Energy and Buildings 102 (2015) 187–196

Contents lists available at ScienceDirect

Energy and Buildings

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

Cooling asphalt pavement by a highly oriented heat conduction

structure
Du Yinfei, Wang Shengyue ∗ , Zhang Jian
School of Transportation, Southeast University, Nanjing 210096, PR China

a r t i c l e i n f o a b s t r a c t

Article history: In this paper, a highly oriented heat conduction structure of asphalt pavement, with a combination of
Received 30 March 2015 low thermal conductivity layer and three-layered gradient heat conduction structure, was proposed to
Received in revised form 11 May 2015 reduce pavement temperature and decrease nighttime heat release into the atmosphere in summer.
Accepted 12 May 2015
The structure was formed by modifying contrast asphalt pavement by adding different dosages of low
Available online 21 May 2015
thermal conductivity powders to each layer. Also, it made full use of principles of thermal insulation
and gradient heat conduction, and extended the scope of thermal gradient in asphalt layers. The results
Keywords:
showed that, compared with contrast structure, the highest temperature of upper surface of bottom
Cool pavement
Gradient thermal conductivity
layer, which was used to represent the average temperature of middle and bottom layers, reduced by
High orientation 2.3 ◦ C (simulation result) and 2.4 ◦ C (test road result). The average temperatures of middle and bottom
Pavement temperature layers reduced by 1.6 ◦ C (at 2:30 pm) and 1.5 ◦ C (at 6:00 pm), respectively, which were validated by test
Heat budget road results. Calculations of simulation result displayed that the structure released less 12.1% of heat to
the atmosphere during nighttime than contrast structure. According to the results summarized above,
it is concluded that the structure has a continuous cooling capacity, and is expected to reduce high
temperature rutting of asphalt pavement and help to reduce high air temperature at night.
© 2015 Elsevier B.V. All rights reserved.

1. Introduction [11], thermochromic asphalt [21,22], phase-change material [23],

Asphalt pavement has a high absorption of solar radiation, often
leading to a high temperature on the surface of asphalt pavement
and triggering a series of distresses such as rutting [1,2]. High tem-
perature on road surface is one of the important reasons for forming
urban heat island effect [3–5], which usually causes huge energy
consumption in big cities in summer [6,7]. In addition, strong and
continuous thermal radiation from road surface not only reduces
the comfort level of urban residents, also easily induces diseases
such as cardiovascular [8,9]. To mitigate this problem, paving cool
asphalt pavements [10,11] and controlling the heat conduction in
asphalt pavements [12] can be used. High temperature rutting,
which can be reduced by establishing a gradient thermal con-
ductivity in asphalt layers [13,14], can result in hydroplaning and
nonuniform tire-pavement contact stresses, further aggravating
pavement distresses and driving safety [15].
In the past decades, many techniques have been applied to
cool asphalt pavement in summer, such as solar reflective coating
[16–18], retro reflective films [19,20], colored thin layer asphalt
∗ Corresponding author. Tel.: +86 025 83790551.
E-mail address: sywang@seu.edu.cn (W. Shengyue).
permeable pavement [24,25] and asphalt solar collector [26,27].
All these cooling technologies achieved ideal practical effects. For
example, solar reflective coating could reduce the surface tempera-
ture by 12 ◦ C [10]. However, the technology based on the principle
of light reflection usually brings new difficulties which are hard
to overcome so far. These include driver’s dazzling caused by the
strong reflected light [10]. In addition, because of the high rough-
ness of asphalt pavement, the heat reflection belongs to a typical
diffuse. So most of the reflected heat will be re-absorbed by the
objects on both sides of road, causing secondary heat pollution [28].
For permeable pavement, its abundant pores are easy to be blocked
[29]; it is also prone to raveling, which requires an excellent vis-
cosity of asphalt [30]. For methods such as paving energy collector
pavement, since the inner stress distribution of the pavement is
significantly changed [31], rutting is easier to happen and expand.
Establishing a gradient heat conduction channel in asphalt pave-
ment has proved to be an efficient strategy for cooling hot asphalt
pavements in summer [13,14,32]. In general, asphalt pavements
are composed of three asphalt layers. The structures in Refs. [13,14]
were only composed of two layers of gradient heat conduction in
middle and bottom layers, in addition to a low thermal conductiv-
ity layer in top layer. Though the structure in Ref. [32] was designed
by establishing three layers of gradient heat conduction, in practice

http://dx.doi.org/10.1016/j.enbuild.2015.05.020
0378-7788/© 2015 Elsevier B.V. All rights reserved.
188 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196

Fig. 1. Test specimens.

it was found that, due to a high setting of thermal conductivity of
top layer, too much solar radiant heat was permitted to enter pave-
ment, which was not only easy to cause rutting but also inclined
to release more accumulated heat to the surrounding air in the
night. Thus, the heat conduction structure with high thermal con-
ductivity of top layer reported in Ref. [32] does not apply to those
cities which prefer to have cool asphalt pavements at night, such
as London (United Kingdom), a typical city with nighttime urban
heat island effect. According to Refs. [33–36], the urban heat island
intensity of London at night is 5–8 ◦ C higher than that during the
day. Obviously, more heat release will exacerbate the thermal pol-
lution in these cities at night. To avoid too much solar radiation
heat entering the pavement in the day and forming a release peak
in the night, we put forward a highly oriented heat conduction
structure with a combination of low thermal conductivity layer and
three-layered gradient heat conduction structure. Here the highly
oriented heat conduction structure was designed on the base of a
three-layered contrast asphalt pavement.
2. Materials and methods
of limestone. The gradation of Superpave 13 mixture is shown in
Table 1.
In this paper, the thermal properties of asphalt mixture were
changed by adding graphite or floating beads. As the dosage of pow-
der can significantly affect the volume index and performances of
asphalt mixture [39], the maximum dosages of floating beads and
graphite were both less than 15% of asphalt volume.
2.2. Determination of thermal properties of asphalt mixture
The thermal properties of the specimens prepared by asphalt
mixtures with different proportions of powders added (Fig. 1) were
measured by using a thermal conductivity tester (DRM-II, made in
China). In the test, three specimens were classified as one group,
and the sizes of the specimens of each group were 20 × 20 × 6 cm,
20 × 20 × 2 cm and 20 × 20 × 6 cm. Before the test, the instrument
heated the specimens in order to achieve a transient temperature
distribution inside. Based on the principle of unsteady state, the
thermal properties of the specimens were achieved automatically,
as shown in Table 2.

2.1. Materials
The thermal properties of asphalt mixture are mainly influenced
by its void content [37]. So only a type of asphalt mixture, Super-
pave 13, was used to design heat conduction structures. The asphalt
mixture specimens were prepared with void content of 4.0% [38].
SBS modified asphalt was used and the asphalt content was 5.2%.
The coarse aggregate was basalt and the mineral filler was made
2.3. Model establishment
A two-dimensional heat-transfer model of asphalt pavement
was established by using the finite element software ABAQUS.
Model parameters, including boundary conditions and thickness
of each layer of the pavement, are shown in Fig. 2. The upper
boundary conditions of the pavement are shown in Table 3, in

Table 1
Aggregate gradation for Superpave 13 mixture.

Sieve sizes (mm) 16.0 13.2 9.5 4.75 2.36 1.18 0.6 0.3 0.15 0.075
Passing (% by weight) 100 93.1 68.1 42.4 28.7 22.3 13.7 9.3 8.7 6.3

Table 2
Test results of thermal properties of asphalt mixture.

Filler type Percentage by volume (%) Density (g/cm3 ) Thermal conductivity (W/m ◦ C) Heat capacity (J/(kg ◦ C))

0 1.0618 1485.49
5 0.7504 1507.30
Floating beads 2.45
10 0.6916 1526.54
15 0.6529 1541.73

0 1.0618 1485.49
5 1.1143 1322.07
Graphite 2.45
10 1.1802 1257.30
15 1.2207 1229.03
 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196 189

Reflected solar Heat Effective Table 3

radiation convection radiation Parameters for temperature field calculation.
Incident solar
radiation Solar radiation absorptivity 0.9
Convection coefficient [W/(m2 ◦ C)] 20
Initial temperature [◦ C] 25
Top layer 4 cm
Maximum air temperature Tmax [◦ C] 35
Middle layer 6 cm Minimum air temperature Tmin [◦ C] 20
Bottom layer 8 cm

Thermally Base layer 40 cm Thermally Table 4

insulated insulated Thermal properties for model calculation.

Structure layer Density Thermal Heat capacity

Subbase layer 20 cm
(g/cm3 ) conductivity (J/kg ◦ C)
(W/m ◦ C)

Subgrade 5m Base layer 2.2 1.56 911.7

Subbase layer 2.1 1.43 942.9
Subgrade 1.8 1.56 1040.0

Thermally insulated
10 m where T1 = daily mean air temperature (◦ C), T1 = (Tmax + Tmin )/2;
T2 = daily air temperature amplitude (◦ C), T2 = (Tmax − Tmin )/2;
Fig. 2. Pavement structure diagrame used for heat transfer calculation. t0 = initial phase (h), t0 = 9 h; ω = angular frequency (rad),
ω = 2␲/24 rad.Besides the data in Table 2, the thermal properties
needed in the model calculation are listed in Table 4.
which the calculation equations of solar radiation (Eq. (1)) and air
temperature (Eq. (2)) are from the results reported in Ref. [40].
⎧ c
3. Design procedure of the gradient heat conduction
⎪
⎪ 0 0 ≤ t < 12 − structure
⎪
⎪ 2
⎨
c c The main purpose of the gradient heat conduction structure to
q(t) = q0 cos mw(t − 12) 12 − ≤ t ≤ 12 + (1)
⎪
⎪ 2 2 be designed here is to reduce pavement temperature and decrease
⎪
⎪ c
⎩0 12 + < t ≤ 24 the nighttime heat release from asphalt pavement to its surround-
2 ing air in summer. As rutting is highly related to high pavement
where q0 = maximum radiation in a day (J/m2 ), q0 = 0.131mQ, temperature [41,42], and usually occurs in middle and bottom lay-
m = 12/c; Q = total amount of solar radiation (J/m2 ), ers [43], the upper surface’s temperature of bottom layer was used
Q = 20.1 × 106 J/m2 ; c = effective number of hours of solar radiation to represent the temperatures of middle and bottom layers. And its
(h), c = 10 h; ω = angular frequency (rad), ω = 2␲/24 rad. maximum value (SMTB), combined with the total nighttime heat
release (TNHR), was applied to design the structure. The design
Ta = T1 + T2 [0.96 sin(ω(t − t0 )) + 0.14 sin(2ω(t − t0 ))] (2) procedure can be found in the flow chart (Fig. 3).

Proportion design
Measurement of thermal properties of asphalt mixture
of asphalt mixture

Establishment of heat transfer model

Determination of powder type in top layer

Total nighttime
heat release Determination of powder type and
Design dosage in middle layer
index

Upper surface's Determination of powder

maximum temperature dosage in top layer
of bottom layer

Determination of powder type and

dosage in bottom layer

Determination of the structure

Fig. 3. Flow chart of the structure design.

190 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196

Fig. 4. Comparison diagram of design indexes in Group one.

Fig. 5. Comparison diagram of design indexes in Group two.

3.1. Determination of the type of powder to be added in top layer
In order to determine the powder type to be added in top layer
and compare the designed structure with the gradient heat conduc-
tion structure reported in Ref. [32], four structures were designed
and included in Group one, as shown in Table 5. In this table, num-
ber #0 was the contrast structure (CS), and the powder-adding ratio
of number #3 was determined with that of the structure reported
in Ref. [32].
Fig. 4 showed that number #2 had a minimum of night-
time heat release, indicating that the mixture with floating
beads was the most promising used to prevent heat release
into the atmosphere. Besides this, number #2 also had the low-
est upper surface’s temperature of bottom layer. So, to reduce
the absorption of solar radiation, and avoid heat accumulating
and releasing again during nighttime, choosing floating beads
instead of graphite as the additive of top layer was the most
reasonable.

Table 5 3.2. Determination of the powder dosage in middle layer

Heat conduction structures in Group one.

Structure Top layer Middle layer Bottom layer In order to expand the selection scope of the thermal con-
number ductivity of middle layer, 15% floating beads and 15% graphite
#0 0% 0% 0% were added in top and bottom layers respectively. Based on the
#1 0% +10% graphite +15% graphite thermal properties determined previously, five gradient heat con-
#2 +5% floating beads +10% graphite +15% graphite duction structures, named number #4–#8 as shown in Table 6,
#3 +5% graphite +10% graphite +15% graphite
were designed to study the impact of the thermal conductivity of
 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196 191

Fig. 6. Comparison diagram of design indexes in Group three.

Fig. 7. Comparison diagram of design indexes for all structures.

middle layer on the heat transfer characteristic of all the designed effect of the thermal conductivity of middle layer on the TNHR was
structures. not obvious, but the thermal conductivity of top layer had a great
Fig. 5 showed that the TNHR and SMTB of number #4–#8 dis- influence on the TNHR. The SMTB increased with the increase of the
played smaller values than those of the CS. But the TNHR differences thermal conductivity of middle layer, and the trend of the increase
between them (#4–#8) were very small. This suggests that the of the SMTB declined gradually. The reason is that the increase of the
thermal conductivity of middle layer accelerated the heat conduc-
tion from top layer to bottom layer, resulting in a heat accumulation
Table 6 in middle layer and then a temperature increase in this layer. Based
Heat conduction structures in Group two. on the analyses of Fig. 5 and Table 6, the optimum dosage of floating
Structure Top layer Middle layer Bottom layer
beads in middle layer was set to be 10%.
number

#0 0% 0% 0%
3.3. Determination of the powder dosage in top layer
#4 +15% floating beads +10% floating beads +15% graphite
#5 +15% floating beads +5% floating beads +15% graphite
#6 +15% floating beads 0% +15% graphite Since the floating beads content in middle layer has been set to
#7 +15% floating beads +5% graphite +15% graphite be 10%, in order to form a gradient heat conduction structure, the
#8 +15% floating beads +10% graphite +15% graphite content of floating beads in top layer were set to be 15%.
192 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196

Fig. 8. Calculated temperature results at depth of (a) 4 cm, (b) 10 cm and (c) 18 cm.

Fig. 9. Comparison diagram of heat budget.

 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196 193

Fig. 10. Pavement temperature measurement: (a) temperature probe arrangement; (b) automatic temperature recording instrument: (1) temperature recorder; (2) OHCS;
(3) CS; (4) time.

3.4. Determination of the powder dosage in bottom layer
After setting the powder dosages in top and middle layers,
another four heat conduction structures was designed as shown
in Table 7.
From Fig. 6, the curves of the SMTB and TNHR had similar
trends. It can be concluded that, due to the small effect of the ther-
mal conductivity of bottom layer on the solar absorption and heat
conduction of pavement, the differences of the TNHR and SMTB
between gradient heat conduction structures were very small.
In contrast, the TNHR and SMTB of number #10 both displayed
minimum values. So, no matter from cost saving or heat conduction
effect, it should be the ideal structure in Group three.
3.5. Determination of the heat conduction structure
To facilitate comparisons, the TNHR and SMTB of all the designed
structures are shown in Fig. 7.
According to Fig. 7, both curves showed different changing
trends, indicating that the two indexes were influenced by the syn-
ergy of the thermal conductivities of three layers. And this also
suggests that it is necessary to design heat conduction structures to

Fig. 11. Field temperature results at depth of (a) 4 cm, (b) 10 cm and (c) 18 cm.
194 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196

Fig. 12. Comparison diagram of average temperature differences of middle and bottom layers.

be used for different purposes. Although different heat conduction
structures may have different potential applications, to prevent hot
air above the road surface caused by heat release of asphalt pave-
ment at night, a highly oriented heat conduction structure with 15%
(top) + 10% (middle) + 0% (bottom) of floating beads in each layer,
named OHCS, will be more reasonable.
4. Results and discussion
4.1. Simulation results and discussion
From Fig. 9 it can be found that the heat budget of the CS and
GHCS differed very small. Beside this, the OHCS absorbed less heat
during the day and also released less heat at night than the CS and
GHCS. As a result, for the OHCS, the lowest heat accumulation and
release to air made it to be a cool pavement at night. Our calculation
suggested that in the day the OHCS absorbed less 2.53 × 104 J/m2 of
solar radiation and released less 2.42 × 104 J/m2 of heat in the night
than the CS, which accounted for 9.3% and 12.1% of those of the CS
respectively.

4.1.1. Comparison of temperature field 4.2. Field validation of temperature field

The temperature fields of the CS, the structure reported in Ref.
[32] (GHCS) and OHCS within 24 h on the pavement surface, and at In order to verify the practical effect of the optimizing struc-
the depth of 4 cm, 10 cm and 18 cm are shown in Fig. 8. ture OHCS, a test road was paved. Temperature probes were
In Fig. 8 it showed that the temperature differences between pre-embedded at the depth of 4 cm, 10 cm and 18 cm respectively
the three structures increased with depth, and the OHCS always while the test road was paved, as shown in Fig. 10(a). Temperatures
displayed the lowest temperature at each depth. The maximum were then automatically measured by using a temperature recor-
temperature differences between the OHCS and CS were 1.1 ◦ C (at ding instrument every 30 min as shown in Fig. 10(b). Since it is very
1:00 pm), 2.3 ◦ C (at 4:00 pm), and 1.3 ◦ C (at 7:00 pm), respectively. difficult to directly measure the heat budget of the OHCS test road,
The maximum difference of the SMTB such as 2.3 ◦ C, suggested the net heat exchange between the test road and the atmosphere
that the OHCS was the most promising structure to reduce rutting, was given by a calculation method and shown in Fig. 9. For factors
according to the results reported in Refs. [44,45]. such as wind and solar radiation, it is very difficult to control. So,
in order to better explain the real effects of the designed OHCS, the
two factors were not considered in this discussion, and only the
4.1.2. Comparison of heat budget
temperatures within the pavement were evaluated by measuring
The heat absorptions and releases, i.e., heat budgets, of the CS,
its temperature field, as shown in Fig. 11.
GHCS and OHCS within 24 h were compared, as shown in Fig. 9.
Comparing the field measuring data of Fig. 11 with the numeri-
cal simulation result of Fig. 8, it can be found that they had a similar
Table 7 changing trend of temperature. The maximum temperature differ-
Heat conduction structures in Group three. ences between the OHCS and CS were 1.6 ◦ C at the depth of 4 cm
(at 2:30 pm), 2.4 ◦ C at the depth of 10 cm (at 6:30 pm) and 0.8 ◦ C at
Structure Top layer Middle layer Bottom layer
number the depth of 18 cm (at 5:00 pm).
The mean value of temperatures at the depth of 4 cm and 10 cm
#0 0% 0% 0%
#9 +15% floating beads +10% floating beads +5% floating beads
was identified as the average temperature of middle layer. Simi-
#10 +15% floating beads +10% floating beads 0% larly, the mean value of temperatures at the depth of 10 cm and
#11 +15% floating beads +10% floating beads +5% graphite 18 cm was identified as the average temperature of bottom layer.
#12 +15% floating beads +10% floating beads +10% graphite The calculated average temperature differences between the CS and
#4 +15% floating beads +10% floating beads +15% graphite
OHCS are shown in Fig. 12.
 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196
At any moment the average temperature differences were both
positive. The maximum temperature difference reached 1.6 ◦ C (at
2:30 pm) for middle layer and 1.5 ◦ C (at 6:00 pm) for bottom layer,
indicating that the OHCS could effectively reduce the temperature
of the designed pavement.
In summary, the gradient heat conduction structure could sig-
nificantly reduce the absorption of solar radiation and avoid a heavy
heat accumulation in the pavement. The endothermic reduction
will reduce a high temperature of the road during the day, and
directional release of the internal heat to subgrade will avoid an
upward heat release into the atmosphere during nighttime.
5. Conclusions
In order to reduce asphalt pavement temperature in the daytime
and decrease heat release into the atmosphere during nighttime, a
highly oriented heat conduction structure was proposed by estab-
lishing a gradient thermal conductivity in the layered pavement.
Numeral simulation and field test have confirmed that the designed
structure modified by adding 15%(top) + 10%(middle) + 0%(bottom)
dosages of floating beads, could significantly reduce the tempera-
ture of asphalt pavements. The maximum temperature differences
between the designed structure and contrast structure were 1.1 ◦ C,
2.3 ◦ C, 1.3 ◦ C (simulation result) and 1.6 ◦ C, 2.4 ◦ C, 0.8 ◦ C (test road
result) at the depth of 4 cm, 10 cm and 18 cm. The average temper-
atures of middle and bottom layers of the structure were 1.6 ◦ C (at
2:30 pm) and 1.5 ◦ C (at 6:00 pm) lower than those of the contrast
structure. Simulation also displayed that the structure released
2.42 × 104 J/m2 of heat during nighttime less than that of the con-
trast structure, and the reduction of heat release accounted for
12.1% of the total heat release of the contrast structure. The pro-
posed structure is expected to be used to decrease high surface
temperature of asphalt pavement during nighttime in summer.
Besides, maintaining a relatively low temperature is also conducive
to resisting rutting.
Acknowledgments
This study was supported by (1) National Science and Technol-
ogy Support Program (No. 2014BAG05B03) and (2) Opening Project
of Key Laboratory of Road Engineering for Cold Region of CCCC First
Highway Consultants Co., LTD (2013).
References
[1] G.K. Chang, J.N. Meegoda, Micromechanical model for temperature effects of
hot-mix asphalt concrete, Transp. Res. Rec.: J. Transp. Res. Board 1687 (1) (1999)
95–103.
[2] D. Van Thanh, C.P. Feng, Study on Marshall and rutting test of SMA at abnormally
high temperature, Constr. Build. Mater. 47 (2013) 1337–1341.
[3] I. Livada, M. Santamouris, K. Niachou, et al., Determination of places in the great
Athens area where the heat island effect is observed, Theor. Appl. Climatol. 71
(3/4) (2002) 219–230.
[4] A. Synnefa, A. Dandou, M. Santamouris, et al., On the use of cool materials as
a heat island mitigation strategy, J. Appl. Meteorol. Climatol. 47 (11) (2008)
2846–2856.
[5] J.T. Kevern, L. Haselbach, V.R. Schaefer, Hot weather comparative heat balances
in pervious concrete and impervious concrete pavement systems, J. Heat Island
Inst. Int. 7 (2) (2012) 231–236.
[6] D.S. Parker, S.F. Barkaszi Jr., Roof solar reflectance and cooling energy use: field
research results from Florida, Energy Build. 25 (2) (1997) 105–115.
[7] M. Santamouris, N. Papanikolaou, I. Livada, et al., On the impact of urban climate
on the energy consumption of buildings, Solar Energy 70 (3) (2001) 201–216.
[8] K.A. Gray, M.E. Finster, The Urban Heat Island, Photochemical Smog, and
Chicago: Local Features of the Problem and Solution, Northwestern University,
Department of Civil Engineering, 1999.
[9] Y. Zhou, J.M. Shepherd, Atlanta’s urban heat island under extreme heat condi-
tions and potential mitigation strategies, Nat. Hazards 52 (3) (2010) 639–668.
[10] M. Santamouris, N. Gaitani, A. Spanou, et al., Using cool paving materials to
improve microclimate of urban areas – design realization and results of the
flisvos project, Build. Environ. 53 (2012) 128–136.
195
[11] A. Synnefa, T. Karlessi, N. Gaitani, et al., Experimental testing of cool colored thin
layer asphalt and estimation of its potential to improve the urban microclimate,
Build. Environ. 46 (1) (2011) 38–44.
[12] T.P. Lin, Y.F. Ho, Y.S. Huang, Seasonal effect of pavement on outdoor ther-
mal environments in subtropical Taiwan, Build. Environ. 42 (12) (2007)
4124–4131.
[13] Y. Du, Q. Shi, S. Wang, Highly oriented heat-induced structure of asphalt pave-
ment for reducing pavement temperature, Energy Build 85 (2014) 23–31.
[14] Y. Du, Q. Shi, S. Wang, Bidirectional heat induced structure of asphalt pavement
for reducing pavement temperature, Appl. Therm. Eng. 75 (2015) 298–306.
[15] G. Wang, R. Roque, D. Morian, Effects of surface rutting on near-surface pave-
ment responses based on a two-dimensional axle-tire-pavement interaction
finite-element model, J. Mater. Civ. Eng. 24 (11) (2012) 1388–1395.
[16] A. Synnefa, M. Santamouris, K. Apostolakis, On the development: optical
properties and thermal performance of cool colored coatings for the urban
environment, Solar Energy 81 (4) (2007) 488–497.
[17] N. Tran, B. Powell, H. Marks, et al., Strategies for design and construction of high-
reflectance asphalt pavements, Transp. Res. Rec.: J. Transp. Res. Board 2098 (1)
(2009) 124–130.
[18] X. Cao, B. Tang, H. Zhu, et al., Cooling principle analyses and performance eval-
uation of heat-reflective coating for asphalt pavement, J. Mater. Civ. Eng. 23 (7)
(2010) 1067–1075.
[19] F. Rossi, B. Castellani, A. Presciutti, et al., Retroreflective facades for urban heat
island mitigation: experimental investigation and energy evaluations, Appl.
Energy 145 (2015) 8–20.
[20] F. Rossi, A.L. Pisello, A. Nicolini, et al., Analysis of retro-reflective surfaces for
urban heat island mitigation: a new analytical model, Appl. Energy 114 (2014)
621–631.
[21] T. Karlessi, M. Santamouris, K. Apostolakis, et al., Development and testing of
thermochromic coatings for buildings and urban structures, Solar Energy 83
(4) (2009) 538–551.
[22] J.Y. Hu, X. Yu, Experimental study of sustainable asphalt binder, Transp. Res.
Rec.: J. Transp. Res. Board 2372 (1) (2013) 108–115.
[23] B. Guan, B. Ma, F. Qin, Application of asphalt pavement with phase change
materials to mitigate urban heat island effect, in: Water Resource and Envi-
ronmental Protection (ISWREP), 2011 International Symposium on IEEE, vol. 3,
2011, pp. 2389–2392.
[24] T. Asaeda, V.T. Ca, Characteristics of permeable pavement during hot summer
weather and impact on the thermal environment, Build. Environ. 35 (4) (2000)
363–375.
[25] H. Li, J.T. Harvey, T.J. Holland, et al., The use of reflective and permeable
pavements as a potential practice for heat island mitigation and stormwater
management, Environ. Res. Lett. 8 (1) (2013) 015–023.
[26] R.B. Mallick, B.L. Chen, S. Bhowmick, Harvesting energy from asphalt pave-
ments and reducing the heat island effect, Int. J. Sustain. Eng. 2 (3) (2009)
214–228.
[27] V. Bobes-Jesus, P. Pascual-Muñoz, D. Castro-Fresno, et al., Asphalt solar collec-
tors: a literature review, Appl. Energy 102 (2013) 962–970.
[28] M. Iwama, T. Yoshinaka, N. Nemoto, Development of surface treatment technol-
ogy targeting temperature reduction of asphalt pavement, in: Road Engineering
Association of Asia and Austrlasia (REAAA) Conference, 12th, 2006, Philippines.
[29] T. Lucke, S. Beecham, Field investigation of clogging in a permeable pavement
system, Build. Res. Inf. 39 (6) (2011) 603–615.
[30] E. Coleri, J.T. Harvey, K. Yang, et al., Micromechanical investigation of open-
graded asphalt friction courses’ rutting mechanisms, Constr. Build. Mater. 44
(2013) 25–34.
[31] W.T. Van Bijsterveld, L.J.M. Houben, A. Scarpas, et al., Using pavement as solar
collector: effect on pavement temperature and structural response, Transp. Res.
Rec.: J. Transp. Res. Board 1778 (1) (2001) 140–148.
[32] S. Wang, Q. Zhu, Y. Duan, et al., Unidirectional heat-transfer asphalt pavement
for mitigating the urban heat island effect, J. Mater. Civ. Eng. 26 (5) (2014)
812–821.
[33] T. Asaeda, V.T. Ca, A. Wake, Heat storage of pavement and its effect on the lower
atmosphere, Atmos. Environ. 30 (3) (1996) 413–427.
[34] B. Givoni, Climate Considerations in Building and Urban Design, Van Nostrand
Reinhold, 1998.
[35] R. Watkins, J. Palmer, M. Kolokotroni, et al., The London heat island: results
from summertime monitoring, Build. Serv. Eng. Res. Technol. 23 (2) (2002)
97–106.
[36] M. Kolokotroni, I. Giannitsaris, R. Watkins, The effect of the London urban heat
island on building summer cooling demand and night ventilation strategies,
Solar Energy 80 (4) (2006) 383–392.
[37] D.M. Mrawira, J. Luca, Effect of aggregate type: gradation, and compaction
level on thermal properties of hot-mix asphalts, Can. J. Civ. Eng. 33 (11) (2006)
1410–1417.
[38] R.J. Cominsky, G.A. Huber, T.W. Kennedy, et al., The Superpave Mix Design Man-
ual for New Construction and Overlays. No. SHRP-A-407, Strategic Highway
Research Program, Washington, DC, 1994.
[39] X. Liu, S. Wu, Q. Ye, et al., Properties evaluation of asphalt-based compos-
ites with graphite and mine powders, Constr. Build. Mater. 22 (3) (2008)
121–126.
[40] Z. Yan, Analysis of the temperature field in layered pavement system, J. Tongji
Univ. 3 (1984) 76–85.
[41] Q. Xue, L. Liu, Y. Zhao, et al., Dynamic behavior of asphalt pavement structure
under temperature-stress coupled loading, Appl. Therm. Eng. 53 (1) (2013)
1–7.
196 D. Yinfei et al. / Energy and Buildings 102 (2015) 187–196

[42] Z. Suo, W.G. Wong, Nonlinear properties analysis on rutting behavior of bitu-
minous materials with different air void contents, Constr. Build. Mater. 23 (12)
(2009) 3492–3498.
[43] T. Xu, X. Huang, Investigation into causes of in-place rutting in asphalt pave-
ment, Constr. Build. Mater. 28 (1) (2012) 525–530.
[44] A.R. Archilla, S. Madanat, Development of a pavement rutting model from
experimental data, J. Transp. Eng. 126 (4) (2000) 291–299.
[45] R.A. Tarefder, M. Zaman, K. Hobson, A laboratory and statistical eval-
uation of factors affecting rutting, Int. J. Pavement Eng. 4 (1) (2003)
59–68.