Performance Analysis of Flexible Pavement Using Finite Element Method
Performance Analysis of Flexible Pavement Using Finite Element Method
Introduction:
LITERATURE REVIEW:
Beskou 2016
Hadi 2003
Mechanistic method uses layered pavement homogenous linear elastic and isotropic
Background
A FEM modeling is created to extend deacceleration effect along with nonlinear properties of flexible
pavement with a purpose for getting insight what to use for designing purpose based on iterative
modeling of section for economical and safe design. The modeling will be validated by comparing with
experimental data available then after the validation of modeling process modeling will conducted for
various properties to study the effect and purpose the insight for the design in purpose of finite element
method and investigates common pavement design parameters, including surface mixture type, base layer
thickness, base layer type, subbase layer thickness.
Elastic material behavior simplifies but result is not closer to experimental result So focus is given to fem
for the solution of problem in dynamic analysis non linearity heterogeneity fem is being more use due to
development of high-speed computer can be use in difficult geometry boundary conditions and material
which may be possible to get closed solution using analytical methods
The dynamic analysis of these models requires the development of new accurate and efficient numerical
methods of solution and their computer implementation
However, for the dynamic loading A 3d finite element model in ABAQUS for dynamic analysis of
pavement response to transient loading by Yoo and Al -Qadi while Beskou et al have created in ANSYS.
Beskou ,Hadi using nonlinear properties of material while under constant speed loading
As the linear with static is efficiency in analysis but the result obtain is not accurate as compare to actual
performance of flexible pavement the results of these models are rather qualitative, and even though they
help to understand the system behavior thus development of faster computer nonlinear problem can be
also computed efficiently using of nonlinear properties of pavement result of analysis closer to actual
response.
However, these are not able of representing the true reality of flexible pavement response.
Flexible pavement is the most common type of pavement constructed in Nepal. However, for
decades, it has encountered certain challenges which make them not to last up to half of their
proposed design years. There are several problems leading to this, which includes poor design
methodology adopted. Nepal as a country still makes use of empirical method of designs for
designing her pavement. This method has proven by researchers to be archaic and flawed with
numerous limitations. However, this research work is geared towards the implementation of
mechanistic empirical method of design by using finite element software. Dimensional finite-
element programs have been employed in the past two decades for analyzing road pavement
response. In recent years three-dimensional finite-element (3DFE) analysis emerged as a
powerful tool which is capable of capturing pavement response, can also present effect of
vehicle load response on road pavement. The model can be used to perform parametric studies
involving effect of vehicle load on road pavement on different layer of structures help to know
cause of various type of failure occurring on present road condition.
This FEM model can provide reference for the improvement of road pavement and materials
design and dealt with ways to reduce deflections by varying the design configuration, such as
increasing the HMA modulus, the base modulus, sub base modulus, the subgrade modulus and
increasing thickness of each layers.
Some advanced 3D Finite Element models for simulating the dynamic behavior of pavements have
been developed. Yoo and Al-Qadi [2007] developed a 3D Finite Element model in Abaqus for dynamic
analysis of the pavement response to transient Loading. They observed an increase in stresses and
strains of up to 39% due to dynamic effects compared to a corresponding static response. This was
supported by Beskou et al. [2016] who studied dynamic versus static behavior of pavement response
to moving load in a linear elastic 3D Finite Element model developed in Ansys. They found that the
response obtained by dynamic analysis is higher than the corresponding response obtained by static
analysis and that increasing vehicle speed increases the pavement response of dynamic analysis. In
both works the moving load is simulated by moving nodal loads from element to element across the
surface requiring a huge number of elements to ensure high resolution in the entire area the load is
moving. Hence, simulations are computational heavy especially if longer simulations are needed.
Since the 1940s the layered elastic theory has been the most common tool used to
calculate flexible pavement responses to truck loading. In 1943 Burmister et al. [1944]
developed a closed-form solution for a two-layered linear elastic half-space, which was
A number of computer programs have been developed based on the layered elastic theory. Some of
the more well known are Julea, Bisar and Elsym (Ullidtz [1998]). Some programs have introduced
modifications to the original layered elastic theory to cover e.g. viscoelastic material models (VERSUS)
or to adjust the bonding condition at layer interfaces (BISAR 3.0). However these modifications are
only valid under the previous mentioned assumptions for the original layered elastic theory.
Today it is still widely accepted to use static programs as the above mentioned for backcalculation of
material parameters from impact loading of the pavement generated by e.g. an FWD, even though a
number of works have documented the differences in pavement response from using static and
dynamic modeling. Mamlouk and Davies [1984] concluded that dynamic deflections under FWD tests
were greater than corresponding static deflections under some circumstances due to dynamic
amplification of the pavement structure. Several others such as Chatti et al. [2004] and Al-Khoury et
al. [2001] have come to the same conclusion. Kuo and Tsai [2014] stressed the importanceof subgrade
damping in the dynamic analysis of impact loading.
Analytical models of pavement structures subject to moving loads vary in complexity
with regard to the structure such as plates on elastic foundation (Huang and Thambiratnam
(Chen et al. [2001] and Ding et al. [2013], respectively). Compared to FEM, the main
homogeneity and no inertia effects are usually assumed in order to solve the equations.
For static and quasi-static situations, FE-models have been available since the late
1970’s with the increasing capabilities of computers. Zaghloul and White [1993] were
among some of the first to present 3D models for analysis of flexible pavements subject
to moving loads. The development of models for pavement response to moving load
is still in a relatively early stage. Finite Element modelling of moving loads has been
studied with relation to highspeed railways such as Yang et al. [2003] who studied the
effect of train speeds on wave propagation in layered soils. Later, 2.5D Finite Element
methods have been presented which allows for plane section analysis but with a correct
representation of wave propagation in 3D, see e.g. FranCois et al. [2010] and Galvin
et al. [2010]. However, these methods might not be appropriate in relation to deflection
measurements obtained by an RWD which measures and correlates the deflections inthe longitudinal
direction of the pavement. Thus, analysis of the longitudinal section is
of importance
date 5 – 13 – 2020
for the analysis of the flexible pavement the load act on the it has been as concentrated or distributed
of finite extent, may vary with time and move with constant or variable speed as according to their
require accuracy and tools in their hand.
The The pavement models can be simple beams or plates on Winkler elastic foundations or
Originally, the layered system for modeling the flexible pavement was assumed to exhibit linear
elastic material behavior and be subjected to static, stationary, distributed load. Then, the load was
assumed to be stationary but dynamic (impact load) to simulate the experimental process of
deflectometry. Finally, the load was considered to be constant or time dependent and moving.