IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 767
THREE-DIMENSIONAL FINITE ELEMENT MODELING OF PERVIOUS
CONCRETE PAVEMENT: VERTICAL POROSITY DISTRIBUTION
APPROACH

Ashraful Alam
1
, Liv Haselbach
2
, William F. Cofer
3

1
Teaching Assistant,
2
Associate Professor,
3
Professor, Department of Civil and Environmental Engineering, Washington
State University, USA
ashraf@wsu.edu, haselbach@wsu.edu, wcofer@wsu.edu


Abstract
Pervious concrete has the unique characteristic of allowing water to pass through its porous matrix. . Pervious concrete pavement
systems usually have three main layers, pervious concrete on the top, a subbase layer of aggregate for water storage in the middle and
the subgrade (soil) layer below. Finite element modeling of this novel material is challenging due to its complex porous
characteristics. In this paper, a method has been proposed to model pervious concrete pavement using finite element methods, which
includes a modified approach to capture the unique vertical porosity distribution in the pervious concrete layer by averaging the
distribution in three distinct vertical sections. The mechanical properties of the pervious concrete layer are assumed to vary along its
depth since the porosity of the pervious concrete also varies with depth. ABAQUS, a general-purpose finite element software package
was used to develop the model and perform the analysis. The model was validated through a convergence study, and in comparison
with the analytical theory of tensile stress and deflection for traditional concrete pavement. In addition, the significance of the
porosity distribution model was validated by comparing the results from EverFE a specific software for pavement structure analysis.
It was found that there is significant difference in tensile stress if modeled using the modified vertical porosity distribution in the
previous concrete layer to more fully capture its vertical porosity distribution, as compared to an averaged porosity model in the
previous concrete layer. It was also noted that compressive stress demand may have increased importance for pervious concrete, but
only for highly porous applications which are not commonly used.

Keywords: pervious concrete, vertical porosity, finite element, critical loading, stress, deflection
-----------------------------------------------------------------------------***--------------------------------------------------------------------------------
1. INTRODUCTION
Pervious concrete is a cementitious material that is gaining
popularity and acceptance in the construction industry due to
its contribution to sustainable designs. It is regarded as one of
the stormwater best management practices with low impact
design attributes, and its use has numerous environmental
benefits including improved water quality, better maintenance
of water levels, and improved land utilization [1, 2]. Other
benefits include reduced hydroplaning and glare, and reduced
road noise compared to traditional pavements [3, 4].

Pervious concrete was first reported to be used in the
residential walls in Europe in 1852 [2]. In the U.S., pervious
concrete is generally used in residential streets, driveways and
paths, sidewalks, and parking lots for stormwater management
[5]. It has also been used in low volume highway applications
in Minnesota and in Europe, Australia, and Japan [5, 6]. One
reason for its limited use is the ambiguity in characterizing the
strength and performance of pervious concrete [7]. While
these characteristics can be evaluated by experimental study of
field applications, another approach for evaluating the stress
and deflection demand in pervious concrete pavement is the
application of finite element (FE) methods. The FE method
includes modeling and analysis through discretization of field
problems, and it has advantages over most other numerical
analysis methods because of its physical appeal and versatility
[8]. A finite element model simulates the stress and
deformation behavior of a structure or structural member for
variable geometry, boundary conditions, material properties
and intensity of loading. Thus, a structure can be modeled and
analyzed for preliminary performance evaluation prior to
building or constructing an actual prototype.

Different FE modeling techniques have been considered for
rigid and flexible pavements. For example, while investigating
the response of rigid concrete pavement to static loading and
various environmental conditions, thin interface elements were
used to model the behavior of the top layer of the soil under
the slab, and concrete joints were represented by interface
elements [9]. In another study, the rigid pavement structure
was modeled with multi-layer elastic materials and linear
elastic joint behavior was assumed along the length of the
joint in the connected slabs [10]. For flexible pavements,
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 768
models have included a three layer pavement system (i.e.
asphalt concrete layer, base layer, and subbase layer) with
linear elastic, nonlinear elastic, and viscoelastic materials [11]
and a pavement system with semi-infinite solid material to
obtain guidelines for model size and mesh fineness [12]. To
develop a three-dimensional (3-D) FE based pavement design
procedure for airport rigid pavements, two types of
foundations were considered to represent the infinite depth in
the subgrade layer [13]. One approach for modeling the
subgrade layer was by using a Winkler foundation consisting
of discrete springs at the nodal points beneath the subgrade
layer, while the other approach was to consider a linear elastic
foundation with given modulus of elasticity and Poissons
ratio.

While traditional rigid or flexible pavement has a complex
structure, pervious concrete has additional complexity due to
its porous matrix, which varies with depth, and its unique
properties in the subbase and subgrade. Pervious concrete
pavement systems usually have three main layers, pervious
concrete on the top, a subbase layer of aggregate for water
storage in the middle and the subgrade (soil) layer below,
which usually has less compaction for improved infiltration
than under traditional pavement systems. Microstructures in
the pervious concrete layer have been modeled using a 3-D
reconstruction algorithm to determine the linkages between
microstructure, percolation, and permeability properties [14].
However, modeling and analysis of this novel material using
FE methods to evaluate the structural performance has been
limited and there needs to be further validation for defining a
modeling procedure [15]. The objective of the work of this
paper was to develop a 3-D FE modeling procedure to
evaluate the stresses and displacements in pervious concrete
pavement systems. The modeling procedure included defining
a modified vertical porosity distribution in pervious concrete
[16], from which elastic material properties were obtained
from the investigation of pervious concrete samples from
other studies [17]. The analysis results were then to be
validated through convergence theory for FE formulations,
and also compared with classical analytical theory for
traditional rigid concrete pavement. This convergence theory
states that FE formulations converge toward exact
mathematical solutions with mesh refinement [8]. The
analytical theory for traditional rigid concrete pavement is
based on a plate on an elastic foundation and modeled for
edge, center and corner loading. It states that the maximum
flexural stress will be obtained for wheels positioned at the
edge of the pavement, and that wheels positioned at the corner
of the pavement will provide maximum displacement [18].

2. FE FORMULATION OF PAVEMENT SYSTEMS
The FE formulation methods for modeling pavement systems
are divided into three categories: plane strain, axisymmetric,
and 3-D [19 23]. Plane strain models (Fig-1) use 2-D
formulations and they are often chosen by analysts to model
pavement since they require minimal computer memory and
relatively short computational time [19, 20]. However, plane
strain modeling has the limitation of being unable to simulate
the traffic load accurately [24].



Fig-1: Plane strain model of the pavement

The axisymmetric modeling (Fig-2) approach takes slightly
more computational time compared to the plane strain
modeling approach. This approach has the advantage of
modeling an actual 3-D pavement in cylindrical coordinates,
and constant material properties are assumed in the horizontal
planes. The limitations of this model include improper
simplified representation of traffic loading since only a single
circular load can be modeled, and the inability to include
pavement discontinuities and the shoulder.



(a)

Tire Pressure Tire Pressure
X X
Tire
Location
IJRET: International Journal of Research in Engineering and Technology

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @

(b)

Fig-2: (a) Top view of axisymmetric model of a pavement, (b)
Section X-X of the pavement for FE formulation

A 3-D formulation (Fig-3) is the most computationally
demanding and labor intensive way to model the pavement
system or any other type of structural system. This approach
can represent the pavement configuration accurately and can
include different loading patterns and types, pavement
discontinuity, and the shoulder in the pavement.

In this paper, a 3-D FE formulation has been applied
a pervious concrete pavement system to more accurately
represent the pavement configuration and different loading
patterns for critical wheel positions at the edge, center, and
corner of the pavement.


Fig-3: 3-D model of pavement

Tire Pressure
L
i
n
e

o
f

S
y
m
m
e
t
r
y
Tire Pressure as
Tire Contact Area
International Journal of Research in Engineering and Technology eISSN: 2319
__________________________________________________________________________________________
, Available @ http://www.ijret.org

(a) Top view of axisymmetric model of a pavement, (b)
X of the pavement for FE formulation
is the most computationally
demanding and labor intensive way to model the pavement
pe of structural system. This approach
can represent the pavement configuration accurately and can
include different loading patterns and types, pavement
discontinuity, and the shoulder in the pavement.
D FE formulation has been applied to model
a pervious concrete pavement system to more accurately
represent the pavement configuration and different loading
patterns for critical wheel positions at the edge, center, and

D model of pavement
3. CHARACTERISTICS OF PERVIOUS
CONCRETE
The defining material characteristics of pervious concrete
(Fig-4) are unit weight, tensile and compressive strength,
moduli of elasticity, porosity, infiltration rate, and hydraulic
conductivity. It has been found tha
elasticity, infiltration rate, hydraulic conductivity, and
structural performance of pervious concrete are all functions
of its porosity [5, 25 32]
variables such as compactive effort and water to cement ratio
in the concrete mix design. Although the infiltration rate and
hydraulic conductivity increase with increased porosity
31], unit weight, flexural and comp
moduli of elasticity tend to decrease with increased porosity
[5, 26 29, 32]. Thus, porosity may be used as the governing
characteristic of pervious concrete, from which most other
properties may be defined.


Fig-4: Pervious Conc

In relating the material properties of pervious concrete with its
porosity, it is often assumed that porosity is constant
everywhere in the cross section. However, it has been reported
that the porosity in a pervious concrete layer cross section
varies from top to bottom along its depth with the lowest
porosities on the top due to surface compaction
of expressions have been developed to simulate the porosity in
pervious concrete along its depth for varying levels of
compactive effort. To simplify the analyses, instead of
defining the porosity distribution continuously in terms of
depth, the pervious concrete slab was divided into three
vertical sections i.e. the top quarter, the middle half, and the
bottom quarter, and it was assumed tha
vertical section remains constant. Two sets of expressions
were developed for each of these vertical sections to determine
their porosities: one set uses percent volumetric compaction
while the other set represents the solution for t
range of volumetric compaction between 9 to 10%.

Tire Pressure as
Tire Contact Area
eISSN: 2319-1163 | pISSN: 2321-7308
__________________________________________________________________________________________
769
CHARACTERISTICS OF PERVIOUS
The defining material characteristics of pervious concrete
are unit weight, tensile and compressive strength,
moduli of elasticity, porosity, infiltration rate, and hydraulic
conductivity. It has been found that unit weight, strength,
elasticity, infiltration rate, hydraulic conductivity, and
structural performance of pervious concrete are all functions
32]. The porosity varies with many
variables such as compactive effort and water to cement ratio
in the concrete mix design. Although the infiltration rate and
hydraulic conductivity increase with increased porosity [5, 29,
, unit weight, flexural and compressive strength, and
moduli of elasticity tend to decrease with increased porosity
. Thus, porosity may be used as the governing
characteristic of pervious concrete, from which most other
properties may be defined.

Pervious Concrete
In relating the material properties of pervious concrete with its
porosity, it is often assumed that porosity is constant
everywhere in the cross section. However, it has been reported
that the porosity in a pervious concrete layer cross section
es from top to bottom along its depth with the lowest
porosities on the top due to surface compaction [16]. A series
of expressions have been developed to simulate the porosity in
pervious concrete along its depth for varying levels of
o simplify the analyses, instead of
defining the porosity distribution continuously in terms of
depth, the pervious concrete slab was divided into three
vertical sections i.e. the top quarter, the middle half, and the
bottom quarter, and it was assumed that the porosity in each
vertical section remains constant. Two sets of expressions
were developed for each of these vertical sections to determine
their porosities: one set uses percent volumetric compaction
while the other set represents the solution for the common
range of volumetric compaction between 9 to 10%.
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 770
4. ASSUMPTIONS IN MODELING PERVIOUS
CONCRETE
Pervious concrete pavement systems usually have three or
more different material layers, including the concrete top
layer, an aggregate subbase layer, and the soil subgrade below.
In modeling pervious concrete pavement systems using FE
methods, the pervious concrete layer at the top of the
pavement was represented by three vertical sections through
the depth as defined by the averaged vertical porosity
distribution, and each of those sections were considered to be
composed of linearly elastic material. The subbase
(aggregate/gravel) layer and subgrade (soil) layer were also
considered as linearly elastic. Additional assumptions that
have been considered in modeling the pervious concrete
pavement system are listed below:
The modulus of elasticity of pervious concrete is a
function of the porosity, and the porosity varies as the
top quarter, the middle half, and the bottom quarter
along the depth of the pervious concrete layer.
The Poissons ratio is constant along the depth of the
pervious concrete cross section.
The porosity distribution is constant in the
longitudinal and transverse directions of the
pavement. Thus, the modulus of elasticity and
Poissons ratio are the same in all of the material
directions for each layer considered.
Westergaards assumption that the slab is perfectly
bonded with the subbase as well as that the subbase is
perfectly bonded with the subgrade has been applied
(Huang 2004).

5. FE FORMULATION OF PERVIOUS
PAVEMENT SYSTEMS
The ABAQUS general purpose finite element software
package was used to model and analyze the pervious
pavement systems. In this research, a linear analysis of a
section of the pervious pavement system was performed for
static loading conditions.

5.1 Pervious Pavement Configuration
As previously noted, a pervious pavement system usually
consists of a pervious concrete layer, a gravel subbase layer,
and a natural soil subgrade layer (Fig-5). The subbase layer in
a pervious concrete pavement system serves to enhance the
structural performance of the pavement system in addition to
its use for stormwater storage.


Fig 5: Typical Pervious concrete pavement system

Typically, in a pervious concrete pavement system, the depth
of the subbase layer is comparatively higher than the subbase
depth used for traditional concrete pavement systems.
Although the panels in the pervious pavement system are
commonly jointed together at the bottom one-third along the
depth of the pavement in the longitudinal direction, they are
not commonly jointed together in the transverse direction. In
this research an isolated pervious pavement panel was
modeled which represents a single lane of a highway/street
(not jointed in the transverse direction), but also not jointed in
the longitudinal direction as a simplification. The panel
dimension was 6.0 m in the direction of traffic and 3.66 m in
the transverse direction, and it had a total depth of 200 mm
divided into three layers: the top quarter (50 mm), the middle
half (100 mm), and the bottom quarter (50 mm). The subbase
layer had a depth of 250 mm and it was extended 300 mm
beyond the outside edge of the pavement to mimic an actual
pavement configuration (Fig-6).



(a)

Pervious Concrete
Subbase
Subgrade
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 771


(b)

Fig-6: (a) Layer by layer representation of pervious pavement
system for FE formulation, (b) 3-D view of the pavement
model from ADINA

5.2 Material Properties
The essential material properties for linear elastic materials
within the FE method are the modulus of elasticity and
Poissons ratio. As stated in the assumption, the modulus of
elasticity for pervious concrete may be expressed as a function
of porosity where the porosity varies along the depth of the
pervious concrete layer cross section [16]. The relationship
between the modulus of elasticity and the porosity was
obtained from previous experimental investigations of samples
prepared in the laboratory [17]. Subbase and subgrade material
properties were obtained from previous studies [33 38].

5.2.1 Pervious Concrete
In the previous study, pervious concrete samples were
extracted from a field placement and specimens were prepared
in the laboratory to evaluate their structural performance [17].
Unit weight, compressive strength, porosity and exfiltration
rate were determined for the core extracted samples while, for
laboratory prepared samples, unit weight, compressive
strength, modulus of elasticity, Poissons ratio, porosity and
exfiltration rate were determined. For the pervious pavement
field placement, undisclosed admixtures were used to
strengthen the pervious concrete. For this reason, in the FE
formulation, the moduli of elasticity of the laboratory prepared
samples were used.

The relationships between the compressive strength and the
porosity and between the modulus of elasticity and the
compressive strength for the laboratory prepared samples may
be expressed by Eqs. (1, 2), respectively [17]:

'
43.8 114
c
f P =
(1)

6 1.5 '
50.6 10
pc c
E w f

=
(2)

Where
'
c
f
is the compressive strength in MPa,
P
is the
porosity in percent,
E
is the modulus of elasticity in GPa, and
pc
w
is the unit weight of pervious concrete in kg/m3.

Eqs. (3, 4, 5), respectively [16], express the porosity in the top
quarter, middle half, and bottom quarter of a pervious concrete
layer. It should be noted that the percent compaction values
used in deriving these porosity distribution equations were
between 9 and 10%.

1.07 7
top mean
P P =
(3)

mid mean
P P =
(4)

0.93 7
bottom mean
P P = +
(5)

Where
mean
P
is the average porosity of the pervious concrete
layer in percent and
top
P
,
mid
P
, and
bottom
P
are the percent
porosity values at the top quarter, middle half, and bottom
quarter, respectively.

Thus, if the average porosity of a pervious concrete layer is
known, then the corresponding porosity at the top quarter,
middle half, and bottom quarter of that layer can be
determined through Equations 3, 4, and 5. For these porosity
values, compressive strength and the corresponding moduli of
elasticity in the three vertical sections of the pervious concrete
layer can be evaluated from Equations 1 and 2 for a known
unit weight of pervious concrete.

The average unit weight of the pervious concrete samples
considered in the previous laboratory experiments was 1858
kg/m3 and the mean porosity was 27% [17]. The unit weight
and porosity of pervious concrete have also been reported to
be between 1570 kg/m3 and 1938 kg/m3 and between 10% to
29%, respectively [25]. To determine the modulus of elasticity
for the FE formulation, the unit weight and average porosity
considered were 1850 kg/m3 and 25%, respectively. In the FE
formulation of pervious concrete, the Poissons ratio was kept
constant in all three sections through the depth. The
mechanical properties of pervious concrete that were
considered in the FE formulation are listed in Table-1.







IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 772
Table-1: Mechanical properties of pervious concrete for FE
formulation

Unit Weight 1850 kg/m3
Mean Porosity 25%
Modulus of Elasticity Top Quarter 18.55 GPa
Middle Half 15.70 GPa
Bottom Quarter 12.97 GPa
Poissons ratio 0.22

5.2.2 Subbase and Subgrade
The modulus of elasticity of the subbase, typically composed
of unbounded aggregate, depends on the type of aggregate
[35] and, for the subgrade or soil; the dominating factor
affecting the modulus of elasticity is the effective stress,
which is a function of the water level in the soil layer [33]. In
analyzing pavement, the modulus of elasticity of the subbase
layer has been reported to range from 96 MPa to 500 MPa,
and for the subgrade layer, the modulus of elasticity has varied
between 24 MPa and 100 MPa [34, 36 38]. While the
Poissons ratio of the subbase was reported to have a constant
value of 0.35 for all analyses [34, 36 38], the Poissons ratio
of the subgrade was reported to vary from 0.35 ~ 0.45. The
properties of the subbase and subgrade materials considered in
the FE analysis of pervious concrete pavement are given in
Table-2. Average values were chosen for the aggregate
subbase and conservatively low values for clayey soils were
chosen for the subgrade.

Table-2: Mechanical properties of subbase and subgrade

Subbase Subgrade
Unit Weight (kg/m3) 1800 1800
Modulus of Elasticity (MPa) 400 50
Poissons ratio 0.35 0.40

5.3 Loading on the Pavement
The applied tire pressure on the pavement represents an
equivalent 160-kN tandem axle load and the tire type
corresponds to dual wheel tires (Fig-7a). Tire pressure with an
equivalent 160-kN single axle load is a widely accepted
approach to represent different axle load groups and to
calculate load repetitions in pavement [18]. A standard
semitrailer wheel configuration (Fig-7b) for tandem axle dual
wheel tires was used to represent the tire pressure on the
pavement.


(a)



(b)

Fig-7: a) Tandem axle dual wheel tire, b) Wheel configuration

Whether from closed-form formulas, influence charts, or finite
element methods, it was found that, for traditional concrete
pavement, the critical location of the wheel for flexure is at the
edge of the pavement (Fig-8a) and the critical location of the
wheel for deflection is at the corner of the pavement (Fig-8b).
The pervious concrete pavement in this study was analyzed for
these wheel positions to validate the FE model. In addition,
wheel placement at the center of the pavement was considered
to identify any unique mechanistic response (i.e. tensile stress,
compressive stress, or deflection) of pervious concrete that
might occur for this loading condition.



(a)

229
1
5
8
1219
1
8
2
9
3
4
3
Direction of Traffic
Transverse Joint
Tandem axle load
Concrete Shoulder
(if used)
Traffic lane
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 773


(b)



(c)

Fig-8: Wheel position at a) Edge, b) Corner, and c) Center of
the pavement

6. FE ANALYSIS
3-D models of the pervious concrete pavement representing
the coarsest and finest meshes used are shown in Fig-9 for
wheels positioned at the edge of the pavement. The meshing
pattern, which includes finer meshes in and around the tire
location and coarser meshes away from the tire contact area,
was selected for a uniform distribution of tire pressure in the
pavement around the tire contact area and to reduce
computational time.



(a)



(b)

Fig-9: Pervious pavement system with (a) Coarsest meshing
and (b) Finest meshing

In the ABAQUS element library, there are a number of solid
or continuum elements that can be used for linear 3-D FE
analysis. The solid element library includes both first order
(linear interpolation) and second order (quadratic
interpolation) elements. Second order elements require higher
memory storage and computation time, thus first order linear
interpolation brick elements (8-node) were used in this study.
In addition, the elements were enhanced by using
incompatible modes to improve their bending behavior since
for rectangular shape elements the incompatible mode
elements perform similar to the second-order elements. The
incompatible mode adds additional internal degrees of
freedom and is significantly more economical than the second-
order elements.

The FE modeling of the subgrade in the pavement system is
important due to its infiniteness and it is also important to
apply appropriate boundary condition. A dense liquid
foundation or an elastic solid foundation are usually used to
model the subgrade. However, a dense liquid foundation
hinders the use of contact or friction between the layers [39]
and thus, was not chosen for the analysis. To model the
subgrade as an elastic solid foundation, the layers of the brick
element should have to be extended to a depth at which the
mechanistic response of the pavement system is negligible
compared to the previous depth. A parametric study was
performed to determine the depth for an elastic solid
foundation and it was found that at a depth of 2.7 m the
mechanistic response of the pavement system is almost
identical to the previous analyzed depth at 2.4 m. To mimic
the infiniteness of the subgrade layer in the horizontal
direction, it was extended to 2.7 m in longitudinal and
transverse direction of the pavement system.


Transverse Joint
Tandem axle load
Concrete Shoulder
(if used)
Traffic lane
Transverse Joint
Concrete Shoulder
(if used)
Traffic lane
Tandem axle load
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 774
6.1 Convergence Study for the Pervious Pavement
Model
A convergence study was performed for the case of the wheel
position at the corner of the pavement system. The maximum
tensile stress, compressive stress, and deflection values are
plotted against the number of elements in Fig 10. The plot
shows that the stresses and deflection in the pavement system
increase with an increase in the number of elements in the
model. While the deflection and compressive stress converged
for the same number of elements in the pavement system,
tensile stress required much finer mesh to reach the
convergence.



(a)



(b)



(c)

Fig 10: Convergence of a) Deflection, b) Tensile stress, and
c) Compressive stress for different mesh size in the pavement
system
6.2 Evaluation of Stress and Deflection
To evaluate the stress and deflection in the pervious concrete
pavement system, the loading conditions considered were
wheels at the edge, corner, and center of the pavement. The
finest meshing criterion from the convergence study was
adopted to mesh the pavement system for edge and center
loading on the slab. Bar charts have been used in the previous
studies to compare stresses and deflection in the pavement
[15] and the maximum tensile stress, compressive stress, and
deflection obtained for the three loading conditions have been
similarly plotted (Fig-11) to determine the critical loading
conditions for tensile and compressive stress, and for
deflection in the pervious concrete pavement. From Fig 11, the
wheel position at the corner of the pavement is shown to yield
the maximum deflection as well as the maximum compressive
stress in the pavement system. The maximum tensile stress in
the pavement system was obtained for edge loading. Thus, the
critical loading conditions for pervious concrete pavement
system is the same as the analytical theory of traditional
concrete pavement with maximum tensile stress for edge
loading and with maximum deflection for corner loading.



(a)



(b)
2.54
2.55
2.56
2.57
2.58
2.59
20 40 60 80 100 120 140
D
e
f
l
e
c
t
i
o
n

(
m
m
)
Number of Elements (Thousand)
1.260
1.265
1.270
1.275
1.280
1.285
1.290
1.295
20 40 60 80 100 120 140 160
T
e
n
s
i
l
e

S
t
r
e
s
s

(
M
P
a
)
Number of Elements (Thousand)
4.0
5.0
6.0
7.0
8.0
9.0
20 70 120 170
C
o
m
p
r
e
s
s
i
v
e

S
t
r
e
s
s

(
M
P
a
)
Nimber of Elements (Thousand)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Edge Corner Center
D
e
f
l
c
e
t
i
o
n

(
m
m
)
Wheel Position in the Pavement
System
1.20
1.25
1.30
1.35
1.40
1.45
1.50
Edge Corner Center
T
e
n
s
i
l
e

S
t
r
e
s
s

(
M
P
a
)
Wheel Position in the Pavement
System
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 775


(c)

Fig 11: Comparison of maximum a) Deflection, b) Tensile
stress, and c) Compressive stress for different loading
condition in the pavement system

While tensile stress is assumed to be responsible for the
initiation of cracks in traditional concrete pavements [40],
compressive stress in the pavement is typically not monitored
in the published literature [18, 40]. The relatively high
compressive stress in the pervious concrete reflects its porous
matrix and it occurs for the wheel positioned at the corner of
the pavement. The higher compressive stress for corner
loading could be attributed to the confinement of tire pressure
at the corner of the pavement due to corner position of wheels
whereas for edge loading, the tire pressure can be distributed
on both sides of the wheel position. However, the compressive
stresses found are still much lower than the typical
compressive strengths of most pervious concrete mixes.

6.3 Evaluation of Stresses for a Rigid Pavement
Model using EverFE
The significance of using the vertical porosity distribution
approach to capture the unique porosity distribution in the
pervious concrete pavement system by dividing the surface
into three layers with three different moduli of elasticity has
been validated here by analyzing the pervious concrete
pavement system using a conventional pavement analysis
software named EverFE with only one modulus of elasticity
value in the pervious concrete layer [40]. EverFE is a 3-D FE
analysis tool to simulate the mechanistic response of pavement
system for different axle loads and environmental effects. In
EverFE, four layers can be defined with only one layer in the
surface. In the EverFE analysis, the modulus of elasticity and
Poissons ratio of the surface layer was given the same values
as the middle layer in the vertical porosity distribution
approach, representing a simpler average modulus of elasticity
approach.

The critical stresses and deflections from this more simplified
approach are listed in Table-3, in addition to the vertical
porosity distribution model previously discussed with varying
characteristics in the pervious concrete layer. In Table-3, the
maximum deflection and compressive stress were for corner
loading, and the maximum tensile stress was for edge loading
in the simpler model, as was also found previously in the
modified vertical porosity distribution model. It can be seen
that the simpler average modulus of elasticity approach gives
much higher tensile stress compared to the modeling approach
which captures the vertical porosity distribution, which
eventually will lead to a more conservative design. Apart from
the average modulus of elasticity and three vertical section
approach, the over prediction of tensile stress could be due to
the incapability of EverFE to consider a widened subgrade
layer in the longitudinal and transverse direction of the
pavement system.

Table-3: Comparison of the modified vertical porosity versus
average porosity in the pervious concrete layer modeling
approaches

Modified
Vertical
Porosity
Approach
EverFE
Analysis
%
Difference
Deflection (mm) 2.59 -- --
Tensile Stress
(MPa)
1.47 2.28 55.1
Compressive
Stress (MPa)
8.29 2.48 233.3

CONCLUSIONS
A 3-D FE model of pervious concrete pavement was analyzed
incorporating a modified vertical porosity distribution in the
pervious concrete to simulate the effect of its porous matrix.
The validation criteria were convergence theory and the
analytical theory of stresses and deflection of traditional
concrete pavement. The model was also compared with a
simplified average porosity modeling approach in a traditional
pavement analysis software. The pervious pavement model
with the modified vertical porosity distribution showed
consistent convergence with mesh refinement. In comparison
with the analytical theory for traditional concrete pavement, it
was found that the behavior of pervious concrete pavement
dictates that the compressive stress demand in pervious
concrete pavement may be critical under conditions with very
high porosities, but these conditions are typically not found in
current practice. While tensile stress demand is higher for
edge loading compared to, corner loading, compressive stress
demand is clearly highest for corner loading. A significant
finding is that the three vertical section modeling approach
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
Edge Corner Center
C
o
m
p
r
e
s
s
i
v
e

S
t
r
e
s
s

(
M
P
a
)
Wheel Position in the Pavement
System
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 776
approximating the vertical porosity distribution in pervious
concrete has significantly less critical tensile stress and higher
compressive stress compared to the simplified average
pervious concrete layer modeling approach.

Limitations of the proposed FE modeling technique include
the assumption of perfect bonding between the pervious
concrete and subbase, and subbase and subgrade. Further
mesh refinement is also needed for the true convergence of
deflections and stresses. The modeling procedure presented
here is for macro-scale modeling of pervious concrete
pavement systems, and does not consider the effect of random
zero porosity phase inside the pervious concrete. In the case of
pervious concrete, there may be potential stress concentration
at a paste-aggregate or a paste-aggregate-pore interface. Such
modeling requires micro-scale modeling and was not
considered in the full-scale modeling of a pervious concrete
pavement systems presented here.

Future studies may include other types of tire configurations
(i.e. single axle and tridem axle), cyclic loading, consideration
of nonlinear elements in the pavement system, and inclusion
of the friction coefficient between the pervious concrete and
subbase, and subbase and subgrade. To imitate the more
random porous matrix of the pervious concrete layer, meso-
scale modeling could be used. Despite these limitations, it can
be concluded that this newly developed modified vertical
porosity distribution approach for simulating the pervious
concrete layer in a pervious concrete system will facilitate
pervious concrete research using FE formulation to study the
numerous structural and environment effects on this novel
material. This might be the basis for future dynamic and
nonlinear analysis of pervious concrete pavement systems.

ACKNOWLEDGEMENTS
The authors gratefully acknowledge the financial support from
Transportation Northwest. The authors also appreciate the
assistance from Evolution Paving and Will Goede in
evaluating the performance of pervious concrete.

REFERENCES
[1]. Tennis P, Leming M, Akers D (2004) Pervious
Concrete Pavements. Portland Cement Association,
Skokie, Illinois
[2]. ACI Committee 522 (2006) ACI 522R-06: Pervious
concrete. American Concrete Institute
[3]. Wanielista M, Chopra M (2007) Performance
assessment of Portland Cement pervious pavements.
Final Report FDOT Project BD521-02. Available from:
http://www.dot.state.fl.us/research-
center/Completed_Proj/Summary_RD/FDOT_BD521_
02_rpt2.pdf. [Accessed February 4, 2011]
[4]. Hendricks IL (1998) Noiseless concrete pavements.
Available from:
http://www.eupave.eu/documents/graphics/inventory-
of-documents/febelcem-publicaties/noiseless-concrete-
pavements.pdf [Accessed 22 August 2011]
[5]. Kevern JT, Schaefer VR, Wang K (2011) Mixture
proportion development and performance evaluation of
pervious concrete for overlay application. ACI
Materials J 108(4): 439-448
[6]. Rohne RJ, Izevbekhai BI (2009) Early performance of
pervious concrete pavement. TRB Annual Conference
[7]. Goede W, Haselbach L (2011) Investigation into the
Structural Performance of Pervious Concrete. J of
Transportation Engineering,
doi:10.1061/(ASCE)TE.1943-5436.0000305.
[8]. Cook RD, Malkus DS, Plesha ME, Witt RJ (2002)
Concepts and application of finite element analysis. 4
th

Ed. John Wiley & Sons, Inc., New Jersey, USA
[9]. Beegle DJ (1998) Three dimensional modeling of rigid
pavements. MS Thesis, Ohio University
[10]. Brill DR, Hayhoe GF, Lee, X (1997) Three-
dimensional finite element modeling of rigid pavement
structures. Aircraft/Pavement Technology in the Midst
of Change. Proceedings of the 1997 Airfield Pavement
Conference, Seattle, Washington, August, ASCE,
Editor Frank V. Herman, 151-165
[11]. Helwany S, Dyer J, Leidy J (1998) Finite element
analysis of flexible pavements. J. of Transportation
Engineering,124(5): 491-499
[12]. Kuo C, Chou F (2004) Development of 3-D finite
element model for flexible pavements. J of Chinese
Institute of Engineers, 27(5): 707-717
[13]. Brill DR, Parsons ID (2001) Three-dimensional finite
element analysis in airport pavement design. The
International J of Geomechanics 1(3): 273-290
[14]. Bentz DP (2008) Virtual Pervious Concrete:
Microstructure, Percolation, and Permeability. ACI
Materials J 105(3): 297-301
[15]. Alam A, Haselbach L, Cofer W (2012) Validation of
the Performance of Pervious Concrete in a Field
Application with Finite Element Analysis (FEA). J of
ASTM International 9(4)
[16]. Haselbach LM, Freeman RM (2006) Vertical Porosity
Distributions in Pervious Concrete Pavement. ACI
Materials Journal 103(6): 452-458
[17]. Goede, WG (2009) Pervious Concrete: investigation
into structural performance and evaluation of the
applicability of existing thickness design methods. MS
Thesis, Washington State University
[18]. Huang YH (2004) Pavement Analysis and Design.
Prentice Hall, Upper Saddle River, New Jersey
[19]. Van Metzinger WA, McCullough BF (1991) An
Empirical-Mechanistic Design Method Using Bonded
Concrete Overlays for the Rehabilitation of Pavements.
Report 1205-1. Center for Transportation Research,
University of Texas at Austin, Jan
[20]. Lytton RL, Uzan J, Fernando EG, Roque R, Hiltunen
D, Stoffels SM (1993) Development and Validation of
Performance Prediction Models and Specifications for
IJRET: International Journal of Research in Engineering and Technology eISSN: 2319-1163 | pISSN: 2321-7308

__________________________________________________________________________________________
Volume: 02 Issue: 12 | Dec-2013, Available @ http://www.ijret.org 777
Asphalt Binders and Paving Mixes. Report SHRP-A-
357 Strategic Highway Research Program, National
Research Council, Washington, D.C.
[21]. Nam D (1994) Effect on Slab Flexibility on the Vertical
Stiffness of Circular Foundations. MS thesis, University
of Texas at Austin, Austin, TX
[22]. Zaghloul SM, White TD, Drnevich VP, Coree B (1994)
Dynamic Analysis of FWD Loading and Pavement
Response Using a Three-Dimensional Finite Element
Program. Nonrestrictive Testing of Pavements and
Backcalculation of Moduli, 2: 1198, ASTM
[23]. Thompson MR (1998) ILLI-PAVE, Users Manual.
Transportation Facilities Group, Department of Civil
Engineering, University of Illinois, Urbana
[24]. Cho YH, McCullough BF, Weissmann J (1996)
Considerations on finite-element method application in
pavement structural analysis. Transportation Research
Record 1539: 96-101
[25]. Ghafoori N, Dutta S (1995) Pavement thickness design
for no-fines concrete parking lots. J of Transportation
Engineering 121( 6): 476-484
[26]. Crouch LK, Pitt J, Hewitt R (2007) Aggregate Effects
on Pervious Portland Cement Concrete Static Modulus
of Elasticity. J of Materials in Civil Engineering 19(7):
561-568
[27]. Chindaprasirt P, Hatanaka S, Mishima N, Yuasa Y,
Chareerat T (2009) Effects of binder strength and
aggregate size on the compressive strength and void
ratio of porous concrete. International J of Minerals,
Metallurgy and Materials 16(6): 714-719
[28]. Delatte N, Mrkajic A, Miller, DI (2009) Field and
Laboratory Evaluation of Pervious Concrete
Pavements. Transportation Research Record: Journal of
the Transportation Research Board, Transportation
Research Board of the National Academies,
Washington, D.C., 2113: 132139
[29]. Sumanasooriya MS, Neithalath N (2009). Stereology-
and Morphology-Based Pore Structure Descriptors of
Enhanced Porosity (Pervious) Concretes. ACI Materials
J 106(5): 429-438
[30]. Deo, O, Neithalath, N (2010) Compressive behavior of
pervious concretes and a quantification of the influence
of random pore structures features. Material Science
and Engineering A 528: 402-412
[31]. Neithalath N, Bentz DP, Sumanasooriya MS (2010)
Predicting the Permeability of Pervious Concrete
Advances in characterization of pore structure and
transport properties. Concrete International, 35-40
[32]. Henderson V, Tighe S, Norris J (2011) Pervious
concrete pavement: an integrated laboratory and field
study. TRB Annual Conference, CD-ROM
[33]. Coduto DP (1994) Foundation Design Principles and
Practices. Prentice-Hall, Inc., Englewood Cliffs, NJ
07632, USA
[34]. Hjelmstad KD, Zuo Q, Kim J (1997). Elastic Pavement
Analysis Using Infinite Elements. Transportation
Research Record: Journal of the Transportation
Research Board, Transportation Research Board of the
National Academies, Washington, D.C., 1568: 72-76
[35]. Somayaji S (2001) Civil Engineering Materials.
Prentice-Hall, Inc., Upper Saddle River, NJ 07458,
USA
[36]. Hadi MNS, Bodhinayake BC (2003) Non-linear finite
element analysis of flexible pavement. Advances in
Engineering Software 34:657-662
[37]. Saleh MF, Steven B, Alabaster D (2003) Three-
dimensional nonlinear finite element model for
simulating pavement response. Transportation Research
Record: Journal of the Transportation Research Board,
Transportation Research Board of the National
Academies, Washington, D.C., 1823: 132139
[38]. Nishizawa T, Ozeki T, Katoh K, Matsui K (2009) Finite
element model analysis of thermal stresses of thick
airport concrete pavement slabs. Transportation
Research Record: Journal of the Transportation
Research Board, Transportation Research Board of the
National Academies, Washington, D.C., 312
[39]. Kuo C, Hall KT, Darter MI (1995) Three-dimensional
finite element model for analysis of concrete pavement
support. Transportation Research Record: Journal of the
Transportation Research Board, Transportation
Research Board of the National Academies,
Washington, D.C., 1505: 119127
[40]. Croney D, Croney O (1997) Design and performance of
road pavements. McGraw-Hill, New York
[41]. Davids WG, Wang Z, Turkiyyah G, Mahoney JP, Bush,
D (2003) Three-Dimensional Finite Element Analysis
of Jointed Plain Concrete Pavement with EverFE2.2.
Transportation Research Record: Journal of the
Transportation Research Board, Transportation
Research Board of the National Academies,
Washington, D.C., 1853: 9299