Construction and Building Materials 361 (2022) 129711

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Numerical simulation of mechanical response analysis of asphalt pavement

under dynamic loads with non-uniform tire-pavement contact stresses
Tongxu Wang a, Zejiao Dong a, *, Ke Xu b, Shafi Ullah a, Donghao Wang c, Yiheng Li c
a
School of Transportation Science and Engineering, Harbin Institute of Technology, Harbin 150090, China
b
China Airport Planning and Design Institute Co. Ltd, Beijing 100101, China
c
Heilongjiang Communications Investment Group Co. Ltd, Harbin 150090, China

A R T I C L E I N F O A B S T R A C T

Keywords: The selection of accurate input design parameters is important for understanding the dynamic response of an
Tire-pavement contact asphalt pavement structure. The transverse and longitudinal shear stresses that affect the dynamic response of
Finite element method asphalt pavement are not considered in traditional asphalt pavement design. To this end, this study utilizes the
Asphalt pavement design
finite element method (FEM) to analyze the mechanical response of asphalt pavement under dynamic loads with
Superposition effect
Pressure measurement
non-uniform tire-pavement contact stresses. For this purpose, firstly a three-dimensional (3D) finite element (FE)
model of a rubber tire is developed. Secondly, the predicted and laboratory measured stresses are compared to
develop a 3D FE model of the contact stresses. Finally, a 3D FE model of the tire–pavement contact is analyzed to
investigate the influence of shear stresses on the response of the asphalt pavement and the superposition effect of
the mechanical response. The results show that the distribution and magnitude of the vertical contact pressure
under the free-rolling state are the same as those under the static state, and the traction and braking conditions
could change the symmetry of tire-pavement interactions. In addition, the shear stress at the tire-pavement
interface affects the peak response of the surface layer, indicating that shear stress plays a considerable role
in controlling the development of rutting and longitudinal cracks. Considering the viscoelasticity of the asphalt
material, it is observed that the superposition effect of the mechanical response is a comprehensive result of the
space–time domains. In conclusion, this study analyzes the influence of tire-pavement contact on the mechanical
response of asphalt pavement and proposes input design parameters for improving the design of asphalt
pavement.

1. Introduction efficient in calculating pavement response, their application range and

analytical accuracy are limited because of the use of many simplifying
The asphalt pavement gradually deteriorates due to pavement- assumptions. Furthermore, in most cases, analytical solutions cannot be
vehicle interactions [1,2]. For proper design and analysis, it is impor­ customized for complex problems [12]. Compared with the analytical
tant to analyze the response of asphalt pavements under dynamic loads. method, the numerical method has become a popular method for
The mechanical–empirical road design method treats the pavement pavement mechanical analysis due to its simplicity and high efficiency.
structure as an elastic structure based on the elastic layer theory. The Numerical methods can handle the non-linear behavior of pavement
traffic load is assumed to be concentrated or distributed linearly over the materials, complex tire-pavement contact stresses, and irregular geo­
total width of the wheel, and this assumption has been widely adopted metric conditions [13–15]. Special finite element software has been
due to its computation simplicity [3,4]. However, the mechanical developed for the mechanical analysis of asphalt pavement [16–19].
response of asphalt pavement varies significantly due to the tire- In general, FEM is mainly used for the analysis of the influence of
pavement interaction because of the viscoelasticity of the asphalt ma­ material characteristics, external environment and moving load on the
terial [5,6]. Analytical solutions for viscoelastic pavement responses response distribution and performance of asphalt pavement. It has been
vary with the complexity of the assumptions regarding the pavement reported that the nonlinear characteristics of asphalt materials, such as
structure and moving loads [7–11]. Although analytical solutions are viscoelasticity and anisotropy, influence the dynamic response and

* Corresponding author.
E-mail address: hitdzj@hit.edu.cn (Z. Dong).

https://doi.org/10.1016/j.conbuildmat.2022.129711
Received 19 July 2022; Received in revised form 22 October 2022; Accepted 7 November 2022
Available online 16 November 2022
0950-0618/© 2022 Elsevier Ltd. All rights reserved.
T. Wang et al. Construction and Building Materials 361 (2022) 129711

damage development of the pavement [20–22]. The 3D viscoelastic FE 2. Methodology

models were established using FE software such as ABAQUS and
COMSOL, and it was found that the nonlinear characteristics of asphalt To achieve the objective of this study, firstly a 3D FE model of a
materials have a significant impact on the rutting and fatigue perfor­ rubber tire was developed. Secondly, the contact characteristics of the
mance of the pavement, which confirmed the feasibility and accuracy of tire-pavement interface were determined through the FEM and then
the FEM based on experience in these studies [23–26]. In addition, the experimentally validated by the measurement system to provide load
external environment and moving loads are also key factors affecting the input for the simulation of tire-pavement interactions. Finally, the
dynamic response of asphalt pavement, which were also considered in model was analyzed to determine the effect of shear stresses on the
the FE analysis. To study the pavement response in the actual environ­ dynamic response of the asphalt pavement and the superposition effect
ment and moving loads, measured data were collected and utilized to of the structural response under multi-axial loading. Fig. 1 shows the
verify the feasibility of the FEM and modify the simulation parameters to flowchart of this study.
improve the fitting effect of the FEM [27–31]. Due to the complex stress
distribution of tires, the interaction between the tire and pavement can
increase the possibility of rutting and cracking. To simplify the calcu­ 2.1. 3D finite element modeling of the rubber tire
lations, the current design method of asphalt pavement assumes that the
contact area of the tire-pavement interaction is circular and the vehicle 2.1.1. The simplified structure of tire
load is uniformly distributed, which is different from the actual field A 3D FEM model of rubber tire typed 12R22.5 was developed with an
conditions [16,32,33]. However, it is difficult to predict why the damage outer diameter and section width of 1085 mm and 300 mm, respectively.
to asphalt pavement under heavy vehicles gradually develops from the The tire comprised components such as tread, sidewall, bead, steel belt,
surface using a simple uniform load in FE analysis [16,34–37]. There­ carcass ply and inner liner as shown in Fig. 2(a). The carcass ply con­
fore, it is necessary to study the tire-pavement interaction and complex sisted of a single layer of steel cord, and the steel belt consisted of four
stress distribution for the dynamic response analysis of pavement layers of steel cords. The 3D model was developed by rotating the mesh
structure. cross section 360◦ around the tire axis as shown in Fig. 2(b).
Studies have shown that the tire-pavement contact stress consists of
three parts: transverse, longitudinal and vertical [33,35–38]. The dis­ 2.1.2. Material parameters of rubber tire
tribution of tire-pavement contact stress is complex and random, and is The tread and the sidewall, as deformable parts that interact directly
greatly affected by the tire type [39–41], inflation pressure [42–44], with the pavement surface, were composed of super-elastic rubber ma­
load level [30,45], rolling conditions [46,47], tilting angle [33,48], and terials, and the Yeoh model was used for their constitutive modeling as
pavement surface texture [47,49–50]. Given this, there are two main shown in Eq. (1) [60,61]. Similarly, the carcass ply and tire belts were
methods in the study of tire-pavement interaction: direct measurement cord-rubber composites to support traffic loads that can be modeled with
and numerical simulation. Recently, some smart methods have been a rebar model. The material parameters of the tire are listed in Table 1.
developed to measure the tire-pavement contact stresses directly
U = C10 (I1 − 3) + C20 (I1 − 3)2 + C30 (I1 − 3)3 (1)
[42,51], such as smart sensing [31], ultrasound [32], and functional
image processing technology [52,53]. Although ultrasound and func­ where U is the strain energy of the rubber material;C10 ,C20 ,C30 are
tional image processing technology clearly reveal the actual character­ the parameters of the shear properties of the material; I1 is the strain
istics of tire-pavement contact stresses, the accuracy of these methods invariant.
strongly depends on the algorithm itself, which has the disadvantages of
high cost, limited accuracy, and unsuitability for engineering promo­ 2.1.3. Loading conditions
tion. In view of this, the smart sensing method has become the main A uniform normal pressure of 0.83 MPa was adopted to the inner
measurement method owing to its high precision and simple operating liner in order to simulate the inflation pressure of a 12R22.5 tire. The
principle, which can simultaneously measure dynamic and 3D contact standard tire load, a vertical concentrated load of 33.5 kN, was applied
stresses. In other words, the direct measurement method is suitable for to the tire. From this, static contact characteristics of the tire under
obtaining the tire contact-stress distribution under specific testing con­ standard conditions were obtained, and the peak value of the middle rib
ditions. In contrast, the numerical simulation method is simple and was about 1.23 MPa. However, the interaction between tire and pave­
efficient for simulating the tire contact stresses of different types and ment is a dynamic contact problem. The model achieves traction and
rolling conditions. Therefore, the influence of major factors on the braking through the translation of the tire axis and the rotation of the
interaction of the tire-pavement interface [54], such as tire rolling entire tire around the axis. When the translation and rotation speeds are
resistance [55–58], nonlinearity of pavement material [46,59], pave­ coordinated, the tire can roll freely.
ment surface anti-skid [33] and noise [35], has been investigated This paper assumed that the translation velocity is 72 km/h, and the
through numerical simulations. From the literature, the quantitative rotational velocity can be calculated from the translation velocity and
models of tire-pavement contact stresses have been established using the tire radius. The velocities under free-rolling conditions are listed in
direct measurement or numerical simulation methods. It is worth Table 2. The Lagrangian and Eulerian coordinate systems were used to
mentioning that the accuracy of the numerical simulation depends on improve computational efficiency in simulation [62]. Then, the contact
the rationality of the simulation parameters. Due to the complexity of stress and distribution of the tire model under different rolling condi­
the simulation parameters, the simulation results are quite different tions were obtained and analyzed.
from the measured results, and few studies have discussed the simula­ The assumptions of the model are listed in Table 3. For simplicity, the
tion parameters validated by the measured data. longitudinal pattern grooves were retained and the transverse textures
This study concentrates on the mechanical response analysis of a were omitted in the simplified tire structure. The rubber material of a
semi-rigid base asphalt pavement under moving loads with non-uniform tire is a hyperelastic material that is approximately incompressible. The
tire-pavement contact stresses based on the FEM. The FEM results were tire structure contains a variety of composite materials such as carcass
validated by a pressure sensor-based laboratory measurement system to ply and tire belt, which are steel cords embedded into the inner rubber
provide load input for the tire-pavement interaction simulation. The material; therefore, the rebar model was adopted for simulation. In
results of this study provide reliable load input for mechanical response addition, three types of constraints were selected in the model: the ri­
analysis and optimized design of asphalt pavement. gidity constraint between the rim and tire, embedded constraints be­
tween the steel cord and tire rubber, and contact constraints between the
tire and pavement. The pavement surface was assumed to be rigid [44].

2
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Fig. 1. Flowchart of the experimental program.

Fig. 2. Modeling of the radial truck tire typed 12R22.5.

Table 1
Material parameters of the tire model.
Part Yeoh constants (MPa) Density (kg/m3) Young’s modulus (MPa) Poisson ratio Degree (◦ )

C10 C20 C30

Tread 0.77 − 0.07 0.1 1100

Sidewall 1.34 − 0.03 0.15 1100
Inner liner 1.34 − 0.03 0.15 1100
Belt 7800 190,200 0.3 66 ~ 105
Carcass ply 7800 190,200 0.3 0
Bead 7800 190,200 0.3 90

2.2. Measurement of tire-pavement contact using Tekscan pressure sensor sensor was composed of two polyester sheets with a total thickness of
approximately 0.1 mm. On the first sheet, the piezoresistive materials
A pressure sensor-based indoor tire-pavement contact measurement were placed at equal distances. Similarly, the same sensors were placed
system was developed to evaluate the accuracy of the developed model. in the same layout but perpendicular to the first sheet on the second
A hydraulic press was used to load the tire, and a Tekscan sensor system sheet. When the two sheets connected, the resistance at the row-column
with high repeatability and accuracy was used to measure the contact intersection of the sensors decreased as the tire load increased, and the
pressure. The working principle behind Tekscan is based on a resistive pressure displayed in the software increased accordingly. The electronic
polymer whose electrical resistance changes with applied stress. The sheet was incompressible and operated at temperatures from − 40 ◦ C to

3
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Table 2 mechanical response of a typical semi-rigid base asphalt pavement was

Rolling conditions of the tire model. analyzed using a 3D FE model, as shown in Fig. 4. The assumptions of
Inflation Pressure Tire load Translation velocity Free-rolling velocity the model are listed in Table 4. Owing to the symmetry of the loading, an
(MPa) (kN) (km/h) (rad/s) FE model of the half-width pavement structure was established in this
0.83 10 72 37.18 study. To reduce the influence of the model size, a completely fixed
15 72 37.32 boundary condition was used for the bottom of the model, and no
20 72 37.43 boundary conditions were used for the side of the model, where an
25 72 37.51 infinite element mesh generation method was adopted. A linear
30 72 37.56
33.5 72 37.60
continuous element with reduced integration (C3D8R) was chosen for
the interior of the model, whereas an infinite element (CIN3D8) was
chosen for the three sides of the model. The contact conditions between
pavement layers were assumed to be completely continuous.
Table 3
Although moisture variations of the subgrade soil significantly affect
Assumptions of 3D FE model of rubber tire.
the pavement response [63–65], this study focuses on the contact stress
Content Assumptions distribution at the interface of the tire and pavement surface and its
Simplified structure Tread, sidewall, bead, steel belt, carcass ply and inner liner effect on the pavement response. It has been proven that it is reasonable
of tire to use viscoelastic parameters for asphalt materials and elastic param­
Tire tread pattern The longitudinal patterns were retained and the transverse
eters for other materials [25,28]. The time and temperature dependence
patterns were omitted
Rubber material The Yeoh model was adopted due to the hyperelasticity of the visco-elastic parameters of asphalt material was fitted with the
Cord - Rubber The carcass ply and tire belts were cord-rubber composite Prony series and the Williams-Landel-Ferry (WLF) equation using
material material, where the rebar model was adopted ABAQUS software [9,25]. The materials of the semi-rigid base, semi-
Constraint condition The rigidity constraint between the rim and tire, the rigid subbase and subgrade were considered elastic and the effect of
embedded constraints between steel cord and tire rubber,
and contact constraints between tire and pavement
damping was ignored. A S-shape function of the relaxation modulus was
Loading conditions The translation velocity is 72 km/h, and the rotational fitted to determine the time dependence at selected reference temper­
velocity can be calculated from the translation velocity and ature, and the shift factor was determined by Eq. (2). The constants C1
the tire radius and C2 in the WLF equation were calculated by Eq. (3).
Pavement Rigid
t
aT = (2)
ξ
60 ◦ C with ambient humidity between 5 % and 90 %. The hydraulic
press applied a maximum pressure of 250 kN, which met the measure­
ment requirements. The Tekscan sensor consisted of 36,608 sensels with
an effective measuring area of 0.085 m2 as shown in Fig. 3. To evaluate
the tire-pavement contact, several tire inflation pressures and load levels
were selected. The tire inflation pressure included 0.5 MPa, 0.6 MPa,
0.7 MPa, 0.83 MPa, 0.9 MPa, and 1.0 MPa. Similarly, load levels
included 10 kN, 15 kN, 25 kN, 30 kN and 33.5 kN. The contact pressure
was carried out in a static state.

2.3. 3D FE modeling of tire-pavement interaction

Fig. 4. Schematic diagram of asphalt pavement structure under half-axle load.
The combined effect of moving loads and 3D contact stresses on the

Fig. 3. The indoor tire-pavement contact measurement based on Tekscan sensors in the laboratory.

4
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Table 4 Table 6
Assumptions of 3D FE model of tire-pavement interaction. Constants of Prony series.
Content Assumptions τi gi

Material Viscoelastic parameters for asphalt materials and elastic for SMA-13 AC-20
parameters other materials
0.00001 0.60859 0.48456
Contact condition Completely continuous
0.0001 0.20486 0.18250
Boundary Completely fixed boundary condition for the bottom of the
0.001 0.07580 0.14727
conditions model, and no boundary conditions for the side of the model
0.01 0.04505 0.08145
Meshing C3D8 for the interior of the model and CIN3D8 for the side
0.1 0.02161 0.02994
Loading Non-uniformly, moving and 3D loading obtained from section
1 0.00668 0.01818
conditions 3.2.3
10 0.00745 0.00884
100 0.00101 0.00395
1000 0.00196 0.00280
− C1 (T − Tref )
log(aT ) = (3) Note: τi is the i’th specified Prony retardation time constant, gi is the i’th cor­
C2 + (T − Tref ) responding Prony constant.
Where T is the test temperature, Tref is the reference temperature,aT
is the shift factor at Tref, t is the time at T before conversion,ξ is the
Table 7
relaxation time at the reference temperature Tref after conversion, C1 and
WLF equation constants.
C2 are regression constants.
The simplified model of the half asphalt pavement structure was Materials Tref (℃) C1 C2

built with a model geometry of 10 m in the direction of traffic, 5 m in the SMA-13 20.0 27.8 219.8
transverse direction, and 3 m in depth. The 3D contact stress model used AC-20 20.0 32.3 265.1
in this study is verified based on the measured and simulated data. The
load moving speed was 72 km/h, and the material parameters of the
imprint, as shown in Fig. 5 (a). Similarly, the maximum and minimum
pavement structural layers are shown in Table 5, Table 6 and Table 7.
contact pressure under free-rolling condition in Fig. 5 (b) are 1.27 MPa
in the center and 0.97 MPa, respectively on the right side, and the
3. Results and discussions
maximum and minimum under traction condition are 1.21 MPa in the
center and 0.98 MPa, respectively on the left side in Fig. 5 (c). The peak
3.1. Simulation analysis of tire-pavement contact stress characteristics
pressure at the center under free-rolling is greater than that under
braking and traction because the peak pressure moves forward during
3.1.1. Vertical pressure distribution
braking and backward during traction, which may increase the wearing
Fig. 5 shows the transverse peak and the distribution of vertical
of the tire.
pressure, where it can be observed that braking and traction change the
position of the pressure on both sides of the tire. The distribution
3.1.2. Transverse stress distribution
characteristics of vertical pressure are the same in the static and free-
Fig. 6 shows that the transverse stress distribution is slightly affected
rolling states, and the vertical pressure of the three middle ribs is rela­
by the rolling conditions of the tire. Similarly, the transverse stress is
tively stable under different rolling conditions. It is further confirmed
smaller than the vertical pressure and longitudinal stress. The transverse
that the contact area is non-uniform and non-circular, and the pressure
stress developed from the transverse deformation, which was caused by
distribution under dynamic conditions remains transversely symmetric.
the compression of the longitudinal pattern on the pavement. Because
The vertical pressure distribution characteristics under different rolling
the transverse deformation develops into the gaps on both sides of the
conditions are similar, although the amplitudes are different, as listed in
longitudinal pattern, the transverse stress directions on both sides of the
Table 8. That is, the maximum and minimum of vertical contact pressure
middle three ribs are opposite, and the transverse stress directions on the
under static condition are 1.23 MPa and 0.63 MPa, respectively, which
inside of the first and last ribs are also opposite. It is worth mentioning
were obtained from section 2.1.3. It can be observed that the maximum
that the greater the stiffness of the tire rubber material, the greater the
contact pressure under the braking condition is 1.24 MPa at the center of
inflation pressure of the tire, and the smaller the load level. The resulting
the imprint and the minimum is 0.95 MPa on the right side of the
transverse deformation and stresses change accordingly. Therefore, it
can be concluded that the transverse stress is more susceptible to the tire
Table 5 material characteristics and operating conditions. The transverse stress
Material parameters of pavement structural layers. distribution characteristics under different rolling conditions are shown
Layer Material Depth Density Transient Poisson’s in Table 9.
(cm) (kg/ modulus ratio
m3) (MPa) 3.1.3. Longitudinal stress distribution
Asphalt SMA-13 4 2400 1750 0.35 As shown in Fig. 7, the longitudinal shear stress remains transversely
surface symmetric, and the longitudinal shear stress in the free-rolling state is
course smaller than that in the braking and traction states. Due to the friction
Asphalt AC-20 5 2400 3150 0.35
middle
between the tire and the pavement, the position of the maximum lon­
course gitudinal shear stress remains unchanged from braking to traction, but
Asphalt AC-20 7 2400 3150 0.35 the shear stress direction changes: that is, the shear stress is always
bottom positive during traction and negative during braking. The longitudinal
course
stress distribution characteristics under the different rolling conditions
Semi-rigid Lime-fly ash 38 2100 6000 0.3
base aggregate are shown in Table 10. The maximum longitudinal contact stress under
(LFA) the braking condition is − 0.45 MPa, which is distributed at the bottom
Semi-rigid Lime soil (LS) 36 1900 2000 0.35 of the contact area in Fig. 7 (a), and the minimum longitudinal contact
subbase stress in the contact area is zero. The contact area of the longitudinal
Subgrade Soil 210 1900 250 0.40
stress under the traction condition is larger than that under the braking
Note: SMA = Stone mastic asphalt, AC = Asphalt concrete.

5
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Fig. 5. Vertical pressure distribution under different rolling conditions.

Table 8
Distribution characteristics of vertical pressure.
Rolling condition Tire contact pressure type Imprint shape Distribution along tire length Contact pressure (MPa)

Max./Mini.

Static Non-uniform Rectangular Half sine wave 1.23/0.63

Braking Non-uniform Rectangular Half sine wave 1.24/0.95
Free-rolling Non-uniform Rectangular Half sine wave 1.27/0.97
Traction Non-uniform Rectangular Half sine wave 1.21/0.98

Fig. 6. Transverse stress distribution under different rolling conditions.

Table 9
Distribution characteristics of transverse stress. Table 10
Rolling Tire contact pressure Imprint Contact stress Distribution characteristics of longitudinal stress.
condition type shape (MPa) Rolling Tire contact pressure Imprint Contact stress
Max./Mini. condition type shape (MPa)

Braking Non-uniform Rectangular 0.26/-0.28 Max./Mini.

Free-rolling Non-uniform Rectangular 0.23/-0.25 Braking Non-uniform Rectangular − 0.45/0
Traction Non-uniform Rectangular 0.23/-0.24 Free-rolling Non-uniform Oval 0.07/-0.05
Traction Non-uniform Rectangular 0.38/0

Fig. 7. Longitudinal stress distribution under different rolling conditions.

6
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Fig. 8. Measurement results of vertical pressure.

condition, and the maximum longitudinal contact stress of the contact 3.2.2. Validation of contact stress characteristics
area under traction condition is 0.38 MPa. The longitudinal stress under The tire-pavement contact consists of two parts: geometric properties
free-rolling condition is small, it can be observed from Fig. 7 (b) that the and mechanical properties. As shown in Fig. 9 (a), the transverse (hor­
imprint shape is oval. izontal) maximum value of the measurement result is the contact width,
the longitudinal (vertical) maximum is the contact length, and the
3.2. Experimental analysis of contact stress characteristics product of these two is the projected area. Due to the existence of lon­
gitudinal grooves, there is a series of voids in the projected area, and the
3.2.1. Measurement results of tire-pavement contact contact area is the area between the tread and pavement. The average
As shown in Fig. 8, the contact area and pressure are non-uniform at contact pressure was calculated from the tire load and contact area, and
different inflation pressures and load levels. At an inflation pressure of the ratio of the void area to the contact area is called the area ratio.
0.83 MPa, the contact area increases with increasing load, and the shape Furthermore, the accuracy of the proposed model was verified by
of the footprint gradually approaches a drum, as shown in Fig. 8 (a). As comparing the distribution of vertical pressure for transverse and lon­
shown in Fig. 8 (b), under different loads, the peak pressure is always gitudinal distances with the measured results.
located at the center of the contact area, the average contact pressure of
the middle rib remains unchanged, and the pressure at the sidewall in­ 3.2.2.1. Geometrical properties of vertical pressure. The geometrical
creases with increasing load. Similarly, as the inflation pressure in­ properties of the tire vertical pressure include contact length, contact
creases, the contact area gradually decreases and the pressure at the width, contact area, and projected area, as shown in Fig. 9 (a). Taking
center increases significantly. In most cases, the average contact pres­ the standard inflation pressure of 0.83 MPa and load level of 33.5 kN as
sure at the sidewall is close to the selected inflation pressure of 0.83 comparison conditions, it was found that the simulated section width
MPa. Therefore, it can be concluded that the pressure at the sidewall and and overall diameter of the tire are 297 mm and 1087 mm, respectively,
contact area are mainly affected by the load level, while the inflation which are consistent with the standard values of 300 mm and 1085 mm.
pressure has a certain influence on the contact pressure at the center of Furthermore, the errors between the simulated results (S) and measured
the tire and its contact length. The distribution characteristics of the results (M) of these geometric property indicators are less than 3 %, as
vertical pressure measurements are shown in Table 11. shown in Fig. 9 (b) ~ Fig. 9 (f), indicating a good correspondence be­
tween them. Therefore, the FE model of the rubber tire proposed in
Section 2.1 is considered reasonable.

3.2.2.2. Mechanical characteristics of vertical pressure. Considering that

Table 11 the tire load–deflection curve is critical to the acceptance of the tire by
Distribution characteristics of vertical pressure. the manufacturer, this paper compared the simulated and measured
Inflation load Imprint Contact Average contact load–deflection curves. Taking the standard inflation pressure (0.83
pressure (MPa) level shape area (m2) pressure (MPa) MPa) and load level (33.5 kN) as comparison conditions, it can be
(kN) observed from Fig. 10 (a) that the simulated and measured deflections
0.83 10 Oval 0.018 0.53 are 32.6 mm and 33.5 mm, respectively, indicating an error of 1.9 %.
15 Oval 0.026 0.63 That is, the simulated deflections under various loads are in good
20 Drum- 0.032 0.68
agreement with the measured deflections, which proves the rationality
shaped
25 Drum- 0.037 0.71 of the tire model material setting. To further verify the accuracy of the
shaped developed model, the distribution characteristics of the vertical contact
30 Drum- 0.041 0.73 pressure, including the distribution along the transverse and the distri­
shaped bution along the longitudinal distances were analyzed. As shown in
33.5 Rectangular 0.045 0.75
Fig. 10 (b), the simulated maximum vertical pressure for each rib is
0.5 33.5 Rectangular 0.056 0.60
0.6 Rectangular 0.052 0.64 located at the center, while the measured maximum is not at the center
0.7 Rectangular 0.049 0.68 due to the presence of the transverse patterns. Similarly, due to the
0.83 Rectangular 0.046 0.72 presence of transverse patterns, the simulated pressure curve is
0.9 Rectangular 0.045 0.74
smoother than the measured pressure curve, but the change trend and
1.0 Rectangular 0.044 0.77

7
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Fig. 9. Comparison of geometric properties.

Fig. 10. Comparison of mechanical properties.

amplitude of the curves are consistent, as shown in Fig. 10 (c). In reaches a peak in the middle of the asphalt surface course. Therefore, the
summary, the simulated contact pressure is in good agreement with the middle of the asphalt surface course was selected as the analysis layer,
measured pressure, indicating that the developed tire model can accu­ and the center of the dual-tire was selected as the transverse position for
rately predict the tire-pavement contact characteristics. the analysis.
Fig. 12 presents the mechanical responses under 3D stress (3D),
3.2.3. Numerical model of 3D contact stresses vertical pressure (V) and tangential stress (T), including transverse and
Compared with previous studies that applied rectangular and uni­ longitudinal shear stresses. Table 12 lists the comparison of the peak
formly distributed vertical concentrated loads, this study adopted a 3D responses under different loads. The proportion in Table 12 represents
numerical model of the contact stress under different rolling conditions, the degree of influence of the tangential stress, with positive values
which is more in line with actual tire loading. As shown in Fig. 11, the indicating an increase in the response and negative values indicating a
3D contact stress numerical model is divided into five ribs, the widths of decrease in the response. From the results, it can be observed that the
the five ribs from left to right are 42 mm, 33 mm, 30 mm, 33 mm and 42 tangential shear stress affects the peak stress and strain of the asphalt
mm, respectively, and the width of the gap between each rib is surface course to some extent, particularly the longitudinal and trans­
approximately 12 ~ 13 mm. Each rib was divided into 8 ~ 10 rectangles verse responses. Fig. 12 (a) shows that the stress under 3D stress is
along the driving direction with a width of 25 mm and the value rep­ smaller than that under vertical pressure, which also means that it is
resents the 3D stress. To study the mechanical response of asphalt more conservative to select the vertical compressive stress of each
pavement under moving loads, ABAQUS subroutine platforms were asphalt layer as the index of permanent deformation of the asphalt
adopted, and other simulation operations were the same as in previous mixture layer in the current design method. However, the peak strain
studies [25,66–68]. The subroutine DLOAD was utilized to load the under 3D stress is higher than that under vertical pressure, and its effect
vertical pressure, and the subroutine UTRACLOAD was utilized to load is more significant than that under normal stress. Similarly, the
the longitudinal stress and transverse stress under different rolling tangential shear stress also increases the peak shear stress and strain,
conditions. and its effect may even exceed that of the vertical pressure. Therefore, to
reduce the rutting and longitudinal cracking of asphalt pavements, this
study proposes to consider the effect of tangential shear stress in the
3.3. Analysis of asphalt pavement mechanical responses under moving optimization design of asphalt pavement and to include a check of the
loads tangential shear stresses of the asphalt layer.

3.3.1. Effect of tangential shear stresses 3.3.2. Superposition effect of the mechanical response
It can be observed that the mechanical response of the pavement The traditional superposition principle holds that the responses
under a dual-tire load decreases gradually with increasing depth and

8
T. Wang et al. Construction and Building Materials 361 (2022) 129711

position. In view of the viscoelasticity of the asphalt material, it is

necessary to analyze the superposition of the vertical displacement of
the pavement. As shown in Fig. 13 (c), the peak value and influence
range of the vertical displacement also increases with an increase in the
number of tires and axles, and the displacement increases by 0.2 mm −
0.3 mm from each additional axle load.
Fig. 14 (a) shows the time domain results, which can be identified by
the number and value of peaks, and the influence of the rear axle is
greater than that of the front axle. With an increase in the axles, the
transverse stress and vertical pressure increase gradually, and only the
maximum longitudinal stress under triple dual tandem load decreases,
indicating that the superposition effect in the space domain is greater
than that in the time domain. Due to the reduction effect between
adjacent axles in the spatial domain, the longitudinal stress of the first
axle under dual tandem load is smaller than that under dual tires.
Fig. 14 (b) shows the results for the spatial domain. It can be
observed that the stress at the bottom of the subbase has only one peak
due to the superposition effect, and the position of the peak under the
dual tire is located at the center of the wheel gap. Similar to the results of
the time domain analysis, only the longitudinal stress decreases. It is
believed that the influence of the spatial domain is smaller than that of
the time domain, which is due to the excessive distance between the
axles, which reduces the stress between the adjacent axles. In addition,
the transverse and vertical stresses do not increase exponentially, and
the increase from uniaxial to biaxial loading is greater than that from
biaxial to triaxial loading. It should be noted that the longitudinal stress
at the bottom of the subbase is greater than the transverse stress under
uniaxial loading, while the opposite is true under muti-axial loading.
Therefore, this study proposes replacing the longitudinal stress in the
current specification with maximum horizontal stress.

3.4. Comparison with other models

The numerical simulation method simply and effectively simulates

Fig. 11. Numerical model of 3D contact stresses under different roll­ the influence of the main factors on the tire-pavement interface inter­
ing conditions. action [51], such as tire rolling resistance [52–55], non-linearity of

caused by loads are independent of each other, and the responses under Table 12
multi-axial loads are equal to the linear superposition under a single Comparison of dynamic response under different loads.
load [69]. However, this does not work for nonlinear asphalt materials. Peak response 3D T V Proportion
To determine the superposition effect of the asphalt pavement response,
Vertical pressure (kPa) − 805.94 − 12.08 − 816.53 − 1.3 %
four types of loads were selected, as shown in Fig. 13 (a). Where Td, Dt,
Transverse stress (kPa) − 271.45 43.37 − 306.9 − 11.6 %
Dw and Si represent the load patterns of the triple dual tandem, dual Longitudinal stress (kPa) − 324.46 25.56 − 348.22 − 6.8 %
tandem, dual tire and single tire, respectively. Vertical strain (με) − 3523.5 − 98.66 − 3454.5 +2.0 %
As shown in Fig. 13 (b), the bottom of the subbase has no obvious Transverse strain (με) 759.72 187.94 625.75 +21.4 %
Longitudinal strain (με) − 289.47 − 51.01 − 243.35 +19 %
response under the action of the first axle load, however, on the basis of
Shear stress τxy (kPa) 18.91 10.77 8.57 +120.7 %
the first axle, the response becomes obvious under the action of the Shear stress τxz (kPa) − 115.32 − 14.39 − 102.76 +12.2 %
second and third axle loads, that is, the time domain superposition effect Shear stress τyz (kPa) 106.11 39.43 71.11 +49.2 %
appears from the base. In view of the requirements of the current Shear strain εxy (με) 227.14 137.74 89.4 +154.1 %
specification for inorganic binder semi-rigid base asphalt pavement, the Shear strain εxz (με) − 1350.61 − 209.72 − 1207.5 +11.9 %
Shear strain εyz (με) 1478.64 506.31 1137.81 +30.0 %
design index is located at the bottom of the inorganic binder stabilized
layer. Therefore, the bottom of the subbase was selected as the analysis Note: τij is the tangential shear stress, εxy is the tangential shear strain, i, j = x, y,
layer, and the center of the outer tire was selected as the transverse z, represent transverse, longitudinal and vertical direction, respectively.

Fig. 12. Time-history curve of mechanical response under different loads.

9
T. Wang et al. Construction and Building Materials 361 (2022) 129711

Fig. 13. Schematic diagram of multiaxial loads and vertical displacement.

Fig. 14. Superposition results of mechanical response under multiaxial loadings.

pavement material [43,56], and pavement roughness [70–72]. Howev­ 4. Conclusions

er, most of these studies focused on the vertical contact pressure of the
tire-pavement interaction. There is still no widely accepted method for The main objective of this study was to analyze the mechanical
characterizing the 3D contact stresses of tire-pavement interactions. response of a semi-rigid base asphalt pavement under moving loads with
However, the interaction between the tire and the pavement increases non-uniform tire-pavement contact stresses. The feasibility and accu­
the likelihood of rutting and cracking due to the presence of shear racy of the FEM were verified using indoor tire-pavement contact
stresses. Therefore, compared with the general method of vertical measurement system. The 3D contact stresses under different rolling
pressure, this study adopted the 3D contact stress model shown in conditions were obtained, and the influence of shear stresses on the
Fig. 11, which is more in line with the actual service condition of the mechanical response of the asphalt pavement, and the superposition
pavement. effect of the mechanical response were studied. The main findings of this
Furthermore, the vertical pressure was measured using an indoor study are summarized as follows:
tire-pavement contact measurement system to adjust the accuracy of the
developed FE model. Compared with other smart methods to directly 1. The contact characteristics of the 3D model shows that tire rolling
measure tire-pavement interaction, such as ultrasound [29] and func­ has little effect on the transverse stress distribution. The longitudinal
tional image processing technology [49,50], smart sensing has the ad­ shear stress remains transversely symmetrical and varies in the free-
vantages of high precision, simple operating principle, and simultaneous rolling, braking, and traction states. The vertical pressure distribu­
measurement of dynamic and 3D contact stresses. To ensure the accu­ tion is the same in the static and free-rolling states, and is relatively
racy of the results, the contact pressure was measured using a Tekscan stable under different rolling conditions.
sensor system with high repeatability and accuracy, consisting of 36,608 2. The pressure sensor-based indoor tire-pavement contact measure­
sensels with an effective measurement area of 0.085 m2, as shown in ment system provides an accurate verification of the tire-pavement
Fig. 3. The maximum pressure applied by the hydraulic press was 250 contact stress distribution based on FEM.
kN, which met the measurement requirements for vehicles and aircraft 3. The peak response of the asphalt pavement indicates that the vertical
tires. Based on the developed indoor measurement system, future compressive stress of the asphalt layers significantly affects perma­
studies could measure the contact stress of different types of tires under nent deformation. Similarly, tangential shear stress increases the
turning and rolling conditions by means of dynamic pressure sensors. peak value of the dynamic response and must be considered in the
design to reduce rutting and longitudinal cracking of the asphalt
pavement.

10
T. Wang et al. Construction and Building Materials 361 (2022) 129711

4. The superposition effect indicates that the transverse stress is [10] J.H. Lee, J.K. Kim, J.L. Tassoulas, Dynamic analysis of a layered half-space
subjected to moving line loads, Soil Dyn. Earthq. Eng. 47 (2013) 16–31, https://
sometimes greater than the longitudinal stress, because the me­
doi.org/10.1016/j.soildyn.2012.07.013.
chanical response is a combined result of the time and space do­ [11] N.D. Beskou, E.V. Muho, Dynamic response of a finite beam resting on a Winkler
mains. Therefore, this study proposes replacing the longitudinal foundation to a load moving on its surface with variable speed, Soil Dyn. Earthq.
stress with the maximum horizontal stress in current specifications. Eng. 109 (2018) 222–226, https://doi.org/10.1016/j.soildyn.2018.02.033.
[12] C. Wu, H. Wang, J. Zhao, X. Jiang, Q. Yanjun, B. Yusupov, Prediction of
Viscoelastic Pavement Responses under Moving Load and Nonuniform Tire Contact
In summary, the method proposed in this study effectively demon­ Stresses Using 2.5-D Finite Element Method, Math. Probl. Eng. 2020 (2020) 1–16.
strates the tire-pavement contact characteristics and provides a reliable [13] D.H. Allen, D.N. Little, R.F. Soares, C. Berthelot, Multi-scale computational model
for design of flexible pavement–part I: expanding multi-scaling, Int. J. Pavement
load input for mechanical response analysis. In addition, the proposed Eng. 18 (2017) 309–320, https://doi.org/10.1080/10298436.2015.1065999.
pavement design indexes are beneficial for optimizing the pavement [14] R.A. Tarefder, M.U. Ahmed, M.R. Islam, M.T. Rahman, Finite Element Model of
design and improving its service life. Further study is required to mea­ Pavement Response Under Load Considering Cross-Anisotropy in Unbound Layers,
Adv. Civ. Eng. Mater. 3 (2014) 20130102, https://doi.org/10.1520/
sure the transverse and longitudinal contact stresses for the validation of acem20130102.
3D FE models of different types of rubber tires and to study the rela­ [15] C. Du, P. Liu, C. Jin, H. Yang, C. Li, G. Jiang, M. Oeser, Evaluation of the
tionship between multi-axial loads and the standard load defined in the piezoelectric and mechanical behaviors of asphalt pavements embedded with a
piezoelectric energy harvester based on multiscale finite element simulations,
current specification. Constr. Build. Mater. 333 (2022), 127438, https://doi.org/10.1016/j.
conbuildmat.2022.127438.
CRediT authorship contribution statement [16] X. Jiang, C. Zeng, X. Gao, Z. Liu, Y. Qiu, 3D FEM analysis of flexible base asphalt
pavement structure under non-uniform tyre contact pressure, Int. J. Pavement Eng.
20 (2019) 999–1011, https://doi.org/10.1080/10298436.2017.1380803.
Tongxu Wang: Data curation, Formal analysis, Validation, Writing – [17] P. Liu, D. Wang, J. Hu, M. Oeser, SAFEM - Software with graphical user interface
original draft. Zejiao Dong: Conceptualization, Funding acquisition, for fast and accurate finite element analysis of asphalt pavements, J. Test. Eval. 45
(2016) 1301–1315, https://doi.org/10.1520/JTE20150456.
Project administration, Supervision, Writing – review & editing. Ke Xu:
[18] P. Liu, Q. Xing, Y. Dong, D. Wang, M. Oeser, S. Yuan, Application of Finite Layer
Investigation, Methodology, Validation. Shafi Ullah: Validation, Su­ Method in Pavement Structural Analysis, Appl. Sci. 7 (2017) 611, https://doi.org/
pervision, Writing – review & editing. Donghao Wang: Investigation, 10.3390/app7060611.
Methodology, Validation. Yiheng Li: Investigation, Methodology, [19] Y. Norouzi, S.H. Ghasemi, A.S. Nowak, M. Jalayer, Y. Mehta, J. Chmielewski,
Performance-Based Design of Asphalt Pavements Concerning the Reliability
Validation. Analysis, Constr. Build. Mater. 332 (2022) 127393.
[20] G. Danoshini, A. Baqadeem, A.K.S. Al-Shakhrit, N.E. Jasni, K.A. Masri, Statistical
Analysis of Stone Mastic Asphalt Incorporating Kenaf Fibre, CONSTRUCTION. 1
Declaration of Competing Interest (2021) 12–17. https://doi.org/10.15282/construction.v1i1.6282.
[21] S.N. Aliyu Yaro, D.M.H. Sutanto, D.A. Usman, A.H. Jagaba, M.Y. Sakadadi,
The authors declare that they have no known competing financial E. Shaffie, R.P. Jaya, J. Ahmad, A.K. Arshad, M.A. Zihan, F. Shiong, The Influence
of Waste Rice Straw Ash as Surrogate Filler for Asphalt Concrete Mixtures, Adv.
interests or personal relationships that could have appeared to influence
Mater. Sci. Eng. 2022 (2022) 118–125. https://doi.org/10.15282/construction.
the work reported in this paper. v2i1.7624.
[22] E. Shaffie, R.P. Jaya, J. Ahmad, A.K. Arshad, M.A. Zihan, F. Shiong, S.K. Selvaraj,
Data availability Prediction Model of the Coring Asphalt Pavement Performance through Response
Surface Methodology, Adv. Mater. Sci. Eng. 2022 (2022) 1–17.
[23] H. Wang, I.L. Al-Qadi, Importance of Nonlinear Anisotropic Modeling of Granular
Data will be made available on request. Base for Predicting Maximum Viscoelastic Pavement Responses under Moving
Vehicular Loading, J. Eng. Mech. 139 (2013) 29–38, https://doi.org/10.1061/
(asce)em.1943-7889.0000465.
Acknowledgment [24] Y. Zhang, F. Gu, X. Luo, B. Birgisson, R.L. Lytton, Modeling Stress-Dependent
Anisotropic Elastoplastic Unbound Granular Base in Flexible Pavements, Transp.
This work is supported by the National Key R&D Program of China Res. Rec. 2672 (2018) 46–56, https://doi.org/10.1177/0361198118758318.
[25] F.V. Souza, L.S. Castro, Effect of temperature on the mechanical response of
(2018YFB1600100); National Natural Science Foundation of China thermo-viscoelastic asphalt pavements, Constr. Build. Mater. 30 (2012) 574–582,
(U20A20315, 51878228); Heilongjiang Communications Investment https://doi.org/10.1016/j.conbuildmat.2011.11.048.
Group Co., Ltd (JT-100000-ZC-FW-2021-0181); Fundamental Research [26] J. Liao, S. Sargand, Viscoelastic FE Modeling and Verification of a U.S. 30
Perpetual Pavement Test Section, Road Mater, Pavement Des. 11 (2010)
Funds for the Central Universities. 993–1008, https://doi.org/10.1080/14680629.2010.9690316.
[27] X. Ma, Z. Dong, F. Chen, H. Xiang, C. Cao, J. Sun, Airport asphalt pavement health
References monitoring system for mechanical model updating and distress evaluation under
realistic random aircraft loads, Constr. Build. Mater. 226 (2019) 227–237, https://
doi.org/10.1016/j.conbuildmat.2019.07.174.
[1] S. Ullah, S. Wan, C. Yang, X. Ma, Z. Dong, Self-stress and deformation sensing of
[28] Z. Dong, T. Wang, X. Ma, C. Cao, F. Kong, Z. Leng, Dynamic response analysis of
electrically conductive asphalt concrete incorporating carbon fiber and iron
airport asphalt pavement subjected to high-temperature jet wake based on finite
tailings, Struct. Control Heal. Monit. (2022) 1–17, https://doi.org/10.1002/
element simulation, J. Test. Eval. 48 (3) (2020) 20190650.
stc.2998.
[29] X. Ma, W. Quan, Z. Dong, Y. Dong, C. Si, Dynamic response analysis of vehicle and
[2] X. Ma, Z. Dong, Y. Dong, Stiffness identification method for asphalt pavement
asphalt pavement coupled system with the excitation of road surface unevenness,
layers and interfaces using monitoring data from built-in sensors, Struct. Heal.
Appl. Math. Model. 104 (2022) 421–438, https://doi.org/10.1016/j.
Monit. (2022) 147592172210776. https://doi.org/10.1177/14759217221077612.
apm.2021.12.005.
[3] T. Király, P. Primusz, C. Tóth, Simulation of Static Tyre-Pavement Interaction
[30] J. Zhu, S. Sargand, I. Khoury, D. Tarawneh, T. Ma, F. Chen, Dynamic load
Using Two FE Models of Different Complexity, Appl. Sci. 12 (5) (2022) 2388.
responses of perpetual pavement test roads on U.S. 23: Full-scale instrumentation
[4] M. De Beer, Measurement of tyre/pavement interface stresses under moving wheel
and evaluation, Constr. Build. Mater. 331 (2022), 127326, https://doi.org/
loads, Heavy Veh. Syst. 3 (1996) 97–115.
10.1016/j.conbuildmat.2022.127326.
[5] H. Wang, I.L. Al-Qadi, Evaluation of surface-related pavement damage due to tire
[31] I.L. Al-Qadi, H. Wang, P.J. Yoo, S.H. Dessouky, Dynamic analysis and in situ
braking, Road Mater. Pavement Des. 11 (2010) 101–121, https://doi.org/10.1080/
validation of perpetual pavement response to vehicular loading, Transp. Res. Rec.
14680629.2010.9690262.
2087 (1) (2008) 29–39.
[6] J. Chen, H. Wang, M. Li, L. Li, Evaluation of pavement responses and performance
[32] R.S. Dwyer-Joyce, B.W. Drinkwater, In situ measurement of contact area and
with thermal modified asphalt mixture, Mater. Des. 111 (2016) 88–97, https://doi.
pressure distribution in machine elements, Tribol. Lett. 14 (2003) 41–52, https://
org/10.1016/j.matdes.2016.08.085.
doi.org/10.1023/A:1021766216019.
[7] L. Khazanovich, Q.(. Wang, MnLayer: High-performance layered elastic analysis
[33] J.A. Hernandez, Numerical prediction of three dimensional tire-pavement contact
program, Transp. Res. Rec. 2037 (1) (2007) 63–75.
stresses, 2017.
[8] S.M. Kim, B.F. McCullough, Dynamic response of plate on viscous Winkler
[34] R.V. Siddharthan, N. Krishnamenon, M. El-Mously, P.E. Sebaaly, Investigation of
foundation to moving loads of varying amplitude, Eng. Struct. 25 (2003)
tire contact stress distributions on pavement response, J. Transp. Eng. 128 (2002)
1179–1188, https://doi.org/10.1016/S0141-0296(03)00066-X.
136–144, https://doi.org/10.1061/(ASCE)0733-947X(2002)128:2(136).
[9] Z. Dong, X. Ma, Analytical solutions of asphalt pavement responses under moving
[35] Y. Oubahdou, E.-R. Wallace, P. Reynaud, B. Picoux, J. Dopeux, C. Petit, D. Nélias,
loads with arbitrary non-uniform tire contact pressure and irregular tire imprint,
Effect of the tire – Pavement contact at the surface layer when the tire is tilted in
Road Mater. Pavement Des. 19 (2018) 1887–1903, https://doi.org/10.1080/
bend, Constr. Build. Mater. 305 (2021) 124765.
14680629.2017.1354776.

11
T. Wang et al. Construction and Building Materials 361 (2022) 129711

[36] M. De Beer, C. Fisher, F.J. Jooste, Evaluation of non-uniform tyre contact stresses [55] E.Y. Manyo, P. Reynaud, B. Picoux, R. Tautou, F. Allou, C. Petit, D. Nélias,
on thin asphalt pavements, in: Ninth Int. Conf. Asph. Pavements, 2002: pp. 19–22. Tire–pavement tractive rolling contact under turning conditions: towards
[37] D.W. Park, A.E. Martin, E. Masad, Effects of nonuniform tire contact stresses on pavement top-down cracking, Int. J. Pavement Eng. 23 (2022) 841–850, https://
pavement response, J. Transp. Eng. 131 (2005) 873–879, https://doi.org/10.1061/ doi.org/10.1080/10298436.2020.1775231.
(ASCE)0733-947X(2005)131:11(873). [56] M. Guo, X. Zhou, Tire-Pavement Contact Stress Characteristics and Critical Slip
[38] D.-W. Park, E. Fernando, J. Leidy, Evaluation of Predicted Pavement Response with Ratio at Multiple Working Conditions, Adv. Mater. Sci. Eng. 2019 (2019) 1–11.
Measured Tire Contact Stresses, Transportation Research Record: Journal of the [57] M. Guo, X. Li, M. Ran, X. Zhou, Y. Yan, Analysis of contact stresses and rolling
Transportation Research Board 1919 (2005) 160–170. resistance of truck-bus tyres under different working conditions, Sustain. 12 (2020)
[39] D. Beer, C. Fisher, F.J. Jooste, Determination of Pneumatic Tyre-Pavement 1–16, https://doi.org/10.3390/su122410603.
Interface Contact Stresses under Moving Loads and some Effects on Pavements [58] E.Y. Manyo, B. Picoux, P. Reynaud, R. Tautou, D. Nelias, F. Allou, C. Petit,
with Thin Asphalt Surfacing Layers, in: 8th Int. Conf. Asph. Pavements, 1997: pp. Approach of pavement surface layer degradation caused by tire contact using semi-
179–227. analytical model, Materials (Basel). 14 (9) (2021) 2117.
[40] H. Wang, I.L. Al-Qadi, Impact quantification of wide-base tire loading on secondary [59] K.e. Zhong, M. Sun, Z. Liu, K. Zheng, Research on dynamic evaluation model and
road flexible pavements, J. Transp. Eng. 137 (2011) 630–639, https://doi.org/ early warning technology of anti-sliding risk for the airport pavement, Constr.
10.1061/(ASCE)TE.1943-5436.0000245. Build. Mater. 239 (2020) 117820.
[41] H. Wang, I.L. Al-Qadi, I. Stanciulescu, Simulation of tyre-pavement interaction for [60] J. Zhang, G.L. Wang, N.J. Fu, K. Wang, Finite element analysis of some radial tire,
predicting contact stresses at static and various rolling conditions, Int. J. Pavement Adv. Mater. Res. 490-495 (2012) 2414–2418.
Eng. 13 (2012) 310–321, https://doi.org/10.1080/10298436.2011.565767. [61] C. Renaud, J.-M. Cros, Z.-Q. Feng, B. Yang, The Yeoh model applied to the
[42] E. Diserens, Calculating the contact area of trailer tyres in the field, Soil Tillage Res. modeling of large deformation contact/impact problems, Int. J. Impact Eng. 36 (5)
103 (2009) 302–309, https://doi.org/10.1016/j.still.2008.10.020. (2009) 659–666.
[43] T. Keller, Soil Compaction and Soil Tillage – Studies in Agricultural Soil Mechanics, [62] M. Ziefle, U. Nackenhorst, Numerical techniques for rolling rubber wheels:
2004. Treatment of inelastic material properties and frictional contact, Comput. Mech. 42
[44] E.G. Fernando, D. Musani, D.-W. Park, W. Liu, Summary Report : Evaluation of (3) (2008) 337–356.
Effects of Tire Size and Inflation Pressure on Tire Contact Stresses and Pavement [63] J. Pooni, D. Robert, F. Giustozzi, C. Gunasekara, S. Setunge, S. Venkatesan,
Response, 7 (2006) 288. Hydraulic characteristics of stabilised expansive subgrade soils in road pavements,
[45] X. Hu, A.N.M. Faruk, J. Zhang, M.I. Souliman, L.F. Walubita, Effects of tire Int. J. Pavement Eng. 23 (2022) 3129–3146, https://doi.org/10.1080/
inclination (turning traffic) and dynamic loading on the pavement stress–strain 10298436.2021.1883610.
responses using 3-D finite element modeling, Int. J. Pavement Res. Technol. 10 [64] C.E. Cary, C.E. Zapata, Unsaturated soil modeling for airfield pavement design,
(2017) 304–314, https://doi.org/10.1016/j.ijprt.2017.04.005. J. Transp. Eng. 140 (2014) 50–60, https://doi.org/10.1061/(ASCE)TE.1943-
[46] H. Ge, J.C. Quezada, V.L. Houerou, C. Chazallon, Multiscale analysis of tire and 5436.0000619.
asphalt pavement interaction via coupling FEM–DEM simulation, Eng. Struct. 256 [65] G. Gajári, L. Kisgyörgy, S. Ádány, A. Mahler, J. Lógó, A visco-hypoplastic
(2022) 113925. constitutive model for rolled asphalt, Period. Polytech, Civ. Eng. 65 (2021)
[47] H. Wang, I.L. Al-Qadi, I. Stanciulescu, Effect of Surface Friction on Tire-Pavement 798–809, https://doi.org/10.3311/PPci.17515.
Contact Stresses during Vehicle Maneuvering, J. Eng. Mech. 140 (2014) 04014001, [66] X. Ma, Z. Dong, X. Yu, F. Chen, C. Cao, J. Sun, Monitoring the structural capacity of
https://doi.org/10.1061/(asce)em.1943-7889.0000691. airfield pavement with built-in sensors and modulus back-calculation algorithm,
[48] S.I. Lee, A.D. Mwanza, G. Mutembo, L.F. Walubita, Effects of Tire Inclination on the Constr. Build. Mater. 175 (2018) 552–561, https://doi.org/10.1016/j.
HMA Pavement Shear Stress-Strain Response: 2-D Computational Modeling, in: conbuildmat.2018.04.198.
2014: pp. 41–48. https://doi.org/10.1061/9780784478462.006. [67] Z. Dong, X. Ma, X. Gong, M. Oeser, Theoretical evaluation of the measurement
[49] S. Dong, S. Han, Y. Luo, X. Han, O. Xu, Evaluation of tire-pavement noise based on accuracy of fiber Bragg grating strain sensors within randomly filled asphalt
three-dimensional pavement texture characteristics, Constr. Build. Mater. 306 mixtures based on finite element simulation, Struct. Control Heal. Monit. 25 (2018)
(2021) 124935. 1–15, https://doi.org/10.1002/stc.2057.
[50] B. Chen, X. Zhang, J. Yu, Y. Wang, Impact of contact stress distribution on skid [68] W. Xiu-Shan, F. Tao-Tao, G. Yan-Xu, H.u. Yi, Research on China’s high-grade
resistance of asphalt pavements, Constr. Build. Mater. 133 (2017) 330–339, highway design index of asphalt pavement with granular base, Sci. Cold Arid Reg.
https://doi.org/10.1016/j.conbuildmat.2016.12.078. 5 (4) (2013) 498.
[51] K.M. Tarhini, G.R. Frederick, Wheel Load Distribution in I-Girder Highway Bridges, [69] W.N. Findley, J.S.Y. Lai, A Modified Superposition Principle Applied to Creep of
J. Struct. Eng. 118 (1992) 1285–1294, https://doi.org/10.1061/(asce)0733-9445 Nonlinear Viscoelastic Material Under Abrupt Changes in State of Combined Stress,
(1992)118:5(1285). Trans. Soc. Rheol. 11 (1967) 361–380, https://doi.org/10.1122/1.549091.
[52] E.Y. Manyo, P. Reynaud, B. Picoux, R. Tautou, D. Nelias, F. Allou, C. Petit, Towards [70] B. Goenaga, L. Fuentes, O. Mora, A Practical Approach to Incorporate Roughness-
fast modelling of the tire-pavement contact, Eur. J. Environ. Civ. Eng. 25 (2021) Induced Dynamic Loads in Pavement Design and Performance Prediction, Arab. J.
2396–2412, https://doi.org/10.1080/19648189.2019.1628812. Sci. Eng. 44 (2019) 4339–4348, https://doi.org/10.1007/s13369-018-3414-9.
[53] H. Taghavifar, A. Mardani, Potential of functional image processing technique for [71] S. Misaghi, C. Tirado, S. Nazarian, C. Carrasco, Impact of pavement roughness and
the measurements of contact area and contact pressure of a radial ply tire in a soil suspension systems on vehicle dynamic loads on flexible pavements, Transp. Eng. 3
bin testing facility, Meas. J. Int. Meas. Confed. 46 (2013) 4038–4044, https://doi. (2021) 100045.
org/10.1016/j.measurement.2013.07.019. [72] J. Zhao, H. Wang, Mechanistic-empirical analysis of asphalt pavement fatigue
[54] D. Han, G. Zhu, H. Hu, L. Li, Dynamic simulation analysis of the tire-pavement cracking under vehicular dynamic loads, Constr. Build. Mater. 284 (2021) 122877.
system considering temperature fields, Constr. Build. Mater. 171 (2018) 261–272,
https://doi.org/10.1016/j.conbuildmat.2018.03.071.

12