Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152

2012 ATS All rights reserved

139
Modeling the Effect of Rolling Conditions on Stress Development
at Tire-Pavement Contact Patch

J iasheng YANG
a
, Ghim Ping ONG
b
, Tien Fang FWA
c
, Chye Heng CHEW
d


a,b,c
Department of Civil Engineering, National University of Singapore, 1 Engineering Drive
2, Singapore 117576, Singapore
a
E-mail: yangjiasheng@nus.edu.sg
b
E-mail: cveongr@nus.edu.sg
c
E-mail: cvefwatf@nus.edu.sg
d
Department of Civil Engineering, National University of Singapore, 9 Engineering Drive 1,
Singapore 117576, Singapore; E-mail: mpecch@nus.edu.sg


Abstract: The interfacial pressure between tire and pavement is known to exhibit a highly
non-uniform distribution over its contact area. This differs significantly from the uniform
contact pressure distribution used traditionally in pavement design and analysis. Past research
efforts have recognized that the contact stress distributions are dependent on tire type, tire
inflation pressure, wheel load, tire tread patterns and vehicle rolling characteristics. This paper
discusses the development of a numerical model that is capable of describing the contact
stress distributions at the interface between a rolling tire and the pavement. The theoretical
development of the finite element-based simulation model is first described. This model is
then validated against the experiment data available in the literature and is applied to simulate
the stress distributions at the tire-pavement contact area under different rolling conditions.

Key Words: Tire-pavement interaction, Rolling tire, Contact stress distribution, Finite element
method.


1. INTRODUCTION

A reliable prediction of vertical, longitudinal and lateral stress distribution on tire pavement
contact area is of great importance to pavement design and evaluation. For the past decades,
conventional pavement analytical approaches have always assumed that contact vertical stress
is equal to tire inflation pressure uniformly distributed on the contact surface between the tire
and the pavement (Yoder, 1959). To date, this assumption is still prevalently used in many
empirical pavement design procedures. In addition, it is frequently assumed in pavement
design analyses that the tire-load is stationary (i.e., it is a static loading case) (Huang, 1993;
AASHTO, 1996).

It was recognized that the assumptions of uniform contact stress distribution and stationary
tire are inappropriate for pavement design and pavement performance prediction. These
assumptions ignore the fact that localized tire-pavement contact stresses are key factors in the
development of pavement ruts and in the initiation and propagation mechanisms of cracking
in pavements. Experimentally, Lippmann (1985) noted that the vertical contact stress
distribution on the tire-pavement contact interfaces are non-uniform and this finding was
further verified by De Beer et al. (1997). However, these studies were based on the static
loading conditions and could not adequately describe the real traffic conditions where
vehicles are moving (i.e. tires are rolling). Compared to the numerous experimental studies on
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
140
static tire-pavement contact, relatively fewer experiments were conducted for the case of a
rolling tire loading on a pavement surface (De Beer, 1994; De Beer et al., 1997; Douglas et
al., 2000; Douglas, 2009). De Beer (1994) investigated experimentally the contact stress
distributions developed when a smooth Goodyear truck tire is rolling on a pavement surface at
a speed of 8 km/h under different wheel loads. It was found that the behavior of the contact
stress distributions of the rolling tire may differ significantly to that of the static tire, and
speeds have little effect on the contact stress distribution of rolling tires. De Beer et al. (1997)
further surmised that vertical stress profiles for pneumatic tires can exhibit either an n-shape
or and m-shape (as illustrated in Figure 1), depending on the type of tire (passenger car or
truck), loading conditions and pavement material type and surface properties.


Figure 1. Basic 3D contact stresses and basic shapes on the rolling tire from experimental
observations (De Beer et al., 1997)

Experimental efforts to study tire-pavement contact stresses often consume a lot of time and
are very costly. This leads to the prevalent use of numerical modeling to model tire-pavement
interaction. Many methods were proposed to numerically model tire-pavement interaction
stresses (Lippmann, 1985; Howell et al., 1985; Blab, 1999; Park, 2005). Most were, however,
modeled to either analyze tire wear which is of interest to tire manufacturers or to ascertain
stresses at the tire-pavement contact for use in pavement design. The former tends to ignore
the effect of pavement material properties and surface characteristics; while the latter tends to
over-simplify the tire model.

In order to properly investigate the dynamic behavior of tire-pavement interaction under
different rolling conditions, this paper discusses the development of a rolling smooth tire and
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
141
smooth pavement interaction numerical model, which is capable of describing the contact and
traction stress distribution at the tire-pavement interface. The theoretical development of the
finite element-based simulation model is first described. This model is then validated against
the experiment data available in the literature. Finally, the model is applied to simulate the
contact patch stress distributions under different rolling conditions.


2. DEVELOPMENT OF TIRE-PAVEMENT INTERACTION SIMULATION MODEL

The modeling of rolling tire and pavement interaction is a very complex process. In this
paper, a finite element simulation model was developed to model the contact interaction. The
simulation model consists of two different sub-models, namely the tire model and the
pavement model. These two models interact with one another through the contact modeling
mechanism. These sub-models allow the determination of the tire-pavement contact patch and
the stresses developed at the tire-pavement interface, as shown in Figure 2. The software
package ADINA is used to develop the simulation model. The following sub-sections describe
in detail the key processes involved in the simulation model development.



Figure 2. Key components of the simulation model

2.1 Numerical Modeling of Rolling Tire Sub-Model

The modeling of rolling tire is perhaps the most important component of the tire-pavement
interaction modeling and has a direct influence on the accuracy of contact stress distribution.
In this paper, the PIARC smooth tire is used. The main dimensions of PIARC smooth tire is
shown in Figure 3. The initial tire properties are obtained from PIARC tire specification
(PIARC, 2004). The rolling tire is modeled using single-layer shell iso-parametric elements
of 4-node quadrilateral shape, which has been successfully used to simulate the rolling and
sliding tires (Chang et al., 1983; Ong and Fwa, 2007).








TRUCK TIRE
MODEL

INPUT:
Tire Geometry
Material Properties
Tire Inflation Pressure
Wheel Load
Vehicle Speed and Slip
PAVEMENT
MODEL

INPUT:
Pavement Surface Pattern
Coefficient of Friction
Pavement Material
Properties
Pavement Structure
Tire-Pavement
Contact Modeling
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
142




Figure 3. Dimensions of the PIARC smooth tire (PIARC, 2004)

The element stress-strain matrix in local axes system (as illustrated in Figure 4), with
orthogonal material axes a, b and c can be derived as:
[ ]
1/ / / 0 0 0
/ 1/ / 0 0 0
/ / 1/ 0 0 0
0 0 0 1/ 0 0
0 0 0 0 1/ 0
0 0 0 0 0 1/
a ab b ac c
ba b b bc c
ca a ca a c
ab
ac
bc
E v E v E
v E E v E
v E v E E
C
G
G
G





=






(1)
where E
i
is Youngs modulus, G
ij
is shear modulus and is Poissons ratio. The symmetric
relationship between Youngs modulus and Poissons ratio is:
, , , ,
ji ij
i j
v v
i j a b c
E E
= = (2)
In order to calculate the inverse [C], the following constraints are necessary so that the stress-
strain matrix is positive-definite:
1
2
, , , ,
i
ji
j
E
v i j a b c
E

= =



(3)

2 2 2
0.5(1 ) 0.5
a a b a
ab bc ac ab bc ac
c b c c
E E E E
v v v v v v
E E E E
< (4)

T
i
r
e

h
e
i
g
h
t

=

1
3
2
.
5

m
m

O
v
e
r
a
l
l

d
i
a

=

6
4
6

m
m
Rim width =114 mm
R
i
m

d
i
a

=

3
8
1

m
m
Tire width =167 mm
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
143

Figure 4. Definition of local axes system for shell elements

The advancing front algorithm (Chae and Bathe, 1989) is used for mesh generation. Mesh
convergence analysis is performed under the condition of tire pressure 200 kPa and 4980 N.
The results are shown in Figure 5. It is shown clearly from the figure that for the tire model,
approximately 12,800 shell elements are sufficient to give relatively accurate results.

0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
x 10
4
1.45
1.5
1.55
1.6
1.65
1.7
1.75
x 10
4
Mesh number of tire model
C
o
n
t
a
c
t

a
r
e
a

(
m
m
2
)
Setting:
Smooth Pavement
P.I.A.R.C. Slick Tire
Tire Load: 4980 N
Tire Infaltion Pressure: 200 kpa

Figure 5. Mesh convergence analysis for tire model

For the modeling of the pneumatic tire, three structural components, namely tire rim, tire
sidewalls and tire tread, are considered. The tire rim is considered to be a rigid body. The tire
sidewalls and treads are assumed to be of an orthotropic elastic material with composite
elastic properties. Their structural properties are characterized by the parameters: elastic
modulus, shear modulus and Poissons ratio. Figure 6 shows meshes of these element groups
and the tire pavement interaction model with the selected mesh division is shown in Figure 7.

Parameters:
Smooth rigid pavement
PIARC smooth tire
Tire load =4980 N
Tire inflation pressure =200 kPa
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
144

Sidewall Tread Rim

Figure 6. Meshes of tire element groups



Figure 7. Rolling tire-pavement interaction model for PIARC smooth tire

2.2 Pavement Surface Modeling

A typical pavement system is generally composed of three layers: asphalt concrete, base and
sub-grade. The material in each layer can be defined to be isotropic linear elastic (Myers,
2000; Drakos, 2003). In this paper, the pavement surface is assumed to be perfectly rigid to
simplify the modeling process.

2.3 Tire-Pavement Contact Modeling

Pavement surface interacts with rolling tire through the contact surface. Contact behavior has
a great effect on tire dynamic performance. It therefore is very important to accurately model
the tire-pavement contact interaction. This interaction can be classified as a three-dimensional
(3-D) contact problem. To make the modeling of complex contact problems feasible, the
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
145
following general contact conditions must be satisfied (ADINA, 2009):
1) The points of contact surface are not known a-priori.
2) Friction can be modeled using various friction laws.
3) Self-contact and contact between the contactor and contacted surface are permitted.
4) Both sticking and sliding can be modeled.

The relationship between contact dynamics and friction (Bathe, 2009) can be expressed as:
1 and 1 < implies the slip velocity: 0
SX
V = (5a)
while 1 = implies: sign( ) sign( )
SX
V = (5b)
in which is a non-dimensional friction variable defined as:
T
F

=

(6)

where
T
F is tangential force, is normal contact force and is friction coefficient.

The constraint function method is used as the contact algorithm for this contact problem
(Bathe, 1985). The pavement surface is assumed to be target surface and tire is assumed to be
contactor surface. The inherent friction (i.e. skid number at zero speed SN
0
) between smooth
tire and pavement is assumed to be 20.

2.4 Calibration and Validation of Simulation Model

The PIARC smooth tire model is calibrated under a uniform inflation pressure 200 kPa. The
choice of tire tread and sidewall elastic properties requires a careful calibration to ensure that
the static footprint obtained from the simulation is the same as the actual footprint of a
stationary tire on a dry pavement under the same load. Six different loads (ranging from 2,920
N to 4,980 N) are used for the calibration process. The calibrated tire material parameters are
shown in Table 1. It can be observed that the use of the calibrated elastic and shear moduli of
tire tread and its Poissons ratio in the simulation model satisfies the specification
requirements for PIARC smooth tire (i.e. a tire tread elastic modulus of 5.8 1.4 MPa)
(PIARC, 2004).

Table 1. Calibrated parameters for tire and pavement models
(a) Elastic properties for tire rim and pavement
Material Type Elastic modulus (MPa) Poisson's Ratio Density (kg/m
3
)
Steel Rim 2.1E+5 0.3 7,800
Pavement 3.0E+6 0.15 2,600
(b) Orthotropic elastic properties for tire tread and sidewalls
Elastic Modulus (MPa) Poisson's Ratio Shear Modulus (MPa) Density
(
3
Kg/m )
a b c ab ac bc ab ac bc
Sidewall 1.3 6.0 6.0E+06 0.4 0.4 0.4 0.50 0.50 0.50 1,452
Tread 5.8 3.0 4.0E+06 0.2 0.1 0.4 0.10 0.96 0.10 1,143

Table 2 compares the area of simulated and experiment contact footprints for the six different
static wheel loads considered in the study. From the table, it can be observed that by using the
calibrated parameters, the error in the static footprint is less than 5%. This indicates that the
choice of the various calibrated parameters is appropriate.

YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
146
Table 2. Comparison of contact area between experiment and simulation
Vertical Load (N)
Footprint Area (mm
2
)
Simulation Experiment Percentage Error
2920 1.07E+04 1.11E+04 -3.60%
2940 1.08E+04 1.12E+04 -3.57%
3890 1.41E+04 1.45E+04 -2.76%
3970 1.42E+04 1.48E+04 -4.05%
4975 1.60E+04 1.65E+04 -3.03%
4980 1.66E+04 1.70E+04 -2.35%


3. EFFECTS OF ROLLING CONDITONS ON TIRE-PAVEMENT CONTACT STRESS
DISTRIBUTION

The effect of rolling conditions on tire-pavement contact stress distribution is next
investigated. In order to study the relationship between tire dynamics and non-uniform
interfacial pressure between tire tread and pavement, the slip ratio has to be first defined. Slip
ratio (S) can be expressed as
(1 ) 100%
wr
S
v
= (7)
where w

is the wheel angular velocity; r is the radius of the deformed wheel; v is the
vehicle velocity. When S =0 (i.e. zero slip), the tire is in a free rolling state. A positive slip
(i.e. S >0) indicates that the vehicle is in a braking state whereas a negative slip (i.e. S <0)
indicates that the vehicle is in a driving state.

In this paper, the PIARC smooth tire is assumed to have the calibrated material parameters
described in Table 1. Other operating conditions used in this paper are described in Table 3.
Figures 8 to 10 show the three-dimensional contact stress profiles under three different rolling
conditions (i.e. free rolling, braking and driving states). It is illustrated from these figures that
the shape of the stress distribution is similar to that measured experimentally by Douglas
(2009) for passenger car tire rolling on steel plates (Figure 11). Although some differences in
magnitude can be observed due to specific differences in tire and other operating conditions,
the numerical tire model can be said to effectively simulate the variation trend of contact
stress distribution. This capability of finite element tire model to effectively simulate contact
stress distribution is very important to pavement design and evaluation.

Table 3. Operating conditions considered in this study
Parameters Value
Velocity 40 km/h
Tire inflation pressure 200 kPa
Vertical load 4980 N
Skid number at zero speed 20







YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
147


-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-1
-0.5
0
0.5
1
x10
5
T
r
a
c
t
i
o
n

s
t
r
e
s
s

(
N
/
m
2
)
Rearedge
Frontedge
Transversedirection(m)
Longitudinal direction(m)
(a) Traction stress distribution

-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0
2
4
6
x10
5
V
e
r
t
ic
a
l
s
t
r
e
s
s

(
N
/
m
2
)
Transversedirection(m)
Longitudinal direction(m)
Rearedge
Frontedge
(b) Contact stress distribution

Figure 8. Traction and contact stress distribution at driving state (S =-0.4)
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
148


-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-1
-0.5
0
0.5
1
x10
5
T
r
a
c
t
io
n

s
t
r
e
s
s

(
N
/
m
2
)
Frontedge
Rearedge
Transversedirection(m)
Longitudinal direction(m)
Transversedirection(m)
(a) Traction stress distribution

-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0
2
4
6
x10
5
V
e
r
t
i
c
a
l

s
t
r
e
s
s

(
N
/
m
2
)
Transversedirection(m)
Frontedge
Rearedge
Longitudinal direction(m)
(b) Contact stress distribution

Figure 9. Traction and contact stress distribution at free-rolling state (S =0)












YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
149

-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-1
-0.5
0
0.5
1
x10
5
T
r
a
c
t
i
o
n

s
t
r
e
s
s

(
N
/
m
2
)
Transversedirection(m)
Frontedge
Rearedge
Longitudinal direction(m)
(a) Traction stress distribution

-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
-0.1
-0.08
-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
0.1
0
2
4
6
x10
5
V
e
r
t
i
c
a
l

s
t
r
e
s
s

(
N
/
m
2
)
Transversedirection(m)
Rearedge
Frontedge
Longitudinal direction(m)
(b) Contact stress distribution

Figure 10. Traction and contact stress distribution at braking state (S =0.4)



Figure 11. Measured tire longitudinal and transverse stresses for a 195/70 R40 passenger car
tire of 210 kPa inflation pressure (Douglas, 2009)

YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
150
Figure 12 compares vertical stress distributions in different rolling states. The vertical stress
distribution along the tire longitudinal distance almost does not change with the rolling state,
which is similar to findings of past researchers (De Beer, 1994; De Beer, 1997; Douglas,
2000; Douglas, 2009). Moreover, the shape is similar to that reported by Douglas (2009). The
difference in magnitudes between Figure 12 and that reported by Douglas (2009) can be
attributed to the different tires and wheel loads tested.

Unlike the free rolling state, the traction stress distribution shifts a little forward to the front
edge in the driving state, and backward to the rear edge in the state of braking. Figure 13
shows distributions of longitudinal traction force in the different rolling state. It can be clearly
observed that the longitudinal traction force distribution along with tire longitudinal distance
varies significantly with rolling state. The maximum traction force appears in the acceleration
state. The traction force is also found to increase with decreasing slip ratio. Similarly, the
shape presented in the paper is similar to that reported by Douglas (2009) and the difference
in magnitudes is attributed to the different tires and wheel loads tested.
-0.1 -0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08 0.1
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
x10
5
Tirelongitudinal distance(m)
V
e
r
t
i
c
a
l

s
t
r
e
s
s

(
N
/
m
2
)


Slipratio40
Slipratio0
Slipratio-40

Figure 12. Comparison of predicted vertical contact stresses in different slip ratios

-0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08
-1
-0.8
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
x10
5
Tirelongitudinal distance(m)
T
r
a
c
t
io
n

s
t
r
e
s
s

(
N
/
m
2
)


Slipratio40
Slipratio0
Slipratio-40

Figure 13. Comparison of predicted longitudinal traction force in different slip ratios
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
151
4. CONCLUSION

This paper has presented the development of a numerical tire pavement interaction model to
simulate the contact and traction stresses at the tire-pavement interface. Through the use of
finite element modeling, it was found that the model can produce contact patches that are
similar to that observed in experiments. The model is then applied to study the effect of three
different rolling conditions (free-rolling, driving and braking) on the development of vertical
contact and traction stresses. It was found that the simulation model can effectively predict the
variation in contact stress distributions for the three rolling conditions. Due to the various
simplifications assumed in the paper, this model could be further enhanced to simulate and
analyze the stresses developed within the pavement layers under different rolling conditions.


REFERENCES

ADINA R&D, Inc. (2009) ADINA 8.6 Manuals. ADINA R&D, Inc. Massachusetts, USA.
AASHTO (1996) AASHTO Guide for Design of Pavement Structures. American Association
of State Highway and Transportation Officials. Washington D.C.
Bathe, K.J . (1985) Finite Element Procedures. New J ersey, USA.
Blab, R. (1999) Introducing improved loading assumptions into analytical pavement models
based on measured contact stresses of tires. Presented at the International Conference on
Accelerated Pavement Testing, Reno, Nevada, October 18-20.
Chae, S.W., Bathe, K.J . (1989) On automatic mesh construction and mesh refinement in finite
element analyses. Computers and Structures, 32 (3/4), 911-936.
Chang, Y.B., Yang, T.Y. (1983) Linear dynamic analysis of revolutional shells using finite
elements and modal expansion. Journal of Sound and Vibration, 86, 523-538.
De Beer, M. (1994) Measurement of Tyre/Pavement Interface Stresses under Moving Wheel
Loads. Final Report, Center for structural mechanics: University of Pretoria, South
Africa.
De Beer, M., Fisher, M.C., J ooste, F.J . (1997) Determination of pneumatic tyre/pavement
interface contact stresses under moving loads with some effects on pavements with thin
asphalt surfacing layers. Proceedings of the 8th International Conference on Asphalt
Pavements, International Society for Asphalt Pavements, 179-227, Seattle, Washington,
August 8-14.
Drakos, C.A. (2003) Identification of Physical Model to Evaluate Rutting Performance of
Asphalt Mixtures. PhD dissertation. University of Florida, Gainesville.
Douglas, R.A. (2009) Tyre/Road Contact Stresses Measured and Modeled in Three
Coordinate Directions. NZ Transport Agency Research Report 384, Wellington, New
Zealand.
Douglas, R.A., Woodward, W.D.H., Woodside, A.R. (2000) Road contact stresses and forces
under tires with low inflation pressure. Canadian Journal of Civil Engineering, 27,
1248-1258.
Howell, W.E., Perez, S.E., Volger, W.A. (1985) Aircraft tire footprint forces. Tire Pavement
Interface, ASTM STP 929, Pottinger, M.G. and Yager, T.J . (eds.), American Society for
Testing and Materials, Philadelphia, 110-124.
Huang, Y.H. (1993) Pavement Analysis and Design. Prentice-Hall, Inc. Englewood Cliffs,
New J ersey.
Lippmann, S.A. (1985) Effects of Tire Structure and Operating Conditions on the Distribution
of Stress between the Tread and the Road. Tire Pavement Interface, ASTM STP 929,
YANG, J., et al. / Asian Transport Studies, Volume 2, Issue 2 (2012), 139-152
152
Pottinger, M.G. and Yager, T.J . (eds.), American Society for Testing and Materials,
Philadelphia, 91-109.
Myers, L.A. (2000) Development and Propagation of Surface-Initiated Longitudinal Wheel
Path Cracks in Flexible Highway Pavements. PhD thesis. University of Florida.
Ong, G.P., Fwa, T.F. (2007) Wet-pavement hydroplaning risk and skid resistance: modeling,
ASCE Journal of Transportation Engineering, 133 (9), 590-598.
Park, D.W. (2005) Effects of non-uniform tire contract stresses on pavement response. ASCE
Journal of Transportation Engineering, 131 (11), 873-879.
PIARC (2004) Specification for a Standard Test Tyre for Friction Coefficient Measurement of
a Pavement Surface: Smooth Test Tyre. World Road Association.
Yoder, E.J . (1994) Principles of Pavement Design. Wiley and Sons, Inc.