This article was downloaded by: [University of Nebraska, Lincoln]

On: 31 August 2015, At: 10:13

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: 5 Howick Place,
London, SW1P 1WG

International Journal of Pavement Engineering

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/gpav20

Effects of mean annual temperature and mean annual

precipitation on the performance of flexible pavement
using ME design
a a a
Mohd Rosli Mohd Hasan , Jacob E. Hiller & Zhanping You
a
Department of Civil and Environmental Engineering, Michigan Technological University,
1400 Townsend Drive, Houghton, MI 49931, USA
Published online: 16 Mar 2015.

Click for updates

To cite this article: Mohd Rosli Mohd Hasan, Jacob E. Hiller & Zhanping You (2015): Effects of mean annual temperature
and mean annual precipitation on the performance of flexible pavement using ME design, International Journal of Pavement
Engineering, DOI: 10.1080/10298436.2015.1019504

To link to this article: http://dx.doi.org/10.1080/10298436.2015.1019504

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 International Journal of Pavement Engineering, 2015
http://dx.doi.org/10.1080/10298436.2015.1019504

Effects of mean annual temperature and mean annual precipitation on the performance
of flexible pavement using ME design
Mohd Rosli Mohd Hasan1, Jacob E. Hiller2 and Zhanping You*
Department of Civil and Environmental Engineering, Michigan Technological University, 1400 Townsend Drive, Houghton,
MI 49931, USA
(Received 21 January 2015; accepted 31 January 2015)

The purposes of this study were to establish the difference between empirical and mechanistic – empirical approaches in the
flexible pavement design and to quantify the effects of mean annual precipitation and temperature on the flexible pavement
distresses using the Mechanistic-Empirical Pavement Design Guide (MEPDG) software. Seventy-six specific locations from
13 states throughout the USA were selected based on different climate conditions using virtual climate stations based on the
interpolation from the nearest weather stations prior to meeting the objectives. Subsequently, analysis was conducted based
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

on the predicted distresses, including longitudinal cracking, transverse cracking, alligator cracking, asphalt concrete rutting
and total pavement permanent deformation. Generally, the pavement structure and materials have been set as constant to
control the effects of material on the results. On the basis of the MEPDG analysis, the longitudinal cracking of flexible
pavement is significantly affected by both factors (temperature and precipitation), particularly in wet climatic regions.
The mean annual temperature has a great influence on the alligator cracking, transverse cracking and permanent deformation
of flexible pavement. However, neither factors demonstrated a significant impact on the predicted International Roughness
Index of flexible pavement surfaces.
Keywords: mechanistic – empirical; environment condition; climate; asphaltic concrete; pavement distress

1. Introduction moisture and freeze –thaw damage, drain ability of paving

The development of pavement design has greatly changed layers and infiltration potential of the pavement have been
from an observation-based approach (empirical) to a integrated with its physical condition, and can affect the
mechanistic – empirical approach, which offers more design outcome (NCHRP 2004a).
rational and realistic design procedures. Advancement in In Sections 2 and 3 will discuss the evolution of
computational technology has improved engineers’ ability pavement design and the differences between empirical
to incorporate traffic characteristics and climate impacts to and mechanistic –empirical approaches in flexible pave-
predict the pavement response throughout the designated ment design, based on the American Association of State
service life based on the wide range of climatic factors, Highway and Transportation Officials (AASHTO) 1993
materials, traffic considerations and mathematical com- and MEPDG guidelines. Then, this paper will explore the
putations (National Cooperative Highway Research effects of climate (mean annual precipitation and
Program [NCHRP] 2004a, Schwartz and Carvalho 2007). temperature) on the flexible pavement-predicted outputs
As part of the transportation agency’s efforts to using MEPDG software.
improve pavement design, mechanistic –empirical-based
software, named the Mechanistic-Empirical Pavement
Design Guide (MEPDG), has been developed from the 2. Empirical pavement design
NCHRP project 1-37A. One of the major concerns of this Under empirical methodology, pavement design basically
prediction design software was to take into consideration consists of determining required thickness of the pavement
the site-specific environmental or climate conditions in the layers to protect the subgrade layer and support the loads
design outputs, which have not been done extensively in from travelling vehicles based on previous experience and
previous design methodologies. The climate condition was observations of field performance (Schwartz and Carvalho
found to have a significant contribution to the MEPDG 2007). On the basis of the literature review (Hallin et al.
design of flexible pavement, including precipitation, 2007, Schwartz and Carvalho 2007), the AASHO road
temperature, water table depth and freeze – thaw cycles. tests were constructed in August 1956 and the test was
On the other hand, it was also stated that internal factors ended on 30 November 1960 to provide an empirical
such as the susceptibility of the pavement materials, database to be used as pavement design guides (Smith

*Corresponding author. Email: zyou@mtu.edu

q 2015 Taylor & Francis
 2 M.R. Mohd Hasan et al.
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

Figure 1. MEPDG design process.

et al. 2004). However, the laboratory test data and test the AASHTO methods are considered as empirical, the
track experiments are only valid for the materials, climate placement of soil support value by the subgrade resilient
conditions and subgrade properties for which they were modulus and layer coefficients by the resilient modulus of
constructed (Huang 2004, Schwartz and Carvalho 2007), each layer show that these methods trend towards
thereby limiting the applicability of the observed results. mechanistic design approach.
The earliest empirical methods in designing flexible
pavement were begun in the mid-1920s with the
development of a soil classification system. In 1929, 3. Mechanistic – empirical approach
the California Bearing Ratio (CBR) test was developed by The MEPDG is an emerging prediction software that is
the California Highway Department to quantify the used to analyse and design the pavements based on the
required thickness of pavement to provide protection designated inputs, including materials, traffic, climate and
against shear failure of the subgrade. This method was pavement structure (Figure 1) to predict the pavement
further developed by the US Corps of Engineers during the performance over the design life (Baus and Stires 2010,
Second World War and became a very popular design Diefenderfer 2010). The MEPDG software quantifies
method during the era after the war. In addition, the Public pavement responses such as stresses and strains, which are
Roads soil classification system was further modified in used to estimate the pavement distresses and ride quality
1945 by the Highway Research Board, which grouped and over the designated service life (Baus and Stires 2010).
indexed very specific soil types. This classification system This software also comprises two sets of parameters for
was adopted to determine the quality of subgrade and to local calibration of most performance transfer functions
estimate the adequate thickness of pavement structure (Williams and Shaidur 2013). Timm et al. (2010)
(Huang 2004, Schwartz and Carvalho 2007). mentioned that the MEPDG was developed with a broad
According to previous studies (Hallin et al. 2007, climate database based on the data collected from 851
Hudson et al. 2007, Mahoney et al. 2007, Schwartz and (Johanneck and Khazanovich 2010) and 232 (Saha and
Carvalho 2007), major findings from the AASHO road test Bayat 2011) weather stations throughout the USA and
are currently still widely used and have become the basis for Canada, respectively. The designer would be able to select
designing flexible pavements across the USA and in many a specific location or interpolate from the available nearest
countries of the world. The major findings include the use of weather stations. Moreover, the predicted distresses were
the equivalent single-axle loads as a single traffic parameter also calculated based on a mechanistic model and an
to account for both traffic volume and load levels, the empirical damage model that runs contrary to the
serviceability – performance concept, structural number empirically based AASHTO 1993 pavement design
and effects of layer thickness and strength. These findings guide (Yang and Wu 2012).
can be found in the AASHTO method published in 1986 and Yet, Yang and Wu (2012) suggested to state
modified for overlay designs in 1993 (Coree and White departments of transportation that they should investigate
1990, AASHTO 1993). In this design method, the impact of and calibrate the damage models based on the local
different climates on pavement performance is not condition before implementing the MEPDG for a
considered. Huang (2004) acknowledged that, although particular state due to the varieties of climate, pavement
 International Journal of Pavement Engineering 3

structure and materials throughout the USA. Aguiar-Moya problems in the AC surface layer (NCHRP 2004a), which
and Prozzi (2011) argued that the variability needs to be is not captured in the MEPDG analysis directly.
properly accounted for, and not necessarily be controlled The Enhanced Integrated Climate Model (EICM) is a
in the pavement mechanistic – empirical design. one-dimensional heat and moisture flow programme that
Table 1 shows the hierarchical approach that has been simulates changes in each layer of pavement structure as
developed in MEPDG. The different levels are defined by regards to the specific climate conditions over the
the quality of data available and importance level of the pavement’s design life. In the MEPDG guidelines
project. On the basis of the NCHRP 1-37A report, Level 1 is (NCHRP 2004b), it was indicated that the temperature
recommended for heavy traffic highways, followed by profile, changes in the depth of the water table, precipitation,
Level 2 intermediate projects and Level 3 is typically used freeze and thaw cycles, and other related factors are
for low-traffic roads. However, Level 3 was found to be the modelled in MEPDG in a very comprehensive manner. The
most commonly used by highway state agencies, especially EICM is completely linked to the MEPDG and the internal
for pavement management programmes (Schwartz and pavement responses/distress calculations performed are
Carvalho 2007). On the basis of Chapter 3 of Part 3 of based on the input parameters (general information, weather-
Mechanistic – Empirical Design guide (NCHRP 2004c), the related information, ground water-related information,
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

input-level selection for a parameter typically depends on drainage and surface properties, and pavement structure
certain factors, including criticality of a project, sensitivity and materials) that feed through provided interfaces in the
of the pavement performance, availability of information software (Ghalia et al. 2011). Li et al. (2011) further affirmed
during the design process and resources to obtain the inputs. that the general information, including traffic opening time,
While the reliability of data entered into the programme construction dates and pavement structure, is required to
changes by level, the actual calculations for pavement generate the moisture model in the EICM. Johanneck and
responses and the transfer functions to arrive at the Khazanovich (2010) cited that the EICM consists of three
predicted distress levels do not change. important components: the climate–materials structural
model, frost heave and settlement model (Cold Regions
Research and Engineering Laboratories [CRREL] model)
and infiltration–drainage model.
3.1. Climate-related impact in the MEPDG analysis The EICM is an updated version of the Integrated
As a visco-elastic material, the performance of asphalt is Climate Model (ICM) Version 2.0 under the NCHRP 1-37A
significantly influenced by temperature variation. The study in 1997. Before that, the earliest version of ICM was
modulus of asphalt concrete (AC) can vary from 100,000 designed at Texas A&M University for the Federal
psi during hot summer days to about 2,000,000 – Highway Administration (FHWA) in 1989 (Lytton et al.
3,000,000 psi during winter conditions. In addition, the 1993, Larson and Dempsey 1997). Zaghloul et al. (2006)
resilient modulus of soil could be increased up to 120 times mentioned that EICM is a powerful tool in the MEPDG
higher than the value before the freezing condition. software to model and predict the specific location
However, unbound materials are not influenced by temperature, moisture content, resilient adjustment factor,
temperature, except when ice forms below 32oF. The pore water pressure, frost and thaw depth and frost heave
strength of an asphalt pavement structure is greatly reduced throughout the pavement structure profile including the hot
when thawing occurs, especially if the pavement is mix asphalt (HMA) layer, base, sub-base and subgrade.
constructed on a subgrade that is having frost heave issues. Johanneck and Khazanovich (2010) stated that the EICM
The reduction of strength caused by these issues can lead to database is based on the weather stations that are well
premature pavement distresses resulting in failure. On the distributed throughout the USA. Brown et al. (2006)
other hand, excessive moisture also can lead a moisture- affirmed that the EICM offers better outputs compared to
induced damage problem that may cause stripping the previous procedure. The application of the EICM allows
the prediction of moisture and temperature conditions at all
depth in the pavement system, which strongly influenced its
Table 1. Levels in MEPDG hierarchical approach.
mechanical behaviour and the development of distresses
No. of level Descriptions throughout the pavement life.
In the MEPDG, the EICM is used to predict the localised
Level 1 Laboratory evaluation on the material properties
(dynamic modulus) and traffic data (vehicle environmental impacts on the pavement design. There are
class and load distributions) are required several essential factors that need to be taken into
Level 2 Inputs are obtained from empirical correlations consideration, including hourly air temperature, precipi-
with other parameters (resilient modulus based tation, wind speed, percentage sunshine and relative
on CBR value) humidity. These parameters can be obtained either from the
Level 3 Inputs are based on the default values from
the national database specified weather stations throughout the USA or through
mathematical computations based on related factors. For
 4 M.R. Mohd Hasan et al.

example, the number of freeze–thaw cycles is determined 4. Methodology and selected parameters
based on the temperature data within the analysis time period In this investigation, the climate conditions for 76
(NCHRP 2004a). In addition, two other environmental locations across 13 US states from the Long-Term
condition parameters that give significant impact in the Pavement Performance (LTPP) programme were charac-
pavement analysis are ground water table depth and drainage terised by mean annual precipitation and temperature of
properties (Schwartz and Carvalho 2007) which the EICM the site-specific locations based on weather data collected
takes into consideration. The climate-related impact con- from 1981 to 2010 for the National Oceanic and
sidered in the MEPDG design is shown in Figure 2. On the Atmospheric Administration’s National Climatic Data
basis of the EICM, the air temperature is essential to determine Center (2011). While grouping the selected regions,
the long-wave radiation emitted by the air and the convective locations that had more than 42 in. of rainfall per year
heat transfer from surface to air. In addition, the air were clustered as wet conditions, and locations that had a
temperature is also used to establish the duration of a mean annual temperature of less than 48oF were included
freeze–thaw cycle, and to determine the number of freeze– in the freeze region (Hall et al. 2002). Prior to taking into
thaw cycles. Meanwhile, the precipitation data are required to consideration both effects in this study, the selected
characterise the infiltration and ageing process. However, the locations were divided into four different regions that are
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

recorded precipitation during months when the mean freeze-dry, no freeze-dry, freeze-wet and no freeze-wet as
temperature is less than the freezing temperature of water is shown in Table 2. Meanwhile, the water table was
considered to fall as snow. The percentage of sunshine and assumed to be constant at 8 ft, referring to the median
wind speed is needed to use in computing heat balance and the height of the water table throughout the USA based on the
convection heat transfer coefficient at the pavement surface, LTPP database (Aguiar-Moya and Prozzi 2011). The
respectively. Even though the relative humidity is not directly selected locations used in these analyses are as marked in
incorporated in the EICM for analysing flexible pavements, it Figure 3. Subsequently, the MEPDG analyses were
could also affect the long-term performance of pavements. conducted based on the interpolation from the nearest
Another factor that gives significant impact on the pavement weather stations (available in MEPDG software) to the
performance is the ground water table. This parameter is selected locations.
important to incorporate because it can affect the overall Using the MEPDG, the pavement section was designed
accuracy of the foundation and sub-base’s moisture contents as a four-lane highway, with two lanes in each direction.
and equilibrium modulus values (NCHRP 2004a). The design life of the selected pavement was specified to

Precipitation

Sunshine Air Temperature

Serviceability

Relative Humidity

Wind Speed

Design Life

Ground Water Table Freeze and Thaw Cycles

Figure 2. Climate-related impacts in the MEPDG prediction analysis.

 International Journal of Pavement Engineering 5

Table 2. Characterisation of the regions involved in this study The Pearson correlation analysis results are represented by
(Hall et al. 2002). the correlation coefficient and p-values as shown in
Criteria Table 4. The correlation coefficient (r) is a numerical
summary of a bivariate relationship that ranges from
Mean precipitation 2 1.00 to 1.00. Generally, a positive value indicates as a
Region’s climate Mean temperature (in rainfall/year) positive relationship between two variables while a
Freeze – wet ,488F or 8.98C . 42 negative value indicates an inverse relationship between
Freeze – dry ,488F or 8.98C , 21 variables. The r value is typically divided into five scales
No freeze – wet .648F or 17.88C . 42 depending on the level of correlation ranging from very
No freeze – dry .648F or 17.88C , 21
weak to very strong, where:

20 years. Construction of the pavement section was begun . 0.8 –1.0 (very strong);
in June 2013 and open to traffic in August 2013. The . 0.6 –0.8 (strong);
average annual daily traffic was specified to 30,000 . 0.4 –0.6 (moderate);
.
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

vehicles, with 5% growth for the first year and 1.5% 0.2 –0.4 (weak) and
growth every year after. Besides that, roughly 12% truck . , 0.2 (very weak).
traffic has been designed on this section of roadway and As shown in Table 4, there is a very strong correlation
the vehicle class distribution is truck traffic classification between the mean annual precipitation and the longitudi-
(TTC) 1. The initial international roughness index (IRI) nal cracking distress as r ¼ 0.835. The positive value
value was specified at 63 in./mile. The reliability level shows that a higher amount of mean annual precipitation
and operational speed were designed at 90% and 55 mph, will result in higher longitudinal cracking distress. It is
respectively. Meanwhile, the lane factor was set at 95% also found that mean annual precipitation only has a
and 50% for the directional distribution factor. The moderate relationship to the pavement deterioration due to
details of the four layers of pavement structure are alligator cracking distress and weak correlation for the rest
specified in Table 3. of the pavement distresses (r values ranging from 0.2 to
0.4). Meanwhile, the mean annual temperature might be
considered as the main mechanism to the most of the
flexible pavement distress that is signified by the
5. Results and discussion
coefficient value within the range of strong (0.6 –0.8)
5.1. Correlation analysis between climate variables and and very strong (0.8 –1.0) correlation, with the exception
pavement distresses of the longitudinal cracking distress with the moderate
The relationship between the climate factors (mean annual relationship. Furthermore, the mean annual temperature
precipitation and temperature) and the pavement distresses inversely affects the transverse cracking that is indicated
has been analysed using Pearson’s correlation analysis. by the negative value (2 0.907). Theoretically, the Pearson

Figure 3. Locations involved in the MEPDG analysis.

 6 M.R. Mohd Hasan et al.

Table 3. Selected design for the flexible pavement. Table 5. One-way ANOVA effects of precipitation on
pavement distress.
Pavement layer Material Thickness (in.)
Source DF SS MS F p
Road surface AC (asphalt binder: PG 58-16) 6
a
Base Crushed stone 8 Effects of precipitation on longitudinal cracking
Sub-base Crushed gravel 8 Precipitation 71 52,851,511 744,387 8.08 0.027
Subgrade A-4 – Error 4 368,400 92,100
Total 75 53,219,911
Effects of precipitation on alligator crackingb
Precipitation 71 345.51 4.87 1.34 0.436
Error 4 14.51 3.63
Table 4. Pearson correlation analysis results. Total 75 360.02
Effects of precipitation on transverse crackingc
Factors Precipitation 71 68,609,340 966,329 2.96 0.148
Error 4 1,306,051 326,513
Correlation Precipitation Temperature Total 75 69,915,391
Pavement Longitudinal cracking 0.835* 0.528* Effects of precipitation on AC ruttingd
Precipitation 71 11.562 0.163 0.96 0.611
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

distresses (top-down cracking) , 0.001** ,0.001**

Alligator cracking 0.590* 0.743* Error 4 0.682 0.170
(bottom-up cracking) , 0.001** ,0.001** Total 75 12.244
Transverse cracking 2 0.233* 20.907* Effects of precipitation on total pavement deformatione
0.043** ,0.001** Precipitation 71 10.607 0.149 0.87 0.661
AC rutting 2 0.293* 0.721* Error 4 0.689 0.172
0.010** ,0.001** Total 75 11.296
Total pavement 2 0.298* 0.702* Effects of precipitation on IRIf
deformation 0.009* ,0.001** Precipitation 69 8636 125.2 0.49 0.928
Error 6 1533 255.5
*Correlation coefficient (r). Total 75 10,169
**p-Value. a
SD ¼ 303.5, R 2 ¼ 99.31%, R 2 (adj) ¼ 87.02%.
b
SD ¼ 1.905, R 2 ¼ 95.97%, R 2 (adj) ¼ 24.41%.
c
SD ¼ 571.4, R 2 ¼ 98.13%, R 2 (adj) ¼ 64.97%.
correlation values obtained from this study are valid to use d
SD ¼ 0.4129, R 2 ¼ 94.43%, R 2 (adj) ¼ 0.00%.
based on values of p , 0.05 (95% confidence interval). e
SD ¼ 0.4149, R 2 ¼ 93.90%, R 2 (adj) ¼ 0.00%.
f
SD ¼ 15.9842, R 2 ¼ 84.93%, R 2 (adj) ¼ 0.00%.

mean annual precipitation of the selected locations. This can

5.2. Effects of mean annual precipitation on pavement
be related to the nature of top-down cracking that permits
distresses
water infiltration into the underlying pavement layers that
The data obtained from the MEPDG analysis were will cause the structural layer to deteriorate, either in the
statistically interpreted using one-way analysis of variance freeze–wet or no freeze–wet regions. However, the mean
(ANOVA) to determine the effects of mean annual annual precipitation has no significant effects on the
precipitation on the pavement distresses as shown in pavement distress related to alligator cracking, transverse
Table 5. In general, ANOVA is an analysis method that is cracking, AC rutting, total pavement permanent deformation
used to test the null hypothesis where the mean of three or and IRI, which are based on values of p . 0.05.
more populations is equal. One-way ANOVA is performed
based on the three assumptions, where:
. the populations of samples are approximately 5.3. Effects of mean annual temperature on pavement
normally distributed; distresses
. the populations of samples have the same variance
The effects of mean annual temperature in the selected
and
locations on the flexible pavement distresses were also
. the samples from different populations are random
analysed using one-way ANOVA at a 95% confidence
and independent.
interval (a ¼ 0.05). The results of the analysis are shown
The data were analysed at a 95% confidence interval in Table 6. Generally, the mean annual temperature has
(a ¼ 0.05). On the basis of the results, it can be concluded significant effects on all of the flexible pavement
that the precipitation has significant effects on longitudinal distresses. This was statically confirmed by the results
cracking distress. The effect of mean annual precipitation on with a p-value of , 0.05, as tabulated in Table 6. It is clear
longitudinal joint is graphically presented in a scatter plot in from this analysis that flexible pavement is very sensitive
Figure 4. The predicted longitudinal cracking (top-down to temperature and time (Yildirim 2007, Peters et al.
cracking) progressively increases with the increment of 2010). The modulus of asphalt pavement can be differed
 International Journal of Pavement Engineering 7

6000
Region
Freeze-Wet
Freeze-Dry
5000

Longitudinal Cracking (ft/mi)

No Freeze-Wet
No Freeze-Dry

4000

3000

2000

1000
0 20 40 60 80 100 120 140
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

Mean Annual Precipitation (in)

Figure 4. Longitudinal cracking predicted distresses versus mean annual precipitation.

throughout the year, which could be between 2,000,000 Figures 5 – 8 present the effects of mean annual
and 3,000,000 psi during cold winter conditions and precipitation on the flexible pavement distresses. It was
100,000 psi during hot summer days (NCHRP 2004a). found that a higher mean annual temperature results in higher
However, the mean annual temperature has not resulted in longitudinal cracking, alligator cracking and both AC rutting
significant impact on the IRI of pavement surface, which is and total pavement permanent deformation. However, from
based on p-value approximately 0.415 (confidence interval Figure 6, transverse cracking is more pronounced for the
, 95%). pavement placed under the low-temperature climate region.
On the basis of a study, the cracks formed when the
Table 6. One-way ANOVA effects of temperature on pavement
temperature dropped instantly, which caused thermal-
distress.
induced shrinkage stress in the asphalt pavement layer to
Source DF SS MS F p exceed the tensile strength of the asphalt pavement (Das et al.
Effects of temperature on longitudinal cracking a 2012). The total pavement permanent deformation is higher
Temperature 27 34,171,147 1,265,598 3.19 ,0.001 compared to AC rutting due to the accumulated deflection
Error 48 19,048,764 396,849 that occurs in each layer of the pavement structure.
Total 75 53,219,911 Deformation occurs on the asphalt pavement as a result of
Effects of temperature on alligator crackingb long-term exposure to certain levels of stress that are below
Temperature 27 292.06 10.82 7.64 ,0.001
Error 48 67.97 1.42
the yield strength of the material. The permanent
Total 75 360.02 deformation is more severe when subjected to heat for an
Effects of temperature on transverse crackingc extended duration and asphalt pavement will deform upon
Temperature 27 65,006,733 2,407,657 23.54 ,0.001 approaching its softening point (Evans and Wilshire 1993).
Error 48 4,908,658 102,264
Total 75 69,915,391
Effects of temperature on AC ruttingd
Temperature 27 8.1223 0.3008 3.50 ,0.001 5.4. Comparison of pavement distresses in different
Error 48 4.1218 0.0859
Total 75 12.2441
climates
Effects of temperature on total pavement deformatione Figures 9 and 10 show the predicted pavement distress at
Temperature 27 7.3069 0.2706 3.26 ,0.001 different climates that resulted from the MEPDG analysis.
Error 48 3.9890 0.0821 From Figure 9(a), longitudinal cracking (top-down
Total 75 11.2959
Effects of temperature on IRIf cracking) is predicted to be severely reoccurring in the
Temperature 27 3808 141.0 1.06 0.415 wet climate region. The top-down fatigue cracking is
Error 48 6362 132.5 critical in the wet region because it allows water to
Total 75 10,169 infiltrate the underneath pavement layers that may cause
a
SD ¼ 630.0, R 2 ¼ 64.21%, R 2 (adj) ¼ 44.07%. failure of the pavement structure (NCHRP 2004c). On the
b
SD ¼ 1.190, R 2 ¼ 81.12%, R 2 (adj) ¼ 70.50%. other hand, the alligator cracking (bottom-up cracking) is
c
SD ¼ 319.8, R 2 ¼ 92.98%, R 2 (adj) ¼ 89.03%. more prominent at the high-temperature climate region as
d
SD ¼ 0.2930, R 2 ¼ 66.34%, R 2 (adj) ¼ 47.40%.
e
SD ¼ 0.2883, R 2 ¼ 64.69%, R 2 (adj) ¼ 44.82%. shown in Figure 9(b). These cracks start at the bottom
f
SD ¼ 11.5123, R 2 ¼ 37.44%, R 2 (adj) ¼ 2.26%. pavement layer and progress to the pavement surface due to
 8 M.R. Mohd Hasan et al.

4000
Region
Freeze-Wet
3500 Freeze-Dry
No Freeze-Wet

Longitudinal Cracking (ft/mi)

No Freeze-Dry

3000

2500

2000

1500

1000

0 5 10 15 20 25
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

Mean Annual Temperature (C)

Figure 5. Longitudinal cracking versus mean annual temperature.

HMA layer that deflects under wheel loads (NCHRP properties of asphalt binder that is used as a bonding agent
2004c). Obviously, the highest longitudinal and alligator in flexible pavement allows the material to shift slowly or
cracking is predicted to occur in the no freeze –wet climate. deform permanently under the influence of continual
This is due to the condition of asphalt that is highly aged, stresses. The severity of permanent deformation also can be
which is the result of exposure to high temperature and the affected by insufficient compaction during construction,
presence of moisture that weakened the base or sub-base poor mix design and increase in the axle load (Moghaddam
layer of the pavement. However, Figure 10(a) shows that et al. 2011).
the resistance to transverse cracking (thermal cracking) A well-designed AC is essential to ensure that the
must be taken into consideration for pavement design in the pavement performs well with desirable properties from the
freeze climate region. The thermal cracking, also known as viewpoint of performance adequacy, longer service life and
low temperature cracking, commonly occurs due to a better resistance to damages induced by traffic loads and the
sudden drop in temperature, which causes extreme thermal severe weather conditions. The performance of asphalt
contraction and fracture of the AC surface (NCHRP 2004c). mixtures is mainly affected by the temperature and the
Finally, the highest predicted permanent deformation of presence of moisture in the mixes. As a visco-elastic
both AC and total pavement would occur in the no freeze material, the viscous properties give bitumen its ability to
(hot) climate region, especially during summer as shown in flow at high temperatures with also the ability of a long
Figure 10(b). In the case of AC rutting, the viscous loading time while the elastic properties cause the bitumen

17.5
Region
Freeze-Wet
Freeze-Dry
15.0 No Freeze-Wet
Alligator Cracking (%)

No Freeze-Dry

12.5

10.0

7.5

5.0
0 5 10 15 20 25
Mean Annual Temperature (C)

Figure 6. Alligator cracking versus mean annual temperature.

 International Journal of Pavement Engineering 9

3000
Region
Freeze-Wet
2500 Freeze-Dry
No Freeze-Wet

Tranverse Cracking (ft/mi)

No Freeze-Dry
2000

1500

1000

500

0 5 10 15 20 25
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

Mean Annual Temperature (C)

Figure 7. Transverse cracking versus mean annual precipitation.

to become brittle at low temperatures and also present the pavement in different climate conditions, which mainly
possibility of a short loading time. When the bitumen focuses on the mean annual precipitation and temperature.
becomes brittle, AC becomes less resistant to abrasive Seventy-six locations have been selected from 13 different
forces and this leads to pavement deterioration. On the other states throughout the USA including California, Texas,
hand, in the hot climate region and during the hot summer Arizona, Louisiana, Mississippi, Hawaii, Montana, Idaho,
days, the viscosity of asphalt binder tends to reduce and will Wyoming, North Dakota, Illinois, New York and New
slowly generate permanent deformation, which is also Hampshire. Those states were categorised into four
known as rutting. Hence, changes in the temperature of the different climates, which are freeze-wet, freeze-dry, no
asphalt mix will influence the behaviour of the asphalt freeze-wet and no freeze-dry. Subsequently, the effects of
binder, which in turn directly affects the properties of the temperature and precipitation were evaluated based on the
asphalt mixture. predicted distresses using MEPDG analysis, known as
longitudinal cracking (top-down cracking), alligator
cracking (bottom-up cracking), transverse cracking, AC
6. Conclusion rutting and total pavement permanent deformation.
In this study, a mechanistic – empirical approach has been On the basis of the MEPDG analysis results, several
used to evaluate the performance of designated flexible conclusions can be made:

2.5 Regression Line

Asphalt Concrete Layer (AC)
Total Pavement (TP)
Legend Region
2.0
Permanent Deformation (in)

AC Freeze-Wet
AC Freeze-Dry
AC No Freeze-Wet
1.5 AC No Freeze-Dry
TP Freeze-Wet
TP Freeze-Dry
TP No Freeze-Wet
1.0 TP No Freeze-Dry

0.5

0.0

0 5 10 15 20 25
Mean Annual Temperature (C)

Figure 8. Permanent deformation versus mean annual precipitation.

 10 M.R. Mohd Hasan et al.

(a) (b)
4000 14

Alligator Cracking (%)

Longitudinal Cracking
3500 12
3000 10
2500

(ft/mi)
8
2000

3247.27

12.07
6

2743.00

10.40
9.88
1500

1997.62

7.18
4

1356.52
1000
500 2
0 0
et ry et ry et ry et ry
-W e-D -W e-D -W e-D -W -D
ez
e ez ez
e ez ez
e ez ez
e ze
F re F re F re Fre F re F re F re F ree
No No No No
Climate Climate

Figure 9. Predicted distresses of flexible pavement in different climate regions: (a) longitudinal cracking and (b) alligator cracking.
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

(1) The longitudinal cracking (top-down cracking) of Overall, this paper concludes that the mean annual
flexible pavement is significantly affected by precipitation and temperature were found to have varying
mean annual temperature and precipitation. levels of impact on the flexible pavement performance and
(2) The mean annual temperature has great effects on distresses in the selected regions. The findings from this
alligator cracking, transverse cracking and perma- study would be very beneficial for pavement engineers
nent deformation. designing flexible pavements to understand how climate
(3) The total pavement permanent deformation is can effect pavement distress development and promote
higher compared to AC rutting due to accumulated better understanding the failure mechanisms that relate to
deflection occurring in each layer of the pavement the temperature and precipitation. Referring to the LTPP
structure. programme data analysis (Hall et al. 2002; United-States-
(4) The alligator cracking (bottom-up cracking) is Federal-Highway-Administration 2004), the outcomes of
more prominent in the high-temperature climate this study are found to have good agreement with the LTPP
region, especially in regions with high precipi- programme findings as the climatic factors including
tation. The highest longitudinal and alligator temperature and precipitation have a significant effect on
cracking is predicted to occur in the no freeze – the fatigue cracking of AC. The seasonal variation and
wet climate that can be due to the state of asphalt distribution of temperature are among the factors that
binders that are highly aged, which is the result of influenced the development of alligator cracking, which
the exposure to high temperature and presence of were initiated by the longitudinal cracks in the wheel
moisture that weakened the base or sub-base layer paths. Other factors that resulted in fatigue cracking
of the pavement. include traffic loadings and the structural design of the
(5) As expected, the highest predicted permanent pavement (Hall et al. 2001). In addition, average annual
deformation of flexible pavement will occur in the precipitation was found to be significantly correlated with
hot climate region and is more severe during the the rate of rutting of flexible pavement. Precipitation was
hot summer days. found to have a critical influence on the permanent

(a) (b)
2500 2.0
Permanent Deformation

AC Rutting
Tranverse Cracking

2000 Total Pavement Deformation

1.5
1500
(ft/mi)

(in)

1.0
1,883.63

1.65

1000
1,437.36

1.23
1.16
0.90

0.89

500 0.5
0.71
1.00

1.00

0.42

0.41

0 0.0
et ry et ry et ry et ry
z e-W z e-D ze-W z e-D z e-W ze
-D
z e-W ze
-D
ree ree ree ree ree Fr
e e e e
Fr
e e
F F F F F Fr
No No No No
Climate Climate

Figure 10. Predicted distresses of flexible pavement in different climate regions: (a) transverse cracking (b) permanent deformation for
AC and total pavement.
 International Journal of Pavement Engineering 11

deformation of the pavement. However, an increase in Larson, G. and Dempsey, B.J., 1997. Enhanced integrated
mean annual precipitation may not be the only factor climatic model, version 2.0. Urbana, IL: Department of Civil
Engineering, University of Illinois at Urbana-Champaign.
linked to the growth of rut depths (United-States-Federal-
Li, Q., Mills, L., and McNeil, S. The implications of climate
Highway-Administration 2004). change on pavement performance and design 2011.
Washington, DC: National Transportation Library, United
States Department of Transportation.
Lytton, R., et al., 1993. An integrated model of the climatic
effects on pavements. Washington, DC: Transportation
Notes
Research Board, 304 p.
1. Email: mmohdhas@mtu.edu Mahoney, J.P., et al., 2007. Pavement lessons from the 50-year-old
2. Email: jhiller@mtu.edu interstate highway system: California, Oregon, and Washing-
ton. Transportation Research E-Circular (E-C118). Washing-
ton, DC: Transportation Research Board, pp. 88–103.
Moghaddam, T.B., Karim, M.R., and Abdelaziz, M., 2011. A
review on fatigue and rutting performance of asphalt mixes.
References Scientific Research and Essays, 6 (4), 670–682.
Aguiar-Moya, J.P. and Prozzi, J.A., 2011. Effect of field National Cooperation Highway Research Program (NCHRP),
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015

variability of design inputs on the MEPDG. In: 90th annual 1993. AASHTO guide for design of pavement structures.
meeting of the Transportation Research Board. Washington, Washington, DC: AASHTO.
DC: Transportation Research Board, 15 p. National Climatic Data Center, 2011, Jan 5. NOAA’s 1981– 2010
Baus, R. and Stires, N., 2010. Mechanistic-empirical pavement climate normals. [Accessed 15 Mar 2013]. http://www.ncdc.
design guide implementation. Report No: FHWA-SC-10-01. noaa.gov/data-access/land-based-station-data/land-based-
Brown, S.F., et al., 2006. Independent review of the ‘Mechan- datasets/climate-normals/1981-2010-normals-data
istic-Empirical Pavement Design Guide’ and software. NCHRP, 2004a. Guide for mechanistic-empirical design of new
NCHRP research results digest (307). Washington, DC: and rehabilited pavement structures, Part 2: design inputs,
Transportation Research Board. Chapter 3: environmental effects. Washington, DC: NCHRP,
Coree, B.J. and White, T.D., 1990. AASHTO flexible pavement Transportation Research Board, National Research Council.
design method: fact or fiction? Design and evaluation of NCHRP, 2004b. Guide for mechanistic empirical design of new
rigid and flexible pavements. Washington, DC: Transpor- and rehabilitated pavement structures. Washington, DC:
tation Research Board, pp. 206– 216. NCHRP.
Das, P.K., Tasdemir, Y., and Birgisson, B., 2012. Low temperature NCHRP, 2004c. Guide for mechanistic empirical design of new
cracking performance of WMA with the use of the Superpave and rehabilitated pavement structures, Part 3: design
indirect tensile test. Construction and Building Materials, 30, analysis, Chapter 3: design of new and reconstructed flexible
643–649. doi:10.1016/j.conbuildmat.2011.12.013. pavements. Washington, DC: NCHRP.
Diefenderfer, S.D., 2010. Analysis of the mechanistic-empirical Peters, S.J., et al., 2010. Nanocellulose and microcellulose fibers
pavement design guide performance predictions: influence of for concrete. Transportation Research Record: Journal of the
asphalt material input properties. Report No: FHWA/VTRC Transportation Research Board, 2142, 25 – 28. doi:10.3141/
11-R3. Virginia Transportation Research Council, Charlot- 2142-04. http://trb.metapress.com/content/k5774126v620
tesville, VA, USA. p655/
Evans, R.W. and Wilshire, B., 1993. Introduction to creep. Saha, J. and Bayat, A., 2011. Evaluation of the Canadian climate
London: Institute of Materials. information and its effect on pavement performance through
Ghalia, M.A., Hassan, A., and Yussuf, A., 2011. Mechanical and MEPDG prediction. In: Transportation Research Board 90th
thermal properties of calcium carbonate-filled PP/LLDPE annual meeting. Washington, DC: Transportation Research
composite. Journal of Applied Polymer Science, 121 (4), Board.
2413– 2421. doi:10.1002/app.33570. Schwartz, C. and Carvalho, R., 2007. Implementation of the
Hall, K.T., et al., 2001. Rehabilitation strategies for highway NCHRP 1-37A design guide. Baltimore, MD: University of
pavements. Washington, DC: Transportation Research Maryland, Maryland State Highway Administration.
Board. Smith, K.D., Zimmerman, K.A., and Finn, F.N., 2004. The
Hall, K.T., Simpson, A.L., and Correa, C.E., 2002. LTPP data AASHO road test: living legacy for highway pavements.
analysis: effectiveness of maintenance and rehabilitation TR News. Washington, DC: Transportation Research Board,
options. National Cooperative Highway Research Program. pp. 14 – 24.
Washington, DC: Transportation Research Board. Timm, D.H., Turochy, R.E., and Davis, K.P., 2010. Guidance for
Hallin, J.P., et al., 2007. Pavement design in the post-AASHO ME pavement design implementation. Auburn University,
road test era. Interstate Highway System, 1, 1 – 16. AL: Highway Research Center, Auburn University.
Huang, Y.H., 2004. Pavement design and analysis. Englewood United-States-Federal-Highway-Administration, 2004. Key find-
Cliffs, NJ: Prentice-Hall. ings from LTPP analysis, 2000– 2003. Washington, DC: U.S.
Hudson, W., et al., 2007. AASHO road test effect on pavement Federal Highway Administration.
design and evaluation after 50 years. Interstate Highway Williams, R.C. and Shaidur, R., 2013. Mechanistic-Empirical
System, 17, 17 – 30. Transportation Research Circular E- Pavement Design Guide calibration for pavement rehabilita-
C118. tion. Salem, OR: Oregon Department of Transportation,
Johanneck, L. and Khazanovich, L., 2010. Comprehensive 169 p.
evaluation of effect of climate in mechanistic-empirical Yang, X. and Wu, Z., 2012. Effects of subgrade resilient modulus
pavement design guide predictions. Transportation Research and climate inputs on MEPDG. In: GeoCongress 2012@
Record: Journal of the Transportation Research Board, State of the Art and Practice in Geotechnical Engineering.
2170 (1), 45 – 55. doi:10.3141/2170-06.
 12 M.R. Mohd Hasan et al.

ASCE. Reston, VA: American Society of Civil Engineers, pp. Zaghloul, S., et al., 2006. Investigations of environmental and
1448– 1457. doi:10.1061/9780784412121.149. traffic impacts on mechanistic-empirical pavement design
Yildirim, Y., 2007. Polymer modified asphalt binders. Construc- guide predictions. Transportation Research Record: Journal
tion and Building Materials, 21 (1), 66 – 72. doi:10.1016/j. of the Transportation Research Board, 1967 (1), 148– 159.
conbuildmat.2005.07.007. doi:10.3141/1967-15.
Downloaded by [University of Nebraska, Lincoln] at 10:13 31 August 2015