
of the
roots onduring
shear past
strength
of soil
has
discussed
and
examined
few
decades
various
discussed
andtree
examined
during
past
few increment
decades by
by
various
been
discussed
and
examined
during
past
few
decades
by
various
researchers
in
different
ways
and
some
of
them
are
discussed
below
researchers in different ways and some of them are discussed below
researchers in different ways and some of them are discussed below
Waldron
value
Waldron (1977)
(1977) defined
defined the
the increment
increment of
of shear
shear strength
strength
value
and
10
of
root
and Figure
Figure
10 shows
shows
the
model
of flexible
flexible
elastic
root
extending
Waldron
(1977)
definedthe
themodel
increment
of shearelastic
strength
S extending
value and
vertically
thorough
horizontal
shear
zone.
vertically
thorough
horizontal
shear elastic
zone. root extending vertically
Figure
10 shows
theaamodel
of flexible
thorough a horizontal shear zone.

3rd Proff 18-02-2015

Geotechnical Journal Vol. 6 No. 1 2014

Figure 7.Variation of Matric suction after one month. (Indraratna et al. 2006)

Figure 8.Variation of Matric suction after six month. (Indraratna et al. 2006)

Figure 9.Variation of Matric suction after one year. (Indraratna et al. 2006)
.

Geotechnical Journal Vol. 6 No. 1 2014 81

Geotechnical Journal Vol. 6 No. 1 2014

Geotechnical
Journal
Vol.
No.
82
Geotechnical
Journal
No.
Geotechnical
Journal
Vol. 66Vol.
No.611 2014
20141 2014
Geotechnical Journal Vol. 6 No. 1 2014

N
N
N
N

(a)

(b)
Figure 10.(a)
Model
of
flexible elastic root extending vertically
(b)
(a)
(b) elastic
Figure
10.aModel
of flexible
extending
vertically
thorough
thorough
horizontal
shear
zone ofroot
thickness
Z (a)
Undisturbed
soil
(a)
(b)
Figure
10.
Model
of
flexible
elastic root
extending
vertically
Figure
10. shear
Model
of of
flexible
extending
Soil above
N displaced
by
x. elastic
a(b)
horizontal
zone
thickness
Z (a) root
Undisturbed
soilvertically
(b) Soil
Figure
10.
Model ofshear
flexible
elastic
root ZZextending
vertically
thorough
aa horizontal
zone
(a)
soil
thorough
horizontal
shear
zone of
of thickness
thickness
(a) Undisturbed
Undisturbed
soil
above
N displaced
by x.
thorough
a
horizontal
shear
zone
of
thickness
Z
(a)
Undisturbed
soil)
(b)
Soil
above
N
displaced
by
x.
According
to
Waldron
(1981),increment
of
the
shear
strength
(
(b) Soil above N displaced by x.
(b) Soil
above N added
displaced the
by x.coulomb equation since there is no
can
be directly
According
to Waldron
Waldronto
(1981),increment
of shear
the shear strength
According
to
(1981),increment
According
to Waldron
(1981),increment of
of the
the shear strength
strength (( ))
change
in
friction
angle.
(S
)becandirectly
betodirectly
added
to the
coulomb
equation
since
there (is no)
According
Waldron
(1981),increment
of
the
shear
strength
can
added
to
the
coulomb
equation
there
can be directly added to the coulomb equation since
since
there is
is no
(7)is no
change
in
friction
angle.
can
be
directly
added
to
the
coulomb
equation
since
there
change in friction angle.
change in friction angle.
(7)
(7)were
In Waldron (1977) model, failure modes of root system
(7)
defined as (a) stretching (b) slipping (c) breaking. Three different
Waldron
(1977)
model,
failure
modes
ofsystem
root
system
were
In
(1977)
model,
failure
modes
of root
were
defined
InWaldron
Waldron
(1977)
model,
failure
modes
root bond
system
were
values
which depend
on
the tensile
strength
ofofroots,
strength
In
Waldron
(1977)
model,
failure
modes
of
root
system
were
defined
as
(a)
stretching
(b)
slipping
(c)
breaking.
Three
different
as
(a)
stretching
(b)
slipping
(c)
breaking.
Three
different
S
values
defined as
(a) and
stretching
(b) slipping
(c) breaking.
Three
between
root
soil and
dimensional
parameters
of different
roots were
defined
as (a) depend
stretching
slipping
(c)
breaking.
Three
different
values
which
on
the
tensile
strength
of
roots,
bond
strength
which
the tensile
strength
of
roots,
bond
between
valuesdepend
which
depend
on(b)
the
tensilefailure
strength
of
roots,
bond
strength
introduced
for on
above
three
different
modes
ofstrength
roots.
values
which
depend
on
the
tensile
strength
of
roots,
bond
strength
between
root
and
soil
and
dimensional
parameters
of
were
root
and
soil
and
dimensional
parameters
of
roots
were
introduced
for
between root and soil and dimensional parameters of roots
roots
were
between
root
and
soil
anddifferent
dimensional
offorroots
were
introduced
for
above
three
failure
modes
of
Further,
Operstain
(2000)
obtained
andparameters
equation
apparent
above
three
different
failure
modes
of
roots.
introduced
for
above
three
different
failure
modes
ofroots.
roots.
introduced
for above
threeroots
different failure modes of roots.
cohesion induced
by tree
Further,
Operstain
(2000)
obtained
and
equation
for
apparent
Further, Operstain
Operstain (2000)
(2000) obtained
obtained and
and equation
equation for
for apparent
apparent
Further,
(8)
Further,
Operstain
(2000)
obtained
and) equation for apparent
cohesion
induced
by
tree
roots
(
cohesion
induced
by
tree
roots
cohesion induced by tree roots
cohesion induced by tree roots
(8)
)) roots, Tris the (8)
where C is apparent cohesion due((to tree
relative
(8)
root tensile strength contribution and( is )an experimental constant.
where
C
isis apparent
cohesion
to
roots, TTrisis the
relative
where
C of
apparent
cohesion due
due
to tree
tree
the
r
The
value
the
is influenced
by root
arearoots,
ratio and
the relative
tensile
where
C isstrength
apparent
cohesion and
due
to is
tree
roots,
Tris theconstant.
relative
root
tensile
contribution
an
experimental
root
tensile
strength
contribution
and
is
an
experimental
constant.
where
apparent
to tree
roots, Tris
theFrydman
relative
strengthCofisroots
and cohesion
diameter. due
Further
(Operstain
and
root
strength
and
is an
experimental
constant.
The
value
of
the
is influenced
by
area
ratio
tensile
The tensile
value
of
the contribution
influenced
bykroot
root
area
ratio and
and the
the
tensile
2000)
described
asis follows;
root
tensile
strength
contribution
and
is an
experimental
constant.
r
The
valueof
ofroots
the and
is influenced
by root
area
ratio and
the
tensile
strength
diameter.
Further
(Operstain
and
Frydman
strength
of
roots
and
diameter.
Further
(Operstain
and
Frydman
The
value
of
the
T
is
influenced
by
root
area
ratio
and
the
tensile
r as follows;
strength
of roots and
2000)
2000) described
described
as diameter.
follows; Further (Operstain and Frydman
strength
of roots and diameter.
2000) described
as follows;Further (Operstain and Frydman 2000)
described Tr as follows;
Stage 1
Stage 2
Stage
Stage 11
Stage 1

Stage
Stage 22
Stage 2

(9)

(9)
((
)is
((
)of
is the root area
where
(9) ratio,
) the ((tensile))strength
) root and
(9)failure
( is )the root
( cross
) section
(
) the unit area of soil in the
which
for
strength of
and
isis the
root
ratio,
where
is the
the tensile
tensile
of root
root
the
root area
area
where is
surface.
According
to strength
(Operstain
and and
Frydman
2000)
canratio,
be
is
the
tensile
strength
of
root
and
is
the
root
area
ratio,
where
which
is
the
root
cross
section
for
the
unit
area
of
soil
in
the
failure
which is the
rootfollowing
cross section
for the unit area of soil in the failure
computed
using
equation,
which
is
the
root
cross
section
for
the
unit
area
of
soil
in
the
failure
surface.
According
to
(Operstain
and
Frydman
2000)
can
surface. According to (Operstain and Frydman 2000)
can be
be
( (Operstain
) equation,and Frydman 2000)
surface.
According
to
can
be
computed
using
following
(
)
computed using following equation,
computed using following
equation,
)
(( A))and m1 are((experimental
)
where
coefficients and d(r, z, t) is the
(
)
(
)anddiameter
average
at point (r, z) at
time t.
where
Aroot
m1 are experimental
coefficients
and d(r, z, t) is the
where
A
and
m
are
experimental
coefficients
and
where Aroot
anddiameter
m11 are at
experimental
coefficients
and d(r,
d(r, z,z, t)t) isis the
the
average
point
(r,
z)
at
time
t.
where
A
and
m
are
experimental
coefficients
and
d(r,
z, t) isroot.
the
average
root
diameter
at
point
(r,
z)
at
time
t.
All
the
above
models
have
been
developed
for
a
vertical
1
average root diameter at point (r, z) at time t.
average root
at point
(r, z) at time
t.
Therefore
thediameter
practical
applications
are limited
to the more
All
the above
models have
been developed
fordue
a vertical
All the spatial
above distribution
models have
beensystems.
developed for a vertical root.
root.
complex
of root
Therefore
the
practical
applications
are
limited
due
to
All
the above
models have
been developed
for due
a vertical
root.
Therefore
the practical
applications
are limited
to the
the more
more
complex
distribution
of
systems.
Therefore
the practical
applications
are limited
to the based
more
(Docker
and
Hubble
2008)
developed
more due
empirical
complex spatial
spatial
distribution
ofroot
root
systems.
complex
distribution
of root
equationsspatial
for four
Australian
tree systems.
species considering progression
(Docker
and
Hubble
2008)
developed
more
empirical
based
(Docker
and
Hubble
more
(Docker
andIncrement
Hubble 2008)
2008)
developed
more
empirical
based
root
failure.
of
the developed
shear
strength
of empirical
soil
due tobased
root
(Docker
and
Hubble
2008)
developed
more
empirical
based
equations
for
four
Australian
tree
species
considering
progression
equations
four
tree
species
equations
for
four Australian
Australian
treefield
species
considering
progression
system
wasfor
determined
by in-situ
directconsidering
shear test ofprogression
Casuarina
equations
for
four Australian
tree
species
considering
progression
root
failure.
Increment
shear
strength
of
soil
due
to
rootfailure.
failure.
Increment
of
the
shear
strength
ofdue
soilto
due
to root
root
root
Increment
ofof
thethe
shear
strength
of
soil
rootAccacia
system
glauca,
Eucalyptus
amplifolia,
Eucalyptus
elata
and
root
failure.
Increment
offield
the direct
shear
strength
of Casuarina
soil
due
to
root
system
was
by
in-situ
field
direct
shear
test of
Casuarina
system
was determined
determined
by
in-situ
field
direct
shear
of
Casuarina
was
determined
by in-situ
test
of
glauca,
floribunda.
Empirical
equations
forshear
increase
in test
shear
strength
system was
determinedamplifolia,
by
in-situ field
direct
testand
offloribunda.
Casuarina
glauca,
Eucalyptus
Eucalyptus
elata
Accacia
Eucalyptus
amplifolia,
Eucalyptus
elata
and
Accacia
Eucalyptus
Eucalyptus
elata
andshear
Accacia
to root system
of each
tree species
against
root
area
Sglauca,
r(KPa) dueamplifolia,
glauca,
Eucalyptus
amplifolia,
Eucalyptus
elata
and Accacia
floribunda.
Empirical
equations
for
increase
in
shear
strength
floribunda.
Empirical
equations
for
increase
in
shear
strength
Empirical
equations
for
increase
in
shear
strength
S
(KPa)
due to
ratio (RAR) are as follows;
r
floribunda.
Empirical
equations
fortree
increase
shear root
strength
due
to
root
system
of
speciesinagainst
area
SSr(KPa)
dueof
toeach
roottree
system
of each
each
tree
against
root
r(KPa)
root
system
species
against
rootspecies
area ratio
(RAR)
arearea
as
due
to as
root
system of
root area
Sratio
ratio
(RAR)
are
follows;
.
Casuarina
glauca
: each tree species against
r(KPa)
(RAR)
are
as
follows;
follows;
ratio (RAR) are as follows;
..
Casuarina
::
Casuarina glauca
glauca
Eucalyptus
amplifolia
.
Casuarina
glauca
::
Eucalyptus
amplifolia
:
Eucalyptus elata
amplifolia
:
Eucalyptus
Eucalyptus
amplifolia
::
Eucalyptus
elata
::
Eucalyptus
elata
Accacia
floribunda
::
Eucalyptus
elata
Accacia
floribunda
:
Accacia
floribunda
In
the above
study it was ::observed that the amount of shear
Accacia
floribunda
increment computed by the (Waldron 1981) model was 50% more
In
the
above
study
itit observed
was
observed
the
amount
of
shear
above
study
it was
that thethat
amount
of
shear increment
Inthe
the
study
was Therefore
observed
that
the and
amount
of 2008)
shear
than
the above
experimental
values.
(Docker
Hubble
In
the
above
study
it
was
observed
that
the
amount
of
increment
computed
by
the
(Waldron
1981)
model
was
50%
more
computed
by
the
(Waldron
1981)
model
was
50%
more
than
the
increment computed
by three
the (Waldron
1981)
50% shear
more
introduced
a model with
stage failure
of model
roots aswas
in Figure
11.
increment
computed
by
the (Waldron
model
was
50% 2008)
more
than
values.
Therefore
(Docker
and
Hubble
experimental
values. Therefore
(Docker 1981)
and
Hubble
2008)
introduced
than the
the experimental
experimental
values.
Therefore
(Docker
and
Hubble
2008)
the with
experimental
values.
andas
2008)
introduced
aa model
with
three
stage
failure
roots
in
11.
athan
model
three
failure
of roots
as(Docker
inof
11.
introduced
modelstage
with
threeTherefore
stage
failure
ofFigure
roots
asHubble
in Figure
Figure
11.
introduced a model with three stage failure of roots as in Figure 11.
Stage 3
Stage
Stage 33
Stage 3

Figure 11.Three stages of failure of root reinforcement for three idealized and identical roots as estimated from the results of direct in-situ
Figure
11.Three
stages
of
three
and
shear
conducted
in of
this
investigation.
(modified afterfor
Docker
and
Hubble2001)
Figuretests
11.Three
stages
of failure
failure
of root
root reinforcement
reinforcement
for
three idealized
idealized
and identical
identical roots
roots as
as estimated
estimated from
from the
the results
results of
of direct
direct in-situ
in-situ
Figure 11.Three stages of failure of root reinforcement for three idealized and identical roots as estimated from the results of direct in-situ
shear
tests
conducted
in
this
investigation.
(modified
after
Docker
and
Hubble2001)
shear stages
tests
conducted
in this
investigation.
(modified
after
Docker
and Hubble2001)
stages
failures
are;
stage
1: Prior
to application
of an
Three
ofoffailures
stage
1: Prior
to(modified
application
of Docker
an
applied
displacement for each species. They show that the greatest shear
shear tests
conducted
inare;
this
investigation.
after
and Hubble2001)
applied
shear
force
the
roots
are
at
rest
across
the
potential
shear
resistance is provided by A. floribunda, followed by E. elata, E.
shear
force
the
roots
are
at
rest
across
the
potential
shear
plane.
In
Three
stages
of
failures
are;
stage
1:
Prior
to
application
of
an
displacement
for
each
species.
They
show
that
the
greatest
shear
Three
stages
of simplest
failures
are;
stagethey
1: extend
Prior
to
application
ofthis
an amplifolia,
displacement
for
each
species.
They
show
that
the is
greatest
plane.
In the
model
are assumed
to extend
amplifolia,
and
then
C.
glauca
although
there
very shear
little
the
simplest
model
they
are
assumed
to
perpendicular
to
then
C. glauca
although
there
is
very
discernible
Three
stages
of
failures
are; are
stage
1:
Prior
to the
application
of
an
displacement
for
each
species.
They show
that
thelittle
greatest
shear
applied
shear
force
the
roots
at
rest
across
potential
shear
resistance
isand
provided
by
A.
floribunda,
followed
by
E.
elata,
E.
applied
shear
force
the
roots
are
at
rest
across
the
potential
shear
resistance
is
provided
by
A.
floribunda,
followed
by
E.
elata,
E.
perpendicular
toAnthis
plane;shear
stageforce
2: An
applied
shear force
causes
discernible between
differencethese
between
these
latter
three species.
The likely
most
plane;
stage
2:
applied
causes
deflection
of
the
roots
difference
latter
three
species.
The
most
applied
shear
force
the
roots
are
at
rest
across
the
potential
shear
resistance
is
provided
by
A.
floribunda,
followed
by
E.
elata,
E.
plane.
In
the
simplest
model
they
are
assumed
to
extend
amplifolia,
and
then
C.
glauca
although
there
is
very
little
plane. In ofthethe simplest
model
they
are
assumed
to
extend
amplifolia,
and then
C.
glauca
although
there isrecorded
very little
deflection
roots
in
a
wide
shear
zone.
At
this
stage
the
likely
explanation
is
the
much
higher
tensile
strengths
for
in
a wide
zone.
At this
stagethey
the applied
majority
of roots
explanation
is thethen
much
tensile
strengths
recorded
A.
plane.
In shear
theto
model
assumed
to provide
extend
amplifolia,
and
C. higher
glauca
although
there
is very
little
perpendicular
this
plane;
stage
An
shear
force
causes
discernible
difference
between
these
latter
three
species.
Thefor
most
perpendicular
to simplest
this
plane;
stage 2:
2:through
An are
applied
shearforce
force
causes
discernible
difference
between
these
latter
three
species.
majority
ofthrough
roots
provide
resistance
a tensile
(T)
that
A.
floribunda
roots
in comparison
to the
other
three
species.The most
resistance
aroots
tensile
force
(T)
that
isapplied
mobilised
as
soilstage
pressure
floribunda
roots
in comparison
to
the
other
three
species.
perpendicular
to
this
plane;
stage
2:
An
shear
force
causes
discernible
difference
between
these
latter
three
species.
The
most
deflection
of
the
in
a
wide
shear
zone.
At
this
the
likely
explanation
is
the
much
higher
tensile
strengths
recorded
for
ofas the
roots
in (P)
a wide
shear
zone. At
likely explanation is the much higher tensile strengths recorded for
isdeflection
mobilised
soil
pressure
which
acts
against
thethis
root;stage
stagethe
3:
(P)
which acts
against
root;
stagethrough
3:
Sufficient
of
deflection
ofroots
the
rootsthe
inresistance
a wide
shear
zone.
Atdisplacement
this
stage
the
likely
explanation
tensile
strengths
recorded
for
majority
provide
aa tensile
force
(T)
that
A.
roots
comparison
to
other
three
species.
With
the
observation
ofmuch
abovehigher
test
results
(Docker
and Hubble
majority of
ofdisplacement
roots
provide
resistance
through
tensile
force
(T)
that
A. floribunda
floribunda
rootsisin
inthe
comparison
to the
the
other
three
species.
Sufficient
of
the
block
has
mobilised
the
full
tensile
With
thedescribed
observation
above test
results
(Docker
and
Hubbleto2009)
the
blockofhas
the
full
tensile
strength
(Tu)
a sufficient
majority
roots
resistance
through
a tensile
force
(T)
that
A. floribunda
rootstwo
inofcomparison
the other
threeaccording
species.
isis mobilised
asmobilised
soilprovide
pressure
(P)
acts
against
the
root;
stage
3:
2009)
types
of to
roots
systems
the
mobilised
pressure
(P) which
which
thein
stage
3: described
strength
(Tu) as
in soil
a sufficient
quantity
of acts
rootsagainst
to cause
a root;
reduction
in
With
observation
of
above
test
results
(Docker
and
Hubble
types
of roots
systems
according
to
the generated
shear
With the
the two
observation
of against
above
test
results
(Docker
and
Hubble
quantity
ofdisplacement
roots
to cause
a reduction
inhas
measured
resistance.
is
mobilised
as soil
pressure
(P)block
which
acts
againstshear
thetheroot;
3:
Sufficient
of
the
mobilised
fullstage
tensile
generated
shear
resistance
the
block
displacement.
Figure
12
Sufficient
displacement
measured
shear
resistance.of the block has mobilised the full tensile resistance
With
the
observation
of
above
test
results
(Docker
and
Hubble
2009)
described
two
types
of
roots
systems
according
to
the
against
the
block
displacement.
Figure
12
show
the
sketch
2009)
described
two
types
of
roots
systems
according
to
the
Sufficient
displacement
of
the
block
has
mobilised
the
full
tensile
strength
(Tu)
in
a
sufficient
quantity
of
roots
to
cause
a
reduction
in
the sketch of two types of root system and Figure 14 displays
strength (Tu) in a sufficient quantity of roots to cause a reduction in show
2009)
described
two
types
ofFigure
roots
systems
according
to the
generated
shear
resistance
against
the
block
displacement.
Figure
12
As
stated(Tu)
earlier
in-situ
shear
tests
have
beentoconducted
by
Docker
of
twofeatures
types
of
root
system
and
14
displays
main features
of
generated
shear
resistance
against
the
block
displacement.
Figure
12
strength
in
a
sufficient
quantity
of
roots
cause
a
reduction
in
measured
shear
resistance.
As
stated
earlier
in-situ
shear
tests
have
been
conducted
by
Docker
main
of
variation
of
the
two
types.
measured shear resistance.
generated
shear
resistance
against
the
block
displacement.
Figure
12
show
the
sketch
of
two
types
of
root
system
and
Figure
14
displays
&
Hubble
on
blocks
of
soil
containing
roots
of
the
four
riparian
tree
variation
of
the
two
types.
show
the
sketch
of
two
types
of
root
system
and
Figure
14
displays
measured
resistance.
&
Hubble shear
on blocks
of soil containing roots of the four riparian tree
show
the
sketch
of
two types
of two
root
system and Figure 14 displays
As
shear
tests
have
been
by
main
of
of
As stated
stated
earlier
in-situ
shear
tests
have
been conducted
conducted
by Docker
Docker
main features
features
of variation
variation
of the
the
two types.
types.
species
andearlier
Figurein-situ
13
the
plots
of
shear
versus
13 shows
shows
the
plots
ofaverage
average
shearstress
stress
versus
As
stated earlier
in-situ
shear
tests haveroots
beenof
by Docker
&
on
of
containing
the
tree
& Hubble
Hubble
on blocks
blocks
of soil
soil
containing
roots
ofconducted
the four
four riparian
riparian
tree main features of variation of the two types.
&
Hubble
onFigure
blocks13
ofshows
soil containing
roots
of theshear
four stress
riparian
tree
species
and
the
average
versus
species
and
Figure
13
shows
the plots
plots of
of
average
shear
stress
versus
species and Figure 13 shows the plots of average shear stress versus

Geotechnical Journal Vol. 6 No. 1 2014

rd
3Geotechnical
Proff 18-02-2015
Journal Vol. 6 No. 1 2014

Geotechnical Journal Vol. 6 No. 1 2014 83

Type 1;root systems that reached a maximum shear resistance

Type 1;root systems that reached a maximum shear resistance before
before the conclusion of testing and exhibited a definite decrease in
the
of testing
and aexhibited
definite
decrease
in
Typeconclusion
1;rootassystems
that reached
maximumashear
resistance
before
resistance
displacement
increased.
resistance
as displacement
the
conclusion
of testingincreased.
and exhibited a definite decrease in
resistance
displacement
Type
2;rootassystems
those
that
maximum resistance
resistance
Type 2;root
systems
those increased.
that either
either reached
reached aa maximum
or
recorded
no
reduction
from
that level,
level, or
or blocks
blocks where
where the
the shear
shear
or
recorded
no
reduction
from
that
Type 2;root systems those that either reached a maximum resistance
or recorded no reduction from that level, or blocks where the shear

Figure12. The distinct root morphology through an in-situ shear testblock

resistance
continued
to increase through
for the duration
the testtest.
Figure12.
The distinct
root morphology
an in-situofshear
block resistance continued to increase for the duration of the test.

Shear
Shear
Stress
Stress

Soil-Only peak
Soil-Only peak

Shear Displacement
Shear Displacement
Figure 13.Shear stress versus displacement plots for the four tree species and the soil-only tests (modified after Docker and Hubble 2008)
Figure 13.Shear stress versus displacement plots for the four tree species and the soil-only tests (modified after Docker and Hubble 2008)

Figure 14.Diagrammatic representation of two generally distinct root system behaviors at direct shearing. (modified after Docker and Hubble
2009).
Figure 14.Diagrammatic representation of two generally distinct root system behaviors at direct shearing. (modified after Docker and Hubble
2009).
According
to the above facts it is discernible that the spatial
parameters developed for mechanical strengthening for saturated
distribution to
of root
as afacts
major
despite all the
condition
not
realistic.
Therefore a strengthening
combined
effect
to are
be
According
the acts
above
it role,
is discernible
thatother
the factors
spatial
developed
for
mechanical
for saturated
condition
parametersare
developed
for strengthening
mechanical
for has
saturated
in reinforcing
typerole,
1 ordespite
type 2 all
trees
can be
taken
into are
account
and aparameters
should
be defined
accordingly.
distribution
of soil.
root There
acts asisa no
other factors
not
realistic.
Therefore
combined
effect
to be taken
into
account
major
thewhich
condition
not realistic.
Therefore
a has
combined
effect
has
to be
distinguished
theThere
environment
by 1the
The which
true tree
in
reinforcing in
is no type
and
defined
accordingly.
soil.
or species.
type 2 trees
canroot
be
takenparameters
into
and be
parameters
should of
be improvement
defined accordingly.
Figure
15 account
givesshould
a general
understanding
in subsoil
system depends
on the
environmental
described
section
distinguished
in the
environment
by thefactors
species.
The trueintree
root
structure
due
to
combined
effect.
Soil
element
A in figure15
is
Figure
15
gives
a
general
understanding
of
improvement
subsoil
1.2.
Figure 15 gives a general understanding of improvement in
in subsoil
dependson
onthe
theenvironmental
environmental
factors
described
in section
system depends
factors
described
in section
1.2.
directly
under
the
Railway
ballast
and
Soil
element
B
is
in
the
area
structure due
duetotocombined
combined
effect.
element
A in figure15
is
structure
effect.
SoilSoil
element
A in figure15
is directly
of
root zone.
General
unsaturated
soil Soil
mechanics
theories
valid
41.2. DISCUSSION
directly
under
the Railway
ballast
and
element
B isarea
inare
the
area
4 DISCUSSION
under
the
Railway
ballast
and
Soil
element
B
is
in
the
of
root
for
soil zone.
element A andunsaturated
therefore Vanapalliet
al. (1996) equation
can
of root
soil mechanics
4 stated
DISCUSSION
As
in the introduction most of the time sub soil structure
zone.
GeneralGeneral
unsaturated soil mechanics
theories theories
are validare
forvalid
soil
be
sused.
for
soil
element
A
and
therefore
Vanapalliet
al.
(1996)
equation
can
As
statedasin
the
most
time
sub the
soildepth
structure
remains
saturated state.
certain
and
element A and therefore Vanapalliet al. (1996) equation can be sused.
As stated
inpartially
the introduction
introduction
most Itof
ofisthe
the
time that
sub
soil
structure
be
sused.
remains
as
partially
saturated
state.
It
is
certain
that
the
depth
and
the
the
degree
of saturation
of the
vadose
zone increase
the soil
)
)
(
(
(
)
remains
as partially
saturated
state.
It is certain
that theindepth
and
(10)
degree
of which
saturation
of thewith
vadose
zonetrees.
increase
the soil
structure
interacts
mature
As increase
ainresult
of
that,
the
the degree
of saturation
of
the
vadose
zone
in structure
the soil
)
)
(
(
(
)
(10)
which
interacts
mature
trees.
As atrees.
resultAs
of that,
the of
parameters
structure
which with
interacts
with
mature
a result
that, the

Geotechnical Journal
Journal Vol.
Vol. 66 No.
No. 11 2014
2014
Geotechnical

84 Geotechnical Journal Vol. 6 No. 1 2014

In equation 10 ua anduw are pore air and water pressures; s, r, and
are saturated, residual and instant volumetric water contents and all
In equation
equation 10
10
and are
are pore
pore air
air and
and water
water pressures;
pressures; ss,, rr,, and
and
In
and
the
other symbols have
their usual
meaning.
are
are saturated,
saturated, residual
residual and
and instant
instant volumetric
volumetric water
water contents
contents and
and
the
symbols
have
their
meaning.
allvalue
the other
other
symbols10,
have
their usual
usual
meaning.
all
in equation
changes
with time
due to root water uptake
and
a
following
equation
can
be
stated
considering
Fatahi
(2007)s
value
in
equation
10,
changes
with
time
water
uptake
value in equation 10, changes with time due
due to
to root
root
water
uptake
equation
and Reynolds
differential
transport
theorem, Fatahi
and aa following
following
equation
can be
be stated
stated
considering
Fatahi (2007)s
(2007)s
and
equation
can
considering
equation
equation and
and Reynolds
Reynolds differential
differential transport
transport theorem,
theorem,
((

))

((

))

((

))

(11)
(11)

In equation
equation 11
11 S
S is
is the
the Root
Root water
water uptake
uptake sink
sink term
term and
and I
I is the
the
In
In
equationsource
11 S
is the
Root
water
uptake sink
term and
I isis the
irrigation
term.
is the
the
volumetric
moisture
content;
is
irrigation
source term.
is
volumetric
moisture
content;
is
irrigation
term. isisthe
volumetric
content;
k is
hydraulic source
conductivity;
matric
suction. moisture
At the
the same
same
time usual
usual
hydraulic
conductivity;
is matric
suction.
At
time
hydraulic
is matric
Atelement
the same
usual
soil water
waterconductivity;
characteristiccurve
curve
can be
besuction.
used for
for
Atime
soil
characteristic
can
used
element A
..
soil water characteristic curve can be used for element A .
Element B
B has
has great
great effect
effect from
from root
root suction
suction and
and root
root
Element
Element
B hassince
great it
rootTherefore
reinforcement
reinforcement
since
iteffect
is in
infrom
the root
root suction
zone of
ofand
tree.
Therefore
usual
reinforcement
is
the
root
zone
tree.
usual
unsaturated
soil
mechanics
theories
cannot usual
be directly
directly
used soil
for
since
it is in soil
the root
zone of theories
tree. Therefore
unsaturated
unsaturated
mechanics
cannot
be
used
for
element B.
B.
At the
thecannot
same time
time
stiffness
of the
the
root
zoneB.
results
the
mechanics
theories
be directly
used
for root
element
At the
element
At
same
stiffness
of
zone
results
the
increasing
of shear
shear of
strength
in zone
soil around
around
element
A. Therefore
Therefore
it
same
time stiffness
the root
results element
the increasing
of shearit
increasing
of
strength
in
soil
A.
can be
be defined
defined
asaround
shear
capacity
of the
the root
rootitarea
area
zone
which acts
acts
strength
in soil as
element
A. Therefore
canzone
be defined
as
can
aa shear
capacity
of
which
asshear
an external
external
stiffener
with
the
effect
from
rootacts
reinforcement
and
an
stiffener
effect
root
reinforcement
and
aas
capacity
of thewith
rootthe
area
zonefrom
which
as an external
suction. with the effect from root reinforcement and suction.
suction.
stiffener

Figure 15
15 -- Schematic
Schematic diagram
diagram of
of root
root water
water uptake
uptake process
process and
and root
root
Figure
reinforcement.
reinforcement.
Computingthethe
the
effects
from
root reinforcement
reinforcement
and
suction
Computing
effects
root
and
suction
Computing
effects
from from
root reinforcement
and suction
separately
separately
and superimposing
superimposing
them
may
not give
give
realistic
answer
separately
and
them
not
realistic
answer
and
superimposing
them may not
givemay
realistic
answer
since suction
since
suction
has
an
influence
on
root
reinforcement.
Suction
since
on root
reinforcement.
has
an suction
influencehas
on an
rootinfluence
reinforcement.
Suction
changes theSuction
bond
changes the
the bond
bond strength
strength between
between root
root and
and soil
soil as
as well
well it
it alters
alters the
the
changes
strength
between
root and soil
as well it alters
the material
properties
material properties
properties of
of root.
root. In
In this
this case
case the
the failure
failure of
of root
root system
system due
due
material
of root. In this case the failure of root system due to slipping,
to slipping,
slipping, stretching,
stretching, breaking
breaking or
or pulling out
out with
with soil
soil annulus
annulus are
are
to
stretching,
breaking
or pulling
outpulling
with
soil roots
annulus
are greatly
greatly affected
affected
by suction
suction
induced
by tree
tree
as well
well
as the
the
greatly
by
induced
by
roots
as
as
affected
suction
induced by tree roots as well as the adjacent soil
adjacentby
soil
suction.
adjacent
soil
suction.
suction.
CONCLUSIONS
55
CONCLUSIONS
5 CONCLUSIONS
Tree roots
roots can
can be
be used
used economically
economically in
in subgrade
subgrade improvement
improvement after
after
Tree
an effective
effective
analysis
ofeconomically
subsequent effects.
effects.
Past developed
developed
models
to
an
analysis
of
subsequent
Past
models
to
Tree
roots can
be used
in subgrade
improvement
after
compute
rootanalysis
based suction
suction
and the
theeffects.
root reinforcement
reinforcement
effect
have
compute
root
based
and
root
effect
have
an
effective
of subsequent
Past developed
models
been
contributed
considerably
for the
the root
sub reinforcement
grade improvement,
improvement,
been
contributed
considerably
for
sub
grade
to
compute
root based
suction and
effect
however
ascontributed
described considerably
in this
this paper
paperfor
comprehensive
model which
which
however
described
in
aa comprehensive
model
have
beenas
the sub grade improvement,
consist of
ofasmechanical
mechanical
strengthening
ascomprehensive
well as
as root
root based
based
suction
is
consist
strengthening
well
suction
however
described in
this paper aas
model
whichis
vital to
to have
have more
more realistic
realistic results.
results. Moreover
Moreover most
most of
of the
the trees
trees are
are
vital
consist of mechanical strengthening as well as root based suction is
not capable
capable of
of doing
doing the
the sub-soil improvement
improvement due to
to their shallow
shallow
not
vital
to have more
realisticsub-soil
results. Moreover mostdue
of the their
trees are not
root system
system and
and deciduous
deciduous behaviour.
behaviour. Therefore
Therefore it
it is
is really
really essential
essential
root
capable of doing the sub-soil improvement due to their shallow root
to select
select the
the most
most appropriate
appropriate plant
plant i.e.
i.e. good
good root
root system
system and
and
to
system
and deciduous
behaviour. Therefore
it is really essential
to
evergreen.
evergreen.
select
the most appropriate plant i.e. good root system and evergreen.

6 ACKNOWLEDGEMENTS
ACKNOWLEDGEMENTS
66 authors
ACKNOWLEDGEMENTS
The
acknowledge funding from the Australian Research
Council;
industry
partners GHD,
City
of Salisbury
and Transport
The
authors
acknowledge
funding
from
the Australian
Australian
Research
The authors acknowledge
funding
from
the
Research
Construction
Authority
and
a
number
of
past
PhD
students.
A
Council; industry
industry partners
partners GHD,
GHD, City
City of
of Salisbury
Salisbury and
and
Transport
Council;
Transport
substantial
component
of
the
technical
details
presented
in
this
paper
Construction
Authority
and
a
number
of
past
PhD
students.
A
Construction Authority and a number of past PhD students. A
can
be found
in earlierof
bydetails
the first
author in
andthis
hispaper
cosubstantial
component
ofpublications
the technical
technical
details
presented
in
this
paper
substantial
component
the
presented
workers
in the Gotechnique,
ICE. by
can be
be found
found
in earlier
earlier publications
publications
by the
the first
first author
author and
and his
his cococan
in
workers in
in the
the Gotechnique,
Gotechnique, ICE.
ICE.
workers
7 REFERENCES
REFERENCES
Bache,
D. H., et al. (1989). Vegetation In Civil Engineering
77
REFERENCES
Construction: Role And Uses. ICE Proceedings 86(1): 247-249.
Bache, D.
D. H.,
H., et
et al.
al. (1989).
(1989). "" Vegetation
Vegetation In Civil
Civil Engineering
Engineering
Bache,
Bhm,
W. (1980).Methods
Of Studying RootIn Systems.
Berlin:
Construction: Role
Role And
And Uses."
Uses." ICE
ICE Proceedings
Proceedings 86(1):
86(1): 247-249.
247-249.
Construction:
Springer (1979), Pp. 200, Dm 69, , Cambridge University Press.
Bhm, W.
W.D.(1980).Methods
(1980).Methods
Ofsoil
Studying
Root
Systems.
Berlin:
Bhm,
Studying
Systems.
Berlin:
Cameron,
A. (2000). TotalOf
suction Root
regimes
near trees
in a
Springer
(1979),
Pp.
200,
Dm
69,
,
Cambridge
University
Press.
Springer
(1979),
Pp.
200,
Dm
69,
,
Cambridge
University
Press.
semi-arid environment Unsaturated soils for Asia. Proceedings of
the Asian Conference on Unsaturated Soils, UNSAT-ASIA 2000,
Cameron, D.
D. A.
A. (2000).
(2000). "Total
"Total soil
soil suction
suction regimes
regimes near
near trees
trees in
in aa
Cameron,
Singapore:
739-744.
semi-arid environment
environment "" Unsaturated
Unsaturated soils
soils for
for Asia.
Asia. Proceedings
Proceedings of
of
semi-arid
the
Asian
Conference
on
Unsaturated
Soils,
UNSAT-ASIA
2000,
the
Asian
Conference
on
Unsaturated
Soils,
UNSAT-ASIA
2000,
Cameron, D. A. (2001).The extent of soil desiccation near trees in
Singapore:
739-744.
aSingapore:
semi-arid 739-744.
environment. Geotechnical & Geological Engineering
19(3-4): 357-370.
Cameron, D.
D. A.
A. (2001)."The
(2001)."The extent
extent of
of soil
soil desiccation
desiccation near
near trees
trees in
in aa
Cameron,
semi-aridM.environment."
environment."
Geotechnical
&re Geological
Geological
Engineering
semi-arid
Geotechnical
Dobson,
C. and A. J. Moffat
(1995).A&
evaluation ofEngineering
objections
19(3-4):
357-370.
19(3-4):
357-370.
to
tree planting
on containment landfills. Waste management and
research 13(6): 579-600.
Dobson, M.
M. C.
C. and
and A.
A. J.
J. Moffat
Moffat (1995)."A
(1995)."A re
re evaluation
evaluation of
of
Dobson,
objections
toand
tree
planting
on
containment
landfills."
Waste
objections
tree
planting
landfills."
Waste
Docker
B. Bto
Hubble
T. T.on
C. containment
(2001).Strength
And Stability
management
and
research
13(6):
579-600.To Slope Stability. 14th
management
research
13(6):
579-600.
Of
Casuarina and
Glauca
Roots
In Relation
Southeast Asian Geotechnical Conference.Lisse: 745-749.
Docker B.
B. B
B and
and Hubble
Hubble T.
T. T.
T. C.
C. (2001).Strength
(2001).Strength And
And Stability
Stability Of
Of
Docker
Casuarina
Glauca
Roots
In
Relation
To
Slope
Stability.
14th
Casuarina
Glauca
Roots
In
Relation
To
Slope
Stability.
14th
Docker B. B and Hubble T. T. C. (2008). Quantify RootSoutheast Asian
AsianOf
Geotechnical
Conference.Lisse:
745-749.
Southeast
Geotechnical
Conference.Lisse:
Reinforcement
River Bank
Soils By Four745-749.
Austrailian Tree
Species. Science Direct 100 (2008): 401-418.
Docker B.
B. B
B and
and Hubble
Hubble T.
T. T.
T. C.
C. (2008).
(2008). "Quantify
"Quantify RootRootDocker
Reinforcement
Of
River
Bank
Soils Modelling
By Four
Four Austrailian
Austrailian
Tree
Docker
B. B and Of
Hubble
T. T.Bank
C. (2009).
The Distribution
Reinforcement
River
Soils
By
Tree
Species."
Science
DirectStrength
100 (2008):
(2008):
401-418.
Species."
Science
Direct
100
401-418.
Of
Enhanced
Soil Shaer
Beneath
Riparian Trees Of SouthEastern Australia. Ecological Engineering 35 (2009): 921-924.
Docker B.
B. B
B and
and Hubble
Hubble T.
T. T.
T. C.
C. (2009).
(2009). "Modelling
"Modelling The
The
Docker
Distribution
Of Modelling
Enhanced Soil
Soil
Shaer Strength
Strength
Beneath
Riparian
Fatahi,
B. (2007).
Of Influence
Of MatricBeneath
Suction Induced
Distribution
Of
Enhanced
Shaer
Riparian
Trees
Of Vegetation
South-Eastern
Australia."Improvement,
Ecological Engineering
Engineering
35
By
Native
On Sub-Soil
University Of
Trees
Of
South-Eastern
Australia."
Ecological
35
(2009): 921-924.
921-924. Thesis.
(2009):
Wollongong.PhD
Fatahi,B.,Khabbaz,
B. (2007).
(2007). H.Modelling
Modelling
Of Influence
Influence
Of Matric
Matric Suction
Suction
Fatahi,
and Indraratna,
B. (2009). Parametric
studies
Fatahi,
B.
Of
Of
Induced
By Native
Native
Vegetation
On Sub-Soil
Sub-Soil
Improvement,
on
bioengineering
effects
of tree root-based
suction
on ground
Induced
By
Vegetation
On
Improvement,
University Of
Of
Wollongong.PhD
Thesis.
behaviour.
Ecological
engineering
35(10): 1415-1426.
University
Wollongong.PhD
Thesis.
Fatahi, B.,
B.,Khabbaz,
H. &Indraratna,
and Indraratna,
Indraratna,
B. (2009).
(2009).
"Parametric
Fatahi,
Khabbaz, H.
B. (2010).
Bioengineering
Fatahi,
B.,Khabbaz,
H.
and
B.
"Parametric
studies on
on
bioengineering
effects of
of
treewater
root-based
suction
on
ground
improvement
considering
root
uptakesuction
model,
studies
bioengineering
effects
tree
root-based
on
ground behaviour."
behaviour."
Ecological
engineering 35(10):
35(10): 1415-1426.
1415-1426.
Ecological
Engineering
36(2): 222-229
ground
Ecological
engineering
Fatahi,
Khabbaz, H.
& Fatahi, B.
(2012).Bioengineering
Mechanical
Fatahi, B.,
B., Khabbaz,
Khabbaz,
H.H.&Indraratna,
&Indraratna,
B.B.(2010).
(2010).
Bioengineering
Fatahi,
ground improvement
improvement
considering
root
water
uptake
model,
characteristics
of soft considering
clay treatedroot
withwater
fibre uptake
and cement,
ground
model,
Ecological Engineering
Engineering
36(2):
222-229
Geosynthetics
International,
19(2),
252-262.
Ecological
36(2):
222-229
Fatahi,
of
Fatahi, B.,
B., Khabbaz,
Khabbaz,H.H.
H.&Indraratna,
& Fatahi,
Fatahi, B.B.
B.(2014).
(2012).Modelling
Mechanical
Fatahi,
B.,
Khabbaz,
&
(2012).
Mechanical
unsaturated
ground
vegetation
characteristics
of behaviour
soft clay
clayinfluenced
treated by
with
fibre transpiration,
and cement,
cement,
characteristics
of
soft
treated
with
fibre
and
Geosynthetics International,
International,
19(2), 252-262.
252-262.
Geomechanics
and Geoengineering:
An International Journal, DOI:
Geosynthetics
19(2),
10.1080/17486025.2014.880520
Fatahi, B.,
B., Khabbaz,
Khabbaz, H.
H. &Indraratna,
&Indraratna, B.
B. (2014).
(2014). Modelling
Modelling of
of
Fatahi,
Feddes,
R.
A.,
Bresler,
E.
and
Neuman,
S.
P.
(1974).Field
test
of
a
unsaturated
ground
behaviour
influenced
by
vegetation
unsaturated ground behaviour influenced by vegetation
modified
numerical
model for water
uptake by root systems.
Water
transpiration,
Geomechanics
and Geoengineering:
Geoengineering:
An International
International
transpiration,
Geomechanics
and
An
Resources
Research
10(6): 1199-1206.
Journal, DOI:
DOI:
10.1080/17486025.2014.880520
Journal,
10.1080/17486025.2014.880520
Fredlund,
D.A.,
G. Bresler,
and Hung,
Q.Neuman,
(2001).Prediction
of volumetest
change
Feddes, R.
R.
A.,
Bresler,
E.V.
and
Neuman,
S. P.
P. (1974).Field
(1974).Field
test
of aa
Feddes,
E.
and
S.
of
in
an expansive
soilmodel
as a for
result
of uptake
vegetation
andsystems.
environmental
modified
numerical
model
for
water
uptake
by root
root
systems.
Water
modified
numerical
water
by
Water
changes.
Special
Publication 115: 24-43.
ResourcesGeotechnical
Research 10(6):
10(6):
1199-1206.
Resources
Research
1199-1206.

3rd Proff 18-02-2015

Fredlund, D. G.,Rahardjo, H. and Fredlund, M. D. (2012).
Unsaturated soil mechanics in engineering practice Hoboken, N.J,
Wiley-Blackwell.
Gardner, W. R. (1960). Dynamic Aspects Of Water Availability To
Plants. Soil Science 89(2): 63-73.
Gilmen, E. F. (1980) An Illustrated Guide to Pruning, Environmental
Horticulture Department, Ifas, University Of Florida.
Green, S. R. (1992). Radiation balance, transpiration and
photosynthesis of an isolated tree Agricultural and Forest
Meteorology, 64(1993): 201-221.
Hopkins, W. G. (1999). Introduction To Plant Physiology.
JhonWoley& Sons, Usa.
Indraratna, B.,Fatahi, B. and Khabbaz, H. (2006). Numerical analysis
of matric suction effects of tree roots. Proceedings of the Institution
of Civil Engineers: Geotechnical Engineering 159(2): 77-90.
Indraratna, B., Khabbaz, H. & Fatahi, B. (2008). Conceptual
Development and Numerical Modelling of Vegetation Induced
Suction and implications on Rail Track Stabilisation, Proceedings
of the 12th International Conference of International Association for
Computer Methods and Advances in Geomechanics (IACMAG),
Indian Institute of Technology, Bombay, Mumbai, India, 4335-4344.

Geotechnical Journal Vol. 6 No. 1 2014 85

Operstain, V. and Frydman, S. (2000). The Influence Of Vegetation
On Soil Strength. Ground Improvement 4: 81-89.
Potter, W. (2006). The Feasibility Of Improving Rail Infrastructure
By Using Native Vegetation On Clay Soils. University Of South
Australia. MSc. Thesis.
Radcliffe,D., Hayden, T., Watson, K., Crowly, P., and Phillips, R. E.
(1980) Simulation Of Soil Water Within The Root Zone Of A Corn
Crop. Agronomy Journal 72: 19-24
Sattelmacher, B.,Marschner, H. and Kuhne, R. (1990). Effects
Of The Temperature Of The Rooting Zone On The Growth And
Development Of Roots Of Potato (Solanum-Tuberosum). Annals Of
Botany 65(1): 27-36.
Teskey, R. O. and D. W. Sheriff (1995). Water use by Pinusradiata
trees in a plantation. Tree Physiology 16: 273-279.
Vanapalli, S. K., Fredlund, D. G., Pufahl, D. E., Clifton, A. W. (1996).
Model for the prediction of shear strength with respect to soil
suction. Canadian Geotechnical Journal 33(3): 379-392.
Waisel, Y.,Eshel, A. and Kafkafi, U. (2002). Plant roots: the hidden
half. New York, Marcel Dekker.

Kramer, P.J. (1995) Roots and Root systems, Chapter 5; 115-166

Waldron, L. J. (1977). The Shear Resistance Of Root-Permeated

Homogeneous And Stratified Soil1. Soil Sci. Soc. Am. J. 41(5): 843849.

Nimah, M. H. and Hanks, R. J. (1973).Model for estimating soil

water, plant and atmospheric inter relations.Soil Science Society of
America Journal.Vol 37(4) : 522-527

Waldren, L. J. and S. Dakessian (1981). Soil Reinforcement By

Roots: Calculation of Increased Soil Shear Resistance From Root
Properties. Soil Science 132: 427-435.

Nobel, P. S. (1991). Physiochemical And Environmental Plant

Physiology. Academic Press New York.

Weaver, J.E. (1958), Classification of Root Systems of Forbs of

Grassland and a Consideration of Their SignificanceEcology, Vol.
39(3) (July, 1958). Ecological Society of America.

86 Geotechnical Journal Vol. 6 No. 1 2014

Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx

An Analysis of Compacted Pavement Subgrade Behaviour Due to Climatic Effects

1
1
An Analysis of Compacted Pavement
Subgrade
Due to Climatic Effects
J. Kodikara
and T. IslamBehaviour

1
1Department of Civil Engineering,
Monash
J. Kodikara
and T.University,
Islam1 Victoria, Australia
1
E-mail:
Jayantha.Kodikara@monash.edu
Department of Civil Engineering, Monash University, Victoria, Australia
E-mail: Jayantha.Kodikara@monash.edu
ABSTRACT: The paper presents an analysis of pavement subgrade behaviour subjected to climatic effects. A newly developed
framework is used to aid the interpretations. The value of this framework is that it uses the compaction curves as the yield surface
ABSTRACT:
The to
paper
presents
analysis of pavement
subgradecan
behaviour
subjected
climatic
effects.
newly design
developed
framework
while
swelling due
wetting
andan
predominantly
elastic loading
take place
withintothe
surface.
TheAcurrent
approaches
is reviewed
used to aidwith
the interpretations.
The value
of this
framework
is that
it uses
thewetting
compaction
curves as the
surface while
swelling
due to
are
some interpretations
within
this
framework.
The
major
of pavements
andyield
equilibrium
moisture
content
wetting
and
predominantly
elastic
loading
can
take
place
within
the
surface.
The
current
design
approaches
are
reviewed
with
that develops due to suction equilibration are highlighted as primary changes due to climatic effects. The design needssome
to
interpretations within this framework. The major wetting of pavements and equilibrium moisture content that develops due to suction
take
into
account
these
effects
and
road
maintenance
needs
to
sustain
design
assumptions
such
as
maintaining
the
integrity
of
equilibration are highlighted as primary changes due to climatic effects. The design needs to take into account these effects and road
pavement
surface
maintenance
needsseal.
to sustain design assumptions such as maintaining the integrity of pavement surface seal.

1. INTRODUCTION
1. pavements
INTRODUCTION
Road
are valuable infrastructure in a nations asset base,
and
influence
the living
standards
of its people.
Large sums
are
Road pavements
are valuable
infrastructure
in a nations
asset base,
spent
construction
and maintenance
of roads.
Therefore,
and annually
influenceonthe
living standards
of its people.
Large
sums are
improvements
pavement
and design
bring
spent annuallyinonroad
construction
andanalysis
maintenance
of roads.can
Therefore,
significant
savingsinto road
communities.
road pavement
improvements
pavementUnfortunately,
analysis andthedesign
can bring
significant
savings
to communities.
Unfortunately,
the road
area
has received
less technical
advancement
than the traditional
pavement area
less be
technical
than the
geotechnical
area.has
Onereceived
reason may
that theadvancement
focus of pavement
traditional
geotechnical
One and
reason
mayplanning
be that issues,
the focus
engineers
is spread
across area.
transport
traffic
theof
pavement engineers
spread
transport andaspects
traffic of
planning
geometrical
design of isroads
andacross
the geotechnical
road
issues,For
the this
geometrical
of roadsofand
geotechnical
design.
reason, design
the behaviour
thetheroad
pavement aspects
may
of
road
design.
For
this
reason,
the
behaviour
of
the
road of
pavement
not have received due attention from a geotechnical point
view.
may not
have appears
receivedtodue
attention
a geotechnical
point
Another
reason
be that
roads from
are long
structures that
canof
view.over
Another
reason
appears
to be
that
roads are phase
long structures
spread
tens of
kilometres
during
one
construction
and the
that can
spread overis tens
of kilometres
duringtoone
construction
control
of construction
challenging
in comparison
a building
site
phasefoundations
and the control
of construction
is these
challenging
in comparison
where
are constructed.
Under
circumstances,
topto a building site where foundations are constructed. Under these
down empirical approaches are commonly employed for pavement
circumstances, top-down empirical approaches are commonly
design.
employed for pavement design.
The pavement structure may be considered as a number of geomaterial
The built
pavement
structure
may be
a number
layers
over the
soil subgrade.
Theconsidered
stiffnesses as
of layers
aboveof
layers
built over increased
the soil subgrade.
stiffnesses
thegeomaterial
subgrade are
progressively
so that theThe
strongest
layerof
so that the
is layers
at the above
top of the
the subgrade
pavement.are
Theprogressively
basic conceptincreased
of this bottom-up
strongest layer
is atthe
theload
top is
of spread
the pavement.
concept of
construction
is that
quicklyThe
on basic
the uppermost
this bottom-up
is only
that the
loadstress
is spread
quickly
on the
stronger
and stifferconstruction
layers so that
a small
is finally
received
stronger
stiffer layers
so that only
smallsimplest
stress is
at uppermost
the subgrade,
whichand
is normally
the weakest
layer.a The
finally received
at the involves
subgrade,thewhich
weakest
pavement
design approach
use ofisthenormally
subgrade the
California
layer. The simplest pavement design approach involves the use of
bearing ratio (CBR) of the pavement subgrade to assign a pavement
the subgrade California bearing ratio (CBR) of the pavement
layer thickness. Charts or formulae have been developed to estimate
subgrade to assign a pavement layer thickness. Charts or formulae
a good quality unbound pavement layer required on a subgrade with
have been developed to estimate a good quality unbound pavement
a given
value
technique,
it is also
layer CBR
required
on(Austroads,
a subgrade2010).
with aUsing
giventhis
CBR
value (Austroads,
possible
design
multiple
unbound
layers
with increasing
2010). toUsing
this
technique,
it is also
possible
to design quality
multiple
over
a givenlayers
subgrade.
unbound
with increasing quality over a given subgrade.
The alternative technique is so-called mechanistic modelling, where
The alternative
technique
is so-called
modelling,
where
fatigue
relationships
are commonly
usedmechanistic
to design pavement
layers.
fatigue
relationships
are
commonly
used
to
design
pavement
layers.
The common fatigue relationships follow the approach given in
The common
fatigue
relationships
followofthe
approach
given
Equation
1, where
the allowable
repetitions
traffic
loads N
arein
all
are
Equation
1,
where
the
allowable
repetitions
of
traffic
loads
N
all
given:
given:

a
N all

(1)

Theconstants
constantsa aand
andb bdepend
dependononthe
thematerial.
material.The
Thestrain
strain is isthethe
The
verticalstrain
strainexerted
exertedononthethesubgrade
subgradedue
duetotoa astandard
standardaxle
axleload,
load,
vertical
whichis isnormally
normallya single-axle
a single-axledual-tyre
dual-tyreload
loadconfiguration
configuration
with
which
with
anan
axleload
loadofof80kN.
80kN.This
Thismeans
meansthat
thatthetheactual
actualtraffic
trafficload
loadspectrum
spectrum
axle
comprisinga arange
rangeofofaxle
axleconfigurations
configurationswith
withvarying
varyingaxle
axleloads
loads
comprising
needs
converted
standard
axles,
commonly
referredtoto
needs
toto
bebe
converted
toto
standard
axles,
commonly
referred
asas
thethe
. The
design
traffic
will
allow
designlife
lifeofof
designtraffic
trafficN N
design
. dThe
design
traffic
will
allow
forfor
thethedesign
d
the pavement (say 40 years) and the possible growth in traffic
during this period. A suitable pavement layer system on a given

the pavement (say 40 years) and the possible growth in traffic during
this
period. A
pavement
on a given subgrade is
subgrade
is suitable
then chosen
such layer
that Nsystem
d = Nall. The determination of
then
chosen such
thatof
Nda=numerical
Nall. The determination
requires
the
the use
program suchofas a finite
element
requires
use
of a numerical
program
as a finite
element
model to
compute
model
to compute
the such
relevant
strain.
Commonly
these
layers
theincluding
relevant the
strain.
Commonly
these layers
subgrade
are assumed
to beincluding
elastic. the subgrade are
assumed to be elastic.
As the subgrade is the weakest link in a pavement system, the
Aspreparation
the subgrade
the weakest
linkvery
in aimportant.
pavement Normally
system, thethe
of is
subgrade
soil is
preparation
of
subgrade
soil
is
very
important.
Normally
theto
subgrade is either compacted or new compacted fill is placed
subgrade
is
either
compacted
or
new
compacted
fill
is
placed
raise the road surface. Therefore, the compaction specification to
and
raise
the road
Therefore,Currently,
the compaction
specification
process
havesurface.
major influences.
the actual
behaviourand
of a
process
havesubgrade
major influences.
Currently, thestate
actual
behaviour
of a
pavement
from the as-compacted
to during
pavement
pavement
fromclear.
the as-compacted
state
during because
pavementthe
operationsubgrade
is not very
This behaviour
is to
complex
operation
is not will
verynot
clear.
behaviour is state
complex
because
theits
clay subgrade
stayThis
in as-compacted
but will
change
clay
subgradeprimarily
will not stay
state butInwill
properties
due in
to as-compacted
climatic influences.
thechange
current
itspavement
propertiesdesign
primarily
duethese
to climatic
influences.
thehandled
currentin
guides,
behavioural
changesInare
pavement
design
guides,without
these behavioural
changes are handled
in an
an ad hoc
manner
detailed consideration.
For instance,
the without
as-compacted
of the pavement
subgrade
may be
adalthough
hoc manner
detailedstate
consideration.
For instance,
although
dry optimum,
soaked
may besubgrade
used in pavement
design
thewell
as-compacted
state
of theCBR
pavement
may be well
dryto
allow for
any future
major
of the subgrade.
optimum,
soaked
CBR may
be wetting
used in pavement
design to This
allowpaper
for
presents
analysis
of pavement
subgrade behaviour
to climatic
any
future an
major
wetting
of the subgrade.
This paperdue
presents
an
effects ofand
traffic loading.
MPK framework
newly
developed
analysis
pavement
subgradeThe
behaviour
due to climatic
effects
and
by the
authorsThe
is used
aid the interpretation.
traffic
loading.
MPKtoframework
newly developed by the authors
is used to aid the interpretation.
INFLUENCE
OF COMPACTION
STATE
2. 2.
INFLUENCE
OF COMPACTION
STATE
2.1 MPK Framework
2.1 MPK Framework
Kodikara (2012) presented a novel framework (dubbed the MPK
Kodikara (2012) presented a framework)
novel framework
(dubbed
MPK
(Monash-Peradeniya-Kodikara)
to explain
and the
predict
framework)
to explain
and predict
the(Monash-Peradeniya-Kodikara)
behaviour of compacted soils. The
novel aspect
of this framework
soils. The
novel aspect
of this
is the
that behaviour
it extends of
the compacted
traditional concept
of Proctors
compaction
framework
is
that
it
extends
the
traditional
concept
of
Proctors
curve to represent the soil volumetric behaviour. In addition, it
compaction curve to represent the soil volumetric behaviour. In
utilises moisture content (given as moisture ratio ew =wGs, where
addition, it utilises moisture content (given as moisture ratio e
w is the gravimetric moisture content and Gs is the specific gravity w
=wGs, where w is the gravimetric moisture content and Gs is the
of soil) to replace the traditionally used matric suction, which is
specific gravity of soil) to replace the traditionally used matric
normally
parameter
to quantify,
especially
in theto
field.
The
suction,a difficult
which is
normally
a difficult
parameter
quantify,
framework
has
been
used
to
explain
most
characteristic
behaviours
especially in the field. The framework has been used to explain
ofmost
compacted
clays, including
volumetric
deformationclays,
under loading
characteristic
behaviours
of compacted
including
and
unloading,deformation
swelling behaviour
wetting,
collapse behaviour
volumetric
under during
loading
and unloading,
swelling
during
wetting,
swelling
pressure
development
during
constrained
behaviour
during
wetting,
collapse
behaviour
during
wetting,
swelling
andpressure
crackingdevelopment
during constrained
swelling
duringdrying.
constrained swelling and
cracking
constrained
drying.
The
typicalduring
compaction
curve is
approximately an inverted parabolic
curve with the maximum dry density (s) corresponding to optimum
The typical compaction curve is approximately an inverted parabolic
moisture content. However, this curve depends on the applied energy
curve
withapplied
the maximum
density
( d ) corresponding
level
(or the
stress) and dry
a family
of compaction
curves can be to
optimum
moisture
content.
However,
this
curve
depends
on the
obtained corresponding to different energy levels or applied stresses.
applied
energysuch
level
(or the
applied stress)
family of
Figure
1a shows
a family
of compaction
curves.and
Theamaximum
compaction
curvesa can
beoptimums,
obtained referred
corresponding
to different
density
points define
line of
to as LOO,
which
energy falls
levelsator
1a shows
such soils.
a family
typically
80applied
to 90% stresses.
degree ofFigure
saturation
for clayey
Theof
compaction curves. The maximum density points define a line of
optimums, referred to as LOO, which typically falls at 80 to 90%

3rd Proff 18-02-2015

Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.??

?? 2015 ISSN
xxxx-xxxx
Geotechnical
Journal
Vol. 6
Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx
ew
degree of saturation can be defined as e .It can be considered that

degree
of saturation
for clayey
The will
degree
of air
saturation
can be
soil
compacted
to the dry
side of soils.
optimum
have
free to move,
degree of maintaining
saturation
formostly
clayeyatmospheric
soils. The degree
of saturation
canthe
be
ew
therefore
air pressure,
whereas
). It can be considered that soil compacted to the dry
defined ase
soil
compacted
of optimum
willcompacted
have air trapped.
In
canwet
be side
considered
that soil
to the dry
defined
as w e).toItthe
this
case,
airepressure
develop
during
fast loading.
shown in
side
of optimum
willcan
have
air free
to move,
thereforeAsmaintaining
Figure
the MPK
extends
the
of maintaining
compaction
side
of 1b,
optimum
will framework
have
air free
to move,
therefore
mostly
atmospheric
air
pressure,
whereas
thefamily
soil compacted
to the
curves
in
the
3-D space
using
axisthefor
applied
mostly
atmospheric
airwill
pressure,
whereas
compacted
to net
the
wet side
of optimum
haveanother
air
trapped.
Insoil
this
case,orairyield
pressure
wet
of
optimum
will
air trapped.
Inin
this
case,
air the
pressure
stress
(p)
giving
a compaction
surface.
net
stress
is defined
asMPK
the
canside
develop
during
fast have
loading.
As The
shown
Figure
1b,
can
develop
As
in Figure
1b,ratio
framework
extends
theloading.
of shown
compaction
curves
inthetheMPK
applied
stressduring
minusfast
the
airfamily
pressure.
Here, the
moisture
ew3-D
is
space
using
axisfamily
for
applied
or yield
stress in
(p)analysis,
giving
framework
extends
the
of compaction
curves
the
3-Da
used
in place
ofanother
gravimetric
moisture
content.
Fornet
theoretical
compaction
surface.
Theforthis
netapplied
stress
isindefined
the applied
stress
space
using another
axis
or
yield
stress
(p)
giving
it
is
convenient
to consider
surface
e-ew net
- passpace,
as shown
ina
is
used
in
place
of
minus
the
air
pressure.
Here,
the
moisture
ratio
e
compaction
surface.
The
net
stress
is
defined
as
the
applied
stress
w
Figure 1c, which shows the surface depicting the loosest states of soil
gravimetric
moisture
For theoretical
analysis,
it ofis
is used
in place
minus
the airreferred
pressure.
the moisture
ratio
ew Boundary
with
stress,
to Here,
ascontent.
Loading
Wetting
State
Surface
gravimetric
moisture
For
analysis,
it is
as shown
in
convenient to
consider content.
this surface
in etheoretical
ew p space,
(LWSBS).
space,
as
shown
in
convenient
to
consider
this
surface
in
e

p
Figure 1c, which shows the surface depicting
the loosest states of
w
soil with
stress, shows
referred
as Loading
Wetting
State Boundary
Figure
1c, which
thetosurface
depicting
the loosest
states of
Surface
soil
with (LWSBS).
stress, referred to as Loading Wetting State Boundary
Surface (LWSBS).

Figure 1a. Family of compaction curves and the line of optimums

(Kodikara, 2012)
Figure 1a. Family of compaction curves and the line of optimums
(Kodikara, 2012)

Figure 1b. Compaction surface in 3-dimensional space (Kodikara,

2012)
Figure 1b. Compaction surface in 3-dimensional space (Kodikara,
2012)

No. 1 2014 87

Figure 1c. Compaction surface given as Loading Wetting State

Boundary
(LWSBS)
(Kodikara,
Figure
1c.Surface
Compaction
surface
given 2012)
as Loading Wetting State
Boundary Surface (LWSBS) (Kodikara, 2012)
2.2Likely
Likely
Behaviour
when
Compacted
or Side
Dry of
Side
2.2
Behaviour
when
Compacted
Wet Wet
or Dry
of
Optimum
2.2 Optimum
Likely Behaviour when Compacted Wet or Dry Side of
Optimum
2.2.1
Collapsepotential
potential
2.2.1 Collapse
2.2.1
Collapse
potential
Collapse in unsaturated compacted soils is referred to as substantial
Collapse
in unsaturated
compacted
is referred
to as substantial
compression
that can happen
when soils
the soil
is wetted.
it is
Collapse
in unsaturated
compacted
soils
is referred
to Usually,
as substantial
compression
that
can happen
when thesoil
soilwill
is wetted.
Usually,
it is
expected
that
unsaturated
compacted
swell
when
wetted.
compression that can happen when the soil is wetted. Usually, it is
Experimental
evidence
indicates
that collapse
follow
expected
that
compacted
soil
swell
when
wetted.
expected
that unsaturated
unsaturated
compacted
soil will
will can
swell
whenswelling
wetted.
under certain evidence
stress andindicates
initial density
(or void
conditions.
Experimental
that
can
follow
swelling
Experimental
evidence
indicates
that collapse
collapse
canratio)
follow
swelling
Collapse
is
considered
to
be
mostly
responsible
for
the
bump
that
under
certain
stress
and
initial
density
(or
void
ratio)
conditions.
under certain stress and initial density (or void ratio) conditions.
develops atconsidered
overpass approaches
inresponsible
roads, when
the bump
overpass
Collapse
for
the
that
Collapse is
is considered to
to be
be mostly
mostly responsible
for the
bump that
structure at
remains
unsettled
while in
the roads,
approach
mainly
develops
overpass
approaches
whensection
the overpass
comprising
compacted
fill settles
withthe
time.approach
This behaviour
been
structure
remains
unsettled
while
sectionhasmainly
explained within the MPK framework and collapse is likely to occur
comprising compacted fill settles with time. This behaviour has been
when soils are compacted dry of LOO only. The reason for this is
and collapse is likely to occur
explained
within
framework
the collapse
of the
the MPK
cloddy
structure that
occurs when soil is
soils are
are
compacted
dryofofLOO
LOO 2only.
The
reason
is
when
soils
dry
The
reason
for for
thisthis
is the
compacted
drycompacted
side of optimum.
Figureonly.
shows
the loading/wetting
the
collapse
of
the
cloddy
structure
that
occurs
when
soil
collapse
of
the
cloddy
structure
that
occurs
when
soil
is
compacted
behaviour of compacted kaolin depicted in this framework. Theis
compacted
dry
side
ofFigure
optimum.
Figure
2 loading/wetting
shows the
loading/wetting
dry
side ofwas
optimum.
2through
shows
the
behaviour
LWSBS
developed
constant
moisture
content
behaviour
of
compacted
kaolin
depicted
in
this
framework.
The
of
compacted ofkaolin
depicted
in this
framework.
LWSBS
was
compression
loosely
compacted
soils
(at 7 kPa,The
nominal
stress)
LWSBS
was
developed
through
constant
content
atmoisture
various of
moisture
statically in
a compaction
(K0 conditions)
developed
through
constantmould
moisture
content
compression
loosely
compression
of loosely compacted
soils
7 kPa,
nominal
ratios. Thesoils
independent
shownstress)
with (at
path
ABCD
for astress)
soil
compacted
(at 7 kPa,test
nominal
statically
in aiscompaction
statically in a compaction mould (K0 conditions) at various moisture
(path
A
to
B)
and
then
wetted
to
compacted
at 8.96% ( eat

0
.
237
)
mould
(K0 conditions)
various
moisture
ratios.
The
independent
ratios. The independentw test shown with path ABCD is for a soil
test
shown
with
path
ABCD
is
for
a
soil
compacted
at
8.96%
to 2000
kPato
27.35% moisture
content
( e0w.237
0.)725
) and
(path
A then
to B)loaded
and then
wetted
compacted
at 8.96%
( ew
(ew = 0.237) (path A to
B) and then wetted to 27.35% moisture
stress. It can be seen that when the soil is wetted at 50 kPa stress, it
and then
loadedIttocan
2000
kPa
27.35% (e
moisture
content
( ewloaded
0.725
content
= 0.725)
and then
to )2000
kPa stress.
be seen
undergoesw collapse (compression)
along the 50 kPa stress contour to
that
when
the
soil
is
wetted
at
50
kPa
stress,
it
undergoes
collapse
stress.
It
can
be
seen
that
when
the
soil
is
wetted
at
50
kPa
stress,
point C. When the soil is loaded further it follows the LWSBSit
(compression)
along
themoisture
50 kPa content.
stress
contour
point
C. contour
When
the
undergoes
collapse
(compression)
along
the
50 to
kPa
stress
to
corresponding
to new
While
this
example
shows
soil
is C.
loaded
further
it follows
the further
LWSBS
to new
point
When
the the
soil
is loaded
itcorresponding
follows
the LWSBS
LWSBS
can
predict
likely
pathway
during
loading/wetting,
in
corresponding
to new
moisture
content.shows
While
this stress
example
shows
moisture
content.
While
this
example
LWSBS
can
practical compaction,
soil
is compacted
to
a higher
andpredict
then
LWSBS
predict
the likely
pathway
during
loading/wetting,
the
likelycan
pathway
loading/wetting,
in apractical
compaction,
unloaded
(such
as during
when
rollers
depart after
number
of passes).in
practical
compaction,
soil is
compacted
to aunloaded
higher
stress
and
then
However,
the compacted
soil
canand
be then
further
be loaded
soil
is compacted
to a higher
stress
(suchto
as some
when
unloaded
(such
aswhich
depart
after
a number
of In
passes).
operational
stress,
isrollers
normally
lessHowever,
than 200
kPacompacted
or so.
road
rollers
depart
after
awhen
number
of passes).
the
soil
However,
theonly
soilvertical
can
be stress
further
bewhich
loaded
to under
some
pavements,
a nominal
would
prevail
can
be further
becompacted
loaded
to some
operational
stress,
is normally
operational
stress,
which
is
normally
lesstraffic
thanonly
200
orisso.
In road
operational
conditions,
except
when
loading
imposed
less
than 200
kPa or
so. In
road pavements,
akPa
nominal
vertical
pavements,
only
a
nominal
vertical
stress
would
prevail
under
leading
mainly
to
the
momentary
rise
of
vertical
stress.
This
stress would prevail under operational conditions, except when traffic
operational
conditions,
except when traffic loading is imposed
situation is discussed
later.
loading is imposed leading mainly to the momentary rise of vertical
leading mainly to the momentary rise of vertical stress. This
stress. This situation is discussed later.
situation is discussed later.

Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx

88 Geotechnical Journal Vol. 6 No. 1 2014

Figure 2. LWSBS for kaolin and loading/wetting pathway of an

independent soil specimen.

Figure 3. Variation in swelling percentage under nominal vertical

stress for first major wetting (based on Holtz and Gibbs, 1956)

2.2.2 Other
Other behaviours
behaviours

have
As abeen
general
comment,
soils that are
compacted significantly
found
during subsequent
investigations
to have lost
dry of optimum
appear todecrease
compactin(or
density)
compaction
and exhibited
dryincrease
density in
after
some
with
significant
wet/dry
cycles
and
soils
that
are
compacted
time. When such fills have experienced some form of failure,
wet side cases
of optimum
tend have
to expand
or
numerous
of litigation
ensued (decompaction
and the evidence
decrease
in
density)
during
wet/dry
cycles
(Kodikara
et
al.,
of decompaction has been a point of dispute (Burman et al.,
2014). Decompaction has been a problem in compacted fills,
2008).
where fills compacted to agreed specifications of dry density
have been found during subsequent investigations to have
3. MOISTURE
VARIATION
DURING
SERVICE
lost compaction
and exhibited
decrease
in dry density after
some time. When such fills have experienced some form of
Once a failure,
compacted
fill or cases
a pavement
subgrade
compacted,
its
numerous
of litigation
haveis ensued
and the
moistureevidence
conditionofcan
change with time.
As shown
schematically
decompaction
has been
a point
of dispute
in Figure(Burman
4, it is possible
to consider an equilibrium moisture content
et al., 2008).
that soil will eventually achieve in thermodynamic equilibrium with
the surrounding layers, water table (if present at the vicinity) and
atmospheric
interaction.
More precisely,
it will
be the soil suction
3.
MOISTURE
VARIATION
DURING
SERVICE
that will reach equilibrium, since it is the thermodynamic potential
Oncecontrols
a compacted
fill moisture
or a pavement
is compacted,
its
that
the soil
energysubgrade
state. Past
research has
moisture condition
can change
time. As shown
developed
links between
climatewith
(characterised
by the schematically
Thornthwaite
in Figure 4, it is possible to consider an equilibrium moisture
Moisture Index or TMI) and equilibrium suction (commonly given in
content that soil will eventually achieve in thermodynamic
pF or log of water head given in cm (Russam and Coleman, 1961).
equilibrium with the surrounding layers, water table (if present at
Then, of course, it is possible to link the equilibrium suction to an
the vicinity) and atmospheric interaction. More precisely, it will be
equilibrium
moisture
moisture
ratio) referring
the soil suction
thatcontent
will (or
reach
equilibrium,
since itto isa soil
the
water
characteristic
curve
for
stabilised
soil.
This
approach
is relevant
thermodynamic potential that controls the soil moisture energy
state.
when
the water
is well links
belowbetween
the ground
surface,
otherwise by
its
Past research
hastable
developed
climate
(characterised
influence
may
dominate.
the Thornthwaite Moisture Index or TMI) and equilibrium suction
(commonly
in 4,
pFiforthelog
of iswater
head given
in cm (Russam
As
shown ingiven
Figure
soil
compacted
at a higher
moisture
and
Coleman,
1961).
Then,
of
course,
it
is
possible
link the
content than the equilibrium moisture content, the soil to
is likely
to
equilibrium
suction
to
an
equilibrium
moisture
content
(or
moisture
undergo drying with intermediate variations in water content (due to
ratio) referring to a soil water characteristic curve for stabilised soil.
climatic events) and then reach the equilibrium state. Over the long
This approach is relevant when the water table is well below the
term, moisture content may fluctuate around the equilibrium during
ground surface, otherwise its influence may dominate.
seasonal variations if they are significant. On the other hand, if the
soil
is compacted
at 4,
a lower
moisture
content than
the equilibrium
As shown
in Figure
if the soil
is compacted
at a higher
moisture
moisture
content,
the
soil
is
likely
to
wet
up
to
equilibrium
content than the equilibrium moisture content, the soil is state
likelyand
to
undergo
cyclic
behaviour
during
seasonal
events.
Exceptions
such
undergo drying with intermediate variations in water content to
(due
to
progressive
wetting
drying
be events such
flooding
that
climatic events)
andor
then
reachwould
the equilibrium
state.asOver
the long
can
the soil
wet almost
to saturation,
of the during
initial
term,make
moisture
content
may fluctuate
aroundregardless
the equilibrium
compaction
moistureif state.
seasonal variations
they are significant. On the other hand, if the
soil is compacted at a lower moisture content than the equilibrium
moisture content, the soil is likely to wet up to equilibrium state and
undergo cyclic behaviour during seasonal events. Exceptions to such
progressive wetting or drying would be events such as flooding that
can make the soil wet almost to saturation, regardless of the initial
compaction moisture state.

It is well established that the behaviour of compacted clays can

differ when compacted wet or dry side of the LOO. Some of these
characteristics can be summarised as follows:
Soil stiffness and shear and tensile strength tend to be higher
Soil stiffness and shear and tensile strength tend to be higher
on the dry side of optimum, but many soils tend to peak at
on the dry side of optimum, but many soils tend to peak at
some moisture content drier than optimum, and then even
some moisture content drier than optimum, and then even show
show a reduction in stiffness or strength with decrease of
a moisture
reductioncontent.
in stiffness or strength with decrease of moisture
content.
Soil saturated permeability tends to be orders of magnitude
higher
dry of
optimum and
normally
achieve a
Soil
saturated
permeability
tendswould
to be orders
of magnitude
minimum
percentage
points
wet ofachieve
optimum.
On the
higher
dry offew
optimum
and would
normally
a minimum
basis
of a range
of test
results,
Kodikara
few
percentage
points
wet
of optimum.
Onand
the Rahman
basis of a(2002)
range
that Kodikara
moisture and
content
that (2002)
will give
minimum
ofindicated
test results,
Rahman
indicated
that
hydraulic
conductivity
is about
0.151hydraulic
x PI wetter
than the
moisture
content
that will give
minimum
conductivity
where
PI is the
is why
isoptimum,
about 0.151
x PI wetter
thansoil
theplasticity
optimum,index.
where This
PI is the
soil
landfill
clay
liners
are
normally
compacted
wet
of
optimum,
plasticity index. This is why landfill clay liners are normally
targeting the lowest permeability achievable for a certain
compacted wet of optimum, targeting the lowest permeability
compaction energy.
achievable for a certain compaction energy.
However, when the soil is compacted wet of optimum, the
likely shrinkage
during
initial drying
to bethehigher,
However,
when the soil
is compacted
wet oftends
optimum,
likely
which can
increase
potential
and can
the
shrinkage
during
initial the
drying
tends toforbe cracking
higher, which
severitythe
of potential
cracking for
during
desiccation.
increase
cracking
and the severity of cracking
during
Soil swelling
percentage (or strain given with respect to a
desiccation.
soil volume) tends to be higher on the dry side of optimum
Soil
swelling
(or strain
given variation
with respect
to a soil
than
on the percentage
wet side. Figure
3 shows
of swelling
volume)
tends
to
be
higher
on
the
dry
side
of
optimum
than on
percentage (with respect to original volume) at nominal
the
wet when
side. Figure
3 shows
variation
swelling
percentage
stress
subjected
to initial
majorofwetting.
For
this soil,
(with
respect that
to original
volume)
at nominal
stress
it is evident
percentage
of swelling
does not
varywhen
very
subjected
initial
wetting. For
thisratio
soil, for
it is Sevident
that
much at tothe
samemajor
as-compacted
void
r less than
percentage
of swelling
notposition
vary very
at the same
0.62, meaning
that the does
swelled
on much
the saturation
line
as-compacted
ratio
Sr less than 0.62,
meaning
the
(void ratio) void
would
be for
approximately
the same,
as that
shown
with the
arrow on
1. the
However,
as the
initial
swelled
position
saturation
linesoil
(void
ratio)compaction
would be
state approaches
close to
LOO
thethe
swelled
(and
approximately
the same,
as the
shown
with
arrowposition
1. However,
droptoalong
the
of compaction
swelling given
efinal) willclose
asthe
theamount
soil initial
stateby
approaches
the LOO
saturation
line (as shown
arrowsof2,swelling
3, 4, 5 and
6).by
Arrow
the
swelled position
(and thebyamount
given
efinal)
7 shows
the swelling
that can
when by
starting
from
will
drop along
the saturation
lineoccur
(as shown
arrows
2, 3,a
density
(lower
but significantly
of LOO.
4,high
5 and
6). Arrow
7 void
showsratio)
the swelling
that candry
occur
when
However,
be noted
that void
this amount
swelling is
starting
fromit ashould
high density
(lower
ratio) butofsignificantly
foroffirst
swelling
andit should
if significant
takes
dry
LOO.
However,
be notedwetting/drying
that this amount
of
to
place,
this
final
position
is
considered
to
move
up
or
e
final
swelling is for first swelling and if significant wetting/drying
increase.
takes place, this final position is considered to move up or e
to increase.

final

As a general comment, soils that are compacted significantly

dry of optimum appear to compact (or increase in density)
with significant wet/dry cycles and soils that are compacted
wet side of optimum tend to expand (decompaction or
decrease in density) during wet/dry cycles (Kodikara et al.,
2014). Decompaction has been a problem in compacted fills,
where fills compacted to agreed specifications of dry density

3rd Proff 18-02-2015

wi

Starting fromwai wi
Starting
from
from
a wa
wetStarting
condition
i
wetwet
condition
condition
Starting from a
Moisture
wet condition

Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx
SriSri
Lankan
Lankan
Geotechnical
Geotechnical
Society
Society
International
International
Conference
Conference
SLGS
SLGS
Vol.
Vol.
?? ??
No.??
No.??
????
2015
2015
ISSN
ISSN
xxxx-xxxx
xxxx-xxxx
Geotechnical
Journal
Vol. 6
Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx

Drying prior to sealing

Drying
Drying
prior
prior
to sealing
to sealing
Drying prior to sealing

Moisture
Moisture
content
content
content
Moisture
content
weq
Equilibrium moisture content
weqweq
Equilibrium
Equilibrium
moisture
moisture
content
content
weq
Dryingmoisture
after
Equilibrium
content
sealing
Drying
Drying
after
after
Starting from a
sealing
sealing
Drying after
Starting
Starting
from
from
a a
dry
condition
sealing
drydry
condition
condition
Starting from a
dry condition

No. 1 2014 89

t eq
t eqt eq
Time to reach
equilibrium
m.c.
eq
Time
Time
totreach
to reach
equilibrium
equilibrium
m.c.
m.c.
Time to reach
equilibrium m.c.

Seasonal wetting (w wet)

Seasonal
Seasonal
wetting
wetting
(w(w
) )
wetwet
drying
SeasonalSeasonal
wetting (w
wet) (w dry )
Seasonal
Seasonal
drying
drying
(w(w
) )
dry dry
Seasonal drying (w dry )

Time
Time
Time
Time

Figure 4. Schematic representation of progressive moisture variation

Figure
4. 4.
Schematic
Schematic
representation
representation
ofof
progressive
progressive
moisture
moisture
variation
variation
ofFigure
a subgrade
after compaction.
ofFigure
of
a subgrade
a subgrade
after
after
compaction.
compaction.
4. Schematic
representation of progressive moisture variation
of a subgrade after compaction.
4.
BEHAVOURAL
CHANGES
WITH VARIATION
MOISTURE
4.4.
BEHAVOURAL
CHANGES
WITH MOISTURE
CHANGES
WITH
4. BEHAVOURAL
BEHAVOURAL
CHANGES
WITH MOISTURE
MOISTURE
VARIATION
VARIATION
4. VARIATION
BEHAVOURAL
CHANGES
WITH MOISTURE
A major problem
with pavements is rutting that can develop when the
A major VARIATION
problem with pavements is rutting that can develop when
pavement
subgrades
become
weakerisdue
to moisture
ingress.
Within
A
major
problem
with
that
develop
when
A
major
problem
withpavements
pavements
isrutting
rutting
that
can
develop
when
the pavement subgrades
become weaker
due
to can
moisture
ingress.
MPK
framework,
an
example
of
a
major
wetting
event
is
shown
Figure
the
pavement
subgrades
become
weaker
due
to
moisture
ingress.
the
pavement
subgrades
become
weaker
due
to
moisture
ingress.
A major
problem
with pavements
is of
rutting
that wetting
can develop
when
Within
MPK
framework,
an example
a major
event
is
MPK
framework,
ofofaat amajor
event
is is
5,Within
where
aFigure
soil
specimen
moisture
content
4.04%
(econtent
= event
0.107)
Within
MPK
anexample
example
major
wetting
the pavement
subgrades
become
weaker
due
towetting
moisture
ingress.
w
shown
5,framework,
whereata aan
soil
specimen
aof
moisture
of
specimen
a major
content
ofof
shown
5,where
where
asoil
soil
specimen
amoisture
moisture
content
Within
MPK
an
example
ofat aatof
wetting
event
is
is shown
compacted
a5,framework,
stress
of a1000
kPa,
to
20
kPa
then
ewFigure
=to0.107)
is compacted
to unloaded
a stress
1000
kPa,stress,
unloaded
4.04%
(Figure
shown
5,w =where
a moisture
soil specimen
atofat
a1000
moisture
content
of
e
e
=w =
0.107)
0.107)
is0.627)
is
compacted
compacted
to toa content
stress
a stress
of
1000
kPa,
kPa,unloaded
unloaded
4.04%
(23.67%
4.04%
(wFigure
wetted
to
(e
this
stress
level
and
to 20 kPa stress, then wetted to 23.67% (ew = 0.627) moisture
=stress,
0.107)
compacted
to23.67%
a stress
unloaded
4.04%
(ewthis
to
to
2020
kPa
stress,
then
then
wetted
wetted
to23.67%
(wof
ew=1000
=0.627)
0.627)
moisture
moisture
finally,
compressed
to
ais
stress
of to
2000
kPa
at(ethat
moisture
content.
content
atkPa
stress
level
and
finally,
compressed
tokPa,
a stress
of
content
content
atthis
this
stress
stress
level
level
and
and
finally,
compressed
compressed
tothat
toa astress
stress
ofaof
to
20
stress,
then
wetted
to
23.67%
(ecan
0.627)
moisture
This
wetting
can
be
considered
a finally,
major
wetting
event
occurred
w =be
2000
kPaatkPa
at
that
moisture
content.
This
wetting
considered
2000
2000
kPa
kPa
at
at
that
that
moisture
moisture
content.
content.
This
This
wetting
wetting
can
can
be
be
considered
considered
content
at
this
stress
level
and
finally,
compressed
to
a
stress
during
operation.
Figure
2b
also
shows
a
subsequent
loading
which
major wetting event that occurred during operation. Figure 2b alsoaofa
major
major
wetting
wetting
event
event
that
thatoccurred
occurred
during
during
operation.
operation.
Figure
2b2balso
also
2000
at that
moisture
content.
This
wetting
canFigure
be
considered
can
again
intercept
the
LWSBS,
where
soil
can
undergo
yielding
ata
shows
akPa
subsequent
loading
which
can
again
intercept
the
LWSBS,
shows
shows
a
subsequent
a
subsequent
loading
loading
which
which
can
can
again
again
intercept
intercept
the
the
LWSBS,
LWSBS,
major
wetting
event
that
occurred
during
operation.
Figure
2b
also
where stress
soil can
undergo
yielding
a lower stressThis
level,
leadingmay
to
a lower
level,
leading
to extraat deformation.
situation
where
where
soil
can
canundergo
undergo
yielding
yielding
atmay
atacan
lower
a be
lower
stress
stressto
level,
level,
leading
leading
to to
shows
asoil
which
again
the
LWSBS,
deformation.
situation
likened
in the
beextra
likened
tosubsequent
ruttingThis
inloading
the
pavement
that
may intercept
occurrutting
after
the
soil
extra
extra
deformation.
deformation.
This
This
situation
situation
may
may
be
be
likened
likened
to
to
rutting
rutting
in
in
the
the
where
soil
can
undergo
yielding
at
a
lower
stress
level,
leading
to
pavement
may
occur after
soil subgrade
has been
weakened
subgrade
hasthat
been
weakened
bythe
a the
significant
wetting
event
or
phase.
pavement
pavement
that
that
may
may
occur
occur
after
after
the
soil
soil
subgrade
subgrade
has
has
been
weakened
weakened
extra
deformation.
This
situation
may
becorresponding
likened
tobeen
rutting
in the
by
a
significant
wetting
event
or
phase.
The
pathways
The
corresponding
pathways
in
e-log(p)
plots
are has
shown
inweakened
Figure
bypavement
by
a significant
a significant
wetting
event
event
oror
phase.
phase.
The
The
corresponding
corresponding
pathways
pathways
thatwetting
may
occur
after
the
soil
subgrade
been
e-log(p)
plots
are
shown
in
Figure
6.
reduction
yield
stress
6.in
The
reduction
inwetting
yield
stress
due
to6.The
wetting
can bein
clearly
seen
in
in
e-log(p)
e-log(p)
plots
plots
are
are
shown
shown
in
in
Figure
Figure
6.
The
The
reduction
reduction
in
in
yield
yield
stress
stress
by
a
significant
event
or
phase.
The
corresponding
pathways
due to wetting can be clearly seen associated with excessive
associated
with
excessive
deformation.
Figure
7
shows
this
effect
in a
due
due
to
to
wetting
wetting
can
can
be
be
clearly
clearly
seen
seen
associated
associated
with
with
excessive
excessive
in
e-log(p)
plots
are
shown
in
Figure
6.
The
reduction
in
yield
stress
deformation. Figure 7 shows this effect in a plot of deformation
plot
of
deformation
against
traffic
load
repetitions.
deformation.
deformation.
Figure
Figure
7
7
shows
shows
this
this
effect
effect
in
in
a
a
plot
plot
of
of
deformation
deformation
due
to
wetting
can
be
clearly
seen
associated
with
excessive
against traffic load repetitions.
against
against
traffic
traffic
load
load
repetitions.
repetitions.
deformation.
Figure
7 shows this effect in a plot of deformation
against traffic load repetitions.

Figure 6. Stress path during loading in e-log(p) plane

Figure
Stress
path
during
loading
e-log(p)
plane
Figure
6. 6.
Stress
path
during
loading
in in
e-log(p)
plane
Figure 6. Stress path during loading in e-log(p) plane

Figure 7. Plastic deformation accumulation with traffic load cycles.

Figure
Plastic
deformation
accumulation
with
traffic
load
cycles.
Figure
7. 7.
Plastic
deformation
accumulation
with
traffic
load
cycles.
Figure
7.
Plastic
deformation
accumulation
with
traffic
load
cycles.
The above analysis within the current MPK framework assumes
Theabove
aboveanalysis
analysiswithin
within
thecurrent
current
MPK
frameworkassumes
assumes
The
MPK
volumetric
deformation
of the
the
soil
under
Kframework
o conditions only.
conditions
conditions
only.
only.
volumetric
deformation
of
the
soil
under
K
volumetric
deformation
of
the
soil
under
K
The
above
analysis
within
the
current
MPK
framework
assumes
o
o
However,
significant
deformation
canMPK
occurframework
under pavement
The
above
analysis shear
within
the current
assumes
However,
However,significant
significant
shear
shear
deformation
deformation
can
canoccur
occur
under
pavement
pavement
conditions
only.
volumetric
deformation
of
the
soil
under
Ko under
conditions,
since
the
soil
is
not
laterally
confined
as
in
Ko
volumetric
deformation
ofisdeformation
under
Ko under
conditions
conditions,
conditions,Under
since
sincethese
thetheshear
soil
soil
isthe
notnotsoil
laterally
laterally
confined
confined
as
as
inarise
inKonly.
However,
significant
candeformation
occur
pavement
oKo
conditions.
conditions,
further
can
However,
significant
deformation
can
occur
under
pavement
conditions.
conditions.
Under
Under
these
these
conditions,
conditions,
further
further
deformation
deformation
can
can
arise
conditions,
since
theshear
soil
is not
laterally
confined
as
inarise
Ko
when
the shear
stress
approaches
close
to the
failure
stress
with
conditions,
since
the
soil
isconditions,
not laterally
confined
as in stress
Kstress
conditions.
when
whenthethe
shear
shear
stress
stress
approaches
approaches
close
to
the
failure
failure
with
conditions.
Under
these
further
deformation
arise
o canwith
wetting.
For
example,
this
situationclose
may
betothe
depicted
as shown
in
Under
conditions,
further
deformation
can
arise
when
thewith
shear
wetting.
wetting.
For
For
example,
example,
situation
situation
may
may
betobedepicted
depicted
as
asshown
shown
in
in
when8,these
the
shear
stressthis
approaches
close
the
failure
stress
Figure
which
shows
thethis
shear
failure
envelopes
corresponding
to
stress
approaches
close
to
the
failure
stress
with
wetting.
For
example,
Figure
Figure
8,
8,
which
which
shows
shows
the
the
shear
shear
failure
failure
envelopes
envelopes
corresponding
corresponding
to
to
wetting.
For
example,
this
situation
may
be
depicted
as
shown
normal and excessive moisture conditions. Also shown is the likely in
normal
normal
and
and
excessive
excessive
moisture
conditions.
conditions.
Also
shown
shown
is
thethe
likely
likely
Figure
8,
which
shows
the shear
corresponding
to
this
situation
may
bemoisture
depicted
asfailure
shown
inAlso
Figure
8,isthat
which
shows
stress
increase
during
traffic
loading.
It envelopes
can
be
seen
under
stress
stress
increase
increase
during
during
traffic
traffic
loading.
loading.
It
It
can
can
be
be
seen
seen
that
that
under
under
normal
and
excessive
moisture
conditions.
Also
shown
is
the
likely
the
shear
failure
envelopes
corresponding
to
normal
and
excessive
normal moisture conditions, the stress state with traffic loading
normal
normal
moisture
conditions,
conditions,
thethe
stress
stress
with
with
traffic
loading
loading
stress
increase
during
traffic
loading.
Itstate
can
be traffic
seen
that
under
moisture
conditions.
Also
shown
is
thestate
likely
stress
increase
during
(point
B)moisture
is
well below
the
failure
surface
and
therefore,
soil
may
(point
(point
B)
B)
is
is
well
well
below
below
the
the
failure
failure
surface
surface
and
and
therefore,
therefore,
soil
soil
may
may
normal
moisture
conditions,
the
stress
state
with
traffic
loading
behave
predominantly
a resilient
manner
traffic loading.
It can beinseen
that under
normalwithout
moisturecontinual
conditions,
behave
behave
predominantly
indeformations
in
aloading
afailure
resilient
resilient
manner
manner
without
without
continual
continual
(point
B)predominantly
is of
well
below
the
surface
therefore,
soilline
may
accumulation
plastic
(similar
thebelow
dashed
the
stress
state
with
traffic
(point
B) and
is to
well
the
failure
accumulation
accumulation
of
of
plastic
plastic
deformations
deformations
(similar
(similar
to
to
the
the
dashed
dashed
line
line
behave
predominantly
in
a
resilient
manner
without
continual
shown
Figure
7. When
the subgrade
is excessivelyin wetted,
surfaceinand
therefore,
soil may
behave predominantly
a resilient
shown
shownin in
Figure
Figure
7.
7.When
When
the
thesubgrade
subgrade
is state
isexcessively
excessively
wetted,
wetted,
accumulation
of
plastic
deformations
(similar
tointhe
dashed
line
however,
traffic
loading
will
bring
the
stress
the
subgrade
manner
without
continual
accumulation
of
plastic
deformations
however,
however,
traffic
traffic
loading
loading
will
will
bring
bring
the
the
stress
stress
state
state
in
in
the
the
subgrade
subgrade
shown
in
Figure
7.
When
the
subgrade
is
excessively
wetted,
much
closer
to failure,
allowing
significant
accumulation
of plastic
(similar
to traffic
the
dashed
line
shown
in
7.
When
the
subgrade
much
muchcloser
closer
to to
failure,
failure,
allowing
allowing
significant
accumulation
accumulation
ofof
plastic
plastic
however,
loading
will
theFigure
stress
state
in the
subgrade
deformations,
which
can
leadbring
tosignificant
rutting
deformations
on
the
is
excessively
wetted,
however,
traffic
loading
will
bring
the
stress
deformations,
deformations,
which
which
can
can
lead
lead
to
to
rutting
rutting
deformations
deformations
on
on
the
much
closer
to
failure,
allowing
significant
accumulation
of
plastic
pavement (similar to the steep line for excessive moisture shown
inthe
state
in the
subgrade
much
closer
to
failure,
allowing
significant
pavement
pavement
(similar
(similar
to
to
the
the
steep
steep
line
line
for
for
excessive
excessive
moisture
moisture
shown
shown
inthe
in
deformations,
which
can
lead
to
rutting
deformations
on
Figure 7).
accumulation
of plastic
deformations,
which can
lead shown
to rutting
Figure
Figure
7).7). (similar
pavement
to the steep
line for excessive
moisture
in
Figure 5. Depiction of a soil specimen compacted and then unloaded
Figure
5. 5.
Depiction
Depiction
of
a soil
athen
soil
specimen
specimen
compacted
compacted
and
and
then
then
unloaded
unloaded
deformations
Figure 7). on the pavement (similar to the steep line for excessive
to Figure
operational
stressofand
wetted due
to a major
wetting
event
to
tooperational
operational
stress
stress
and
then
wetted
wetted
due
to toa major
a major
wetting
event
event moisture shown in Figure 7).
Figure
5. Depiction
ofand
aatthen
soil
specimen
compacted
andwetting
then
unloaded
and
subsequently
loaded
wetted
state.due
and
and
subsequently
subsequently
loaded
loaded
at at
wetted
wetted
state.
state.due to a major wetting event
to
operational
stress
and
then
wetted
and subsequently loaded at wetted state.

Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx

90 Geotechnical Journal Vol. 6 No. 1 2014

pressure generation under traffic loading since the CBR test is

normally performed at a faster rate (e.g., 1 mm/minute). However,
in the current design approaches, these stress paths and the stress
levels are not directly accounted for in deformation accumulation,
such as in foundation design, where finite element analysis can be
performed to examine the operating settlements, commonly under
monotonic loading. A robust three-dimensional model with the
capability to undertake cyclic loading is required to do such
analysis, but with current models based on suction, this approach
has been found to be difficult.

Figure 8. Influence of stress conditions due to excessive wetting as

soil shear strength reduces.
5.

CURRENT
MODELING

PRACTICE

OF

MECHANISTIC

Current practice in pavement design mostly involves the treatment

of subgrade as an elastic layer (in some cases infinitely deep). In
many design guides, the damage accumulation is treated as in
fatigue, where the allowable repetitions of (vertical) stress or strain
on the subgrade surface due to traffic loading on the pavement are
computed (as in Equation 1). The vertical stress or strain on the
subgrade surface is computed using a numerical program (such as
finite or boundary element) assuming elastic pavement layers. An
inherent assumption is that soil will behave well within the influence
of yield limit under repetitive traffic loading, as shown in normal
moisture conditions (1) in Figure 8 resulting in deformation
behaviour similar to (1) in Figure 7. The elastic modulus is normally
the resilient modulus measured in a repetitive loading triaxial
apparatus subjected to a certain stress ratio (ratio of applied to
failure stress) at a designed moisture content. However, in many
situations, the resilient modulus is approximated as a multiple of
CBR at that moisture content. In the Austroads Design Guide
(2010)), modulus in MPa is assumed to be equal to 10 times the
CBR as a percentage value. Because of the lack of certainty of
design moisture content allowing field moisture variations, it is also
common to use a soaked CBR, where the soil specimen is soaked
for a period of time prior to performing the CBR test. Since soaked
CBR is generally much smaller than as-compacted CBR, it is
generally considered that the worst case scenario is used for
design to offset uncertainty in field moisture variation, thereby
giving a conservative design.
Figure 9 shows the variation of soaked CBR in the unloaded (at
nominal stress) compaction plane for basaltic clays in Victoria,
Australia. It is apparent that the same CBR can be obtained at lower
densities when the compacted moisture content is increased. The
reason for this is that along the same CBR contour (say 1% CBR in
Figure 9) at a lower void ratio (higher density) and lower moisture
content, the percentage swelling would be higher, as can be seen in
Figure 3. Therefore, along the same CBR contour (say 1%), the
soaked void ratios are brought closer to each other, although their
positions are far apart at the compaction states. Therefore, similar
CBRs can result for soaked states. In order for that to happen, the
soil may have to behave elastically due to the stress increase
applied. The stress increase is related to the CBR value, since the
CBR gives the force applied. For example, for CBR 1%, the force
applied may be given as 1% of the Standard load 19.8 kN (assuming
this gives the higher CBR value) applied over a circular area of 49.6
mm diameter. This gives rise to a stress increase of about 100 kPa.
Therefore, if the pavements are designed with subgrade properties
assumed to be in soaked CBR state and the applied stress on the
subgrade remains well below such stress levels (e.g., 100 kPa), the
subgrade may not experience significant rutting. However, under
soaked conditions, a major influence is also expected from pore

Figure 9. Variation in CBR in unloaded e-ew plane (adapted from

Hillard, 1981).
6.

POTENTIAL EFFECTS OF CLIMATE CHANGE

It is commonly accepted that the worlds climate is changing based

on measurements, although its course remains a matter of debate.
The likely influence is considered to vary among countries, but it is
predicted that rainfall intensities may increase, despite overall
rainfall increasing or decreasing depending on the geographical
location. As the worlds climate changes, a change in equilibrium
moisture content (or suction) with time can be expected. For those
areas where rainfall becomes less, a decrease in equilibrium
moisture content can be expected, and for those areas where higher
rainfall is expected, an increase in equilibrium moisture content can
be expected. This overall effect is somewhat subdued since it may
happen over a long period (over 50 to 100 years in the medium short
term, where climate predictions are normally done). However, it can
be argued that the major effect will come from the increase in
rainfall intensity and its recurrence. As noted previously, major
wetting has the potential to weaken the subgrade significantly. If the
wetting conditions prevail for longer periods of time, significant
damage could accumulate during traffic passage.
Under such uncertain wetting conditions, maintaining an intact seal
on the pavement will be paramount. During flooding, water ingress
pathways are through either the road surface or from the shoulders
and sides. Even when the road is flooded, if the road has a good
intact seal, the water ingress from the road surface will be slow, and
the lateral ingress will depend on the permeability of the pavement
and subgrade materials. Therefore, if water ingress into the subgrade
under the wheel path areas can be minimised during the flooding
period by the appropriate design of pavement layers, it may be
possible to reduce pavement damage. If the water table is close to
the surface, or the water table has risen with the continual increase
in rainfall, it is inevitable that pavement subgrades will have high
moisture contents. Under such scenarios, elevating the pavement
base would be an option.

3rd Proff 18-02-2015

7.

Sri Lankan Geotechnical Society International Conference SLGS Vol. ?? No.?? ?? 2015 ISSN xxxx-xxxx

CONCLUDING REMARKS

This paper has presented a brief analysis of the effects of climatic

effects on the performance of clayey subgrades of road pavements.
The analysis was primarily explained within the MPK framework
developed at Monash University. It was highlighted that when clay
subgrades are compacted dry of optimum, they have the potential to
swell or collapse under load. However, collapse may occur only
when the compaction levels are small or when operational stresses
are high during wetting. In road pavement scenarios, where the
subgrade is only loaded significantly under heavy traffic loading,
crucial conditions could occur when the subgrade has swelled
during wetting and undergone deformation under subsequent
loading. Collapse can become a problem in clay fills in approaches
to overpasses.
The current use of CBR in pavement design is an approximate
method of ensuring the pavement subgrade does not undergo
excessive deformations during repetitive traffic loading. However,
the connection of this parameter to the soil constitutive behaviour as
needed in advanced mechanistic modelling is only loosely
established. Changes in climatic effects have highlighted the need to
consider the likely moisture variations in subgrade to be expected in
the design life of a pavement. To cater for these conditions, more
advanced but robust numerical models and bottom-up simplified
analyses are required. The proposed MPK framework may provide a
vehicle to achieve this objective.
8.

REFERENCES

Austroads (2010). Guide to Pavement Technology Pavement

Structural Design, Austroads, Sydney, Australia.
Burman, B.C., Mostyn, G. and Piccolo, P. (2008). Experiences
with post-construction retesting of engineered clay fills,
Australian Geomechanics, 43 (4): 1-30.
Hillard, P. (1981). Western Freeway Melton Bypass Section,
lime stabilisation of the clay subgrade, Road Construction
Authority, Victoria, Materials Division Report No. 18X429,
December, 1981.
Holtz, W. G., and Gibbs, H. J. (1956). "Engineering properties of
expansive clays: Transactions." ASCE, 121, 641-677.
Kodikara, J., Islam, T., Wijesooriya, S., Bui, H. and Burman, B.C.
(2014). On controlling influence of the line of optimums on
the compacted clayey soil behavior, Unsaturated Soils:
Research and Applications, Proceedings of the Sixth
International Conference on Unsaturated Soils, UNSAT2014,
Sydney, Australia, 2-4 July, Editors: N. Khalili, A.R. Russell
and A. Khoshghalb, pp. 219-225.
Kodikara, J. (2012). New framework for volumetric constitutive
behaviour of compacted unsaturated soils, Canadian
Geotechnical Journal, Vol. 49, pp. 1227-1243.
Kodikara, J. and Rahman, F. (2002). Moisture content and hydraulic
conductivity relations for compacted clay, Civil Engineering
Transactions, IEAUST, Vol. CE43, pp. 13-18.
Russam, K and Coleman, J.D. (1961). The effect of climatic factors
on subgrade moisture conditions, Geotechnique, Vol. XI, No.
1, pp. 22-28.

Geotechnical Journal Vol. 6 No. 1 2014 91