CHAPTER - 2

LITERATURE
REVIEW
CHAPTER 2 [LITERATURE REVIEW]

2.1 GENERAL

In this chapter review of works reported in literature is discussed on the

different approaches for design and analysis of flexible pavement. As mentioned in
Chapter 1, various empirical methods are evaluated from a certain set of environmental
and material conditions for analyzing flexible pavements.

2.2 LITERATURE REVIEW

Pavement responses such as stress, strain, and deflection caused due to load,
material properties, climatic condition, etc. are determined in mechanistic design,
which is then empirically correlated to the pavement performance. A number of such
mechanistic pavement response models have been developed over the years, ranging
from Boussinesq's one-layer model to multi-layer elastic theories to finite element
models. Boussinesq formulated a simple relationship for determining stresses, strains
and deflections of a homogeneous, isotropic, linear, elastic half-space (infinitely large
area and depth) with Young's modulus and Poisson's ratio subjected to a static point
load [21], Later, Boussinesq's simple closed form equations were fiirther modified by
Yoder and other researchers for a uniformly distributed load by integration [21]. Yoder
and Witczak suggested that Boussinesq theory can be used to estimate subgrade
stresses, strains, and deflections when the base and the subgrade have similar stiffness
[62]. Bunnister developed a solution for two layered and subsequently for a three
layered pavement system [21]. Duncan et al. first applied the finite element method for
the analysis of flexible pavements [15, 21]. Several computer programs based on
Burmister's layered analytical elastic models and finite element method have been
developed to study the flexible pavement response. Various programs such as CHEV
developed by Chevron Research Company, DAMA developed by the Asphalt Institute,
JULEA which is a modified version of WESLEA, BISAR developed by Shell,
ELSYM5, PDMAP (Probabilistic Distress Models for Asphalt Pavements),
DIPLOMAT, ILLI-PAVE, MICH-PAVE developed at Michigan state University [19],
KENLAYER are used. General purpose softwares like ABAQUS, ANSYS, etc are also
used in practice [ 18, 21 ].
A good example of the use of regression equations for pavement design is the
AASHTO method. The design equations published in 1961,,r+fp5;K»86 and 1993 by

\Title: Analysis of flexible pavement by finite element method

 [LITERATURE REVIEW]

AASHTO (American Association of State Highway and Transportation) is based on the

results of the extensive AASHO road test conducted in Ottawa, Illinois, in late 1950's
and early 1960's. In 1984-85, AASHO committee revised and expanded the guide
under NCHRP project, which was issued in 1986. The guide was revised in 1993 with
practically no change in the design method [3, 21].
Helwany Sam et al. [20], illustrates the usefulness of finite element method by
discretising a three layered pavement system with one of the boundary at a distance of
about 8 times the loaded radius subjected to different types of loading. The author
carried out two and three dimensional analysis. In this analysis three layers were
assigned the same elastic modulus, transforming the three systems into a simpler single
layer system. According to the author, the analytical solution of the single layer system
using software DACSAR can be compared with the Boussinesq close form solution
readily available for uniform circular loads. Finite element modeling of pavements if
validated can be useful because it can be used directly to estimate primary responses
instead of performing the costly field experiments. It is also noted that, for a simple
configuration, the resuhs obtained from three-dimensional analysis are nearly identical
to the results of the two-dimensional analysis.
Bassam Saad et al. [6] reported the use of three-dimensional finite element
analysis (FEA) for elastoplastic models with ADINA program. The work focuses on
the analysis of conventional foundation systems flexible pavement, dynamic response,
corresponding to fatigue, strain at the bottom of the asphalt concrete layer and rutting
strain at the top of subgrade material. A convenient FEA model was developed
considering the domain size and traffic load simulation, which was then examined
using sensitivity analysis to interpret the influence of foundation parameters on
pavement response. The effect of use of elastoplastic and linear elastic material was
compared as foundation materials of pavement. From the analysis, it is noticed that the
elastoplasticity caused an increase in rutting strain, maximum tensile strain, and
maximum vertical surface deflection. The model sensitivity analysis considering
elastoplastic materials highlights that subgrade material has little or no impact on
fatigue strain but reduces rutting strain remarkably. Regardless of subgrade quality, the
analysis performed by Saad et al. [7] illustrates that a strong base significantly
decreases fatigue and strain for thick and thin bases as compared to the weak base.
However the reduction in base thickness resulted in increase in fatigue strain where as
it resulted in constant increase in rutting strain.
I Title: Analysis of flexible pavement by finite element method \7\fim 10
CHAPTER 2 [LITERATURE REVIEW]

Diefenderfer et al. [13] developed a model for predicting the pavement

temperature, which is beneficial to those who need to determine the pavement
temperature profile in order to calculate in situ pavement engineering characteristics.
By using ambient temperature trends from historical records, these models can predict
anticipated future pavement temperatures and thus aid researchers in determining the
amount of time that pavements are subjected to critical temperatures.
Immanuel and Timm [22] used layered elastic analysis to compare predicted
vertical stress in the base and subgrade layers to vertical pressures at field obtained
from the National Center for Asphalt Technology (NCAT) Test. The authors found that
the predicted pressure was only a reasonable approximation up to vertical pressures of
82 kPa in the base and 48kPa in the subgrade. Wu et al. [60] calculated the vertical
stress at the bottom of base layer using the multilayer elastic program ELSYM5 and
compared it with the data from the Louisiana ALF. The study found that the calculated
vertical stresses were two to eight times higher than the measured field values. Zuo et
al. [65] observed seasonal temperature and water content variations and evaluated their
effects on predicted pavement life. The results of the parametric study showed that the
temperature averaging period, the temperature gradient in the asphalt, and the timing
and duration of wet base and subgrade conditions affect the estimation of pavement
life.
Sanaa Masad et al. [56] focused on comparative analysis of isotropic and
anisotropic models in flexible pavement along with an evaluation of permanent
deformation and fatigue cracking by using NCHRP 1-37A design guide [33, 34].
Anisotropic properties were assumed for unbound base and sub-base layers to measure
surface deflection. It reduced unrealistic tensile stresses which were assumed in the
isotropic model in a granular base. Finite element predictions were comparatively
analyzed with AASHO road test measured deflections. The anisotropic nonlinear model
depicted longer fatigue lives which reduces the need for large shift factors while
establishing a relationship between empirical laboratory calculations to field
conditions. The study highlights total permanent deformation in asphalt layers
considering anisotropic properties in NCHRP 1-37A model. For the anisotropic model
high tensile stresses were calculated in the lower regions of asphalt layers and total
rutting of the model was more than an isotropic model. The Mechanistic-Empirical
Pavement Design Guide (MEPDG) recommends use of the linear multilayer elastic
program JULEA, to compute flexible pavement responses [33, 34]. Pavement analysis
I Title: Analysis of flexible pavement by finite element method I^UTJ 11
CHAPTER 2 [LITERATURE REVIEW]

procedures that are derived from the theory of elasticity are based on simplifications of
the real condition.
R.M. Mulungye et al. [43] generated a two-dimensional four layer model using
ANSYS/ED to evaluate structural performance by finite element method considering
factors like tyre pressure, wheel configuration and axle load variations of a
transportation truck. The results were verified with in situ full-scale test data. Andrew
Chiasson et al. [4] propose a simulation model that utilizes transient, two-dimensional
finite difference method to calculate the temperature distribution in asphah pavements
in relation to hourly thermal environmental conditions and to calculate associated
thermal stresses. The study uses the stiffness index to evaluate the thermal stress
distribution in asphalt pavements and initial thermal stress maps during a seasonal and
diurnal freeze and thaw cycles.
Kai Su et al. [29] studied evaluation of shear stress in the pavement, so as to
know the actual influence of tyre-pavement contact pressure on pavement shear stress
in asphalt mix. For analysis in 3D FE model, single axle load with dual tyres was
applied. Shear stress value comes maximum at tyre edge. The maximum shear stress
occurs as 60mm below the tyre edge.
Ranadive M.S et al. [46] illustrates that axisymmetric analysis of flexible
pavement is carried out by a computer program ANSYS and different performance
parameters of pavement were studied for varying conditions of thickness. The mcrease
in thickness of the base course and sub-base course layer does not help to reduce stress
and deflection as compared to asphalt concrete layer. Also, it is reported that the
maximum deflection of flexible pavement under wheel load varies with an increase in
thickness of the different layers. The maximum deflection reduces when BC layer
thickness increases, while it increases when base-course and sub-base course thickness
increases. Analytical methods can predict the performance of flexible pavement, which
helps to bypass costly field experiments; hence it is beneficial to society. Umesh
Chandra Sahoo et al. [59] studied the effect of nonlinearity in a granular layer on
critical pavement responses of low volume roads.
Rahman M.T et al. [45] performed the finite element analysis using ABAQUS
focusing on pavement response subjected to various traffic factors like different axle
configuration, tyre imprint areas, and inflation pressure. Three layers of pavement were
considered as asphalt surface, granular base and subgrade which were assumed to
respond linearly and elastically to the static applied load.
\Title: Analysis of flexible pavement by Jinite element method iTT-Ta 1 2
CHAPTER 2 [LITERATURE REVIEW]

Guangming Wang et al. [17] analyzed the effect of truck tyre types based on
tyre geometries and specifications obtained from tyre manufacturers. Two dimensional
finite element models were generated using tyre geometries and structure along with
tyre pavement interaction models. The material property was considered as linear
elasfic for single load application. Software ADINA version 8,3 was used to generate a
2-D model usingfiniteelement analysis.
The IRC: 37-2012 [24] guidelines for the design offlexiblepavement is based
on a mechanistic empirical approach recommends use of the software IITPAVE for
analysis of layered system. Ripunjoy Gogoi et al. [52] studied the correlation if any
between the rutting and fatigue distress. The study does not give any cause or effect of
these stresses but proposes two regression equations which can be used to forecast their
progression. K. Kranthi Kumar et al. [28] studied the use of reclaimed asphalt
pavement (RAP) in bituminous mixes. The RAP was mixed with both hot and cold
bitumen mixes and tested for different mix proportions.
Abdhesh K.Sinha et al. [1] Illustrates the usefulness of FEM to study the
performance of a flexible pavement with different types of local materials in its sub-
base. ANSYS software was used for the FEA and the effect of type of subbase on a life
of the pavement is evaluated. In the study, arightboundary was placed at 110cm from
the outer edge of a loaded area, which is more than 7 times the radius 150mm of the
applied load. A uniform pressure of 575kPa was applied to a circular contact area
having a radius of 150mm causing a single axle load of 40.80kN. Tapase and Ranadive
[57] reported the usefulness of two-dimensional FEA to study the effect of variation in
thickness of different component layers on the critical parameters.

2.3 CLOSURE

A literature survey showed that considerable progress has been made in the
development of response models for mitigating various parameters influencing
pavement performance, ranging from Boussinesq's single-layer model to multi-layer
elastic theories to finite element models. However, it is observed that some of the areas
related to design of pavement are still not addressed properly. Some of the debatable
issues are discussed below which have been not paid full attention by the researchers:
A direct or indirect application of empirical approach in current design
procedure, which is limited to a certain set of environmental and material conditions, is

I Title: Analysis of flexible pavement by finite element method 1 ^ 1 3 13

CHAPTER 2 [LITERATURE REVIEW]

responsible for designing an uneconomical or unsuitable pavement section. Even the

IRC: 37-2012 design guidelines follow the design charts which are extracted for
particular set of conditions. Due to limitations of empirical methods, design is no
longer valid if there is a change in condition. A good pavement design is one that
provides the expected performance for actual field and environmental conditions with
appropriate economic consideration. So, here the need arises to find an aUemative in
the form of analytical tool which can be used to design a pavement based on actual
field data and environmental conditions, instead of following particular design chart
which is developed for particular set of conditions.
The IRC: 37-2012 [12] guidelines for the design of flexible pavement which is
based on a mechanistic empirical approach do not form a rigid standard and sound
engineering judgement considering the local environment and past pavement
performance in the respective regions. So here the need arise to study the behavior of
local environmental conditions while selecting the composition of pavement. Secondly,
the performance of flexible pavement under various combinations of thicknesses and
material properties of component layers is different, so here need is to select and design
an optimum thickness along with use of suitable and economical material for different
component layers.
The growing intensity of commercial vehicles, overloading of trucks beyond
double its capacity and change in daily and cyclic temperature and environmental
factors has been responsible for reducing the life of pavement. A factor which causes
further concern in India is low pavement temperatures in some parts of the country.
Under these conditions, flexible pavements tend to become soft in summer and brittle
in winter [12]. Both positive and negative changes in temperature during day time and
night time as well as due to seasonal changes affects the pavement performance
adversely, so here need is to consider the effect of temperature along with freezing and
thawing cycles during design of pavement. Temperature at any point in the layer
system, the effect of the forces arising out of frost and thawing cycles, the entities such
as the distribution of seepage pressures and gradients should be known before actual
construction of pavement.
The effect of excess wheel loading and modem tyre pressures carmot be
neglected. However, from the past studies and through the recommendations of the
Indian Road Congress and AASHTO code of practice, wheel load, tyre pressure and

\Title: Analysis of flexible pavement by Jinite element method iJm 14

CHAPTER 2 [LITERATURE REVIEW]

their combinations can be shortlisted which can be correlated with actual field
conditions.
The largely used conventional materials like aggregate and bitumen, which can
be partially or completely replaced by alternative materials or waste materials like fly
ash, granulated blast furnace slag, plastic, rubber, e-waste, etc. Hence, the scope is to:
analyze the suitability of waste materials in different component layers of flexible'
pavement.
From the above literature, it is noticed that emphasize is given to the factors
responsible for the deterioration of pavement in the form of fatigue and rutting which is
theoretically related to California Bearing Ratio (CBR) values of the subgrade. Thus
there is a need to estimate the pavement life in the form of a number of standard axles
or cumulative standard axles; which can be correlated with the fatigue life and ruttmg
life for the actual subgrade condition [23,24].
Development of a tailor made procedure/charts for design and analysis of'
flexible pavements using finite element method which will enhance the scope to use^
variety of combinations of materials, conditions, and various parameters which]
correlates with the actual conditions on the field is the thrust area of present research.
Even though researchers presented a number of improvements in the design and
analytical procedures, still there is a need for development of a morerigorousanalytical
tool which will be helpful to study the scope of work defined in chapter 1. This alarms
for more extensive research in this area in the coming years for the design of
sustainable flexible pavements. Number of shortfalls which are explained above are
tried to address in the present work. , ^ { j •'^
I,

t
1 \i- I

I Title: Analysis of flexible pavement by finite element method ii^i'.i-M 1 5