5th INTERNATIONAL CONFERENCE

BITUMINOUS MIXTURES AND PAVEMENTS

Thessaloniki, Greece, 2-3 June 2011

COMPARISON OF MECHANISTIC-EMPIRICAL AND

EMPIRICAL FLEXIBLE PAVEMENT DESIGN PROCEDURES
OF AASHTO: A CASE STUDY

Mehdi Mashayekhi *
M.S Student, Iran University of Science and Technology (IUST),
*
Corresponding Author, mmashayekhi@civileng.iust.ac.ir
Amir Ali Amini
PhD Student, Iran University of Science and Technology (IUST)
Hamid Behbahani
Professor, Iran University of Science and Technology (IUST)
Shams Nobakht
Assistant Professor, Iran University of Science and Technology (IUST)

ABSTRACT
Currently, the most current method for pavement design is AASHTO (1993)
method which is an empirical method. However limitations of the empirical
approach are becoming increasingly apparent with developments and increased
knowledge in the fields of pavement mechanics and material science.
Mechanistic-empirical pavement design Guide (MEPDG) is an effort to address
these limitations. Mechanistic-empirical pavement design Guide is a new
method proposed under NCHRP Project 1-37A and 1-40D which is based on
numerical models. The objective of this paper is to compare the design and
performance between the empirical AASHTO and the MEPDG method using
highest level of data accuracy. Five previously constructed pavement sections
designed by empirical AASHTO method have been used in this study. An
analysis has been made on the sections using MEPDG and new pavement
sections have also been designed and the differences have been illustrated. The
results show that using MEPDG results in thinner AC sections, but the amount
of difference is dependent to performance criteria chosen. For almost all section
used, fatigue cracking is controlling criteria.

KEY WORDS: Flexible pavements design; MEPDG; AASHTO empirical

design.

1
1. INTRODUCTION

For many years, empirical method of AASHTO has been used for design of
pavements. AASHTO empirical method is one of the most used methods in
many countries like Iran. This procedure is based on the results of the extensive
AASHO Road Test conducted in Ottawa, Illinois, in the late 1950s and early
1960s. The AASHO Committee on Design first published a temporary design
guide in 1961. It was revised in 1972 and 1981. In 1984-85, under the project of
NCHRP Project 20-7/24; the guide expanded [1, 4].
This procedure has some limitations, because the original equations were
developed under a given climate condition with a specific set of pavement
materials and subgrade soils.
The proposed AASHTO Mechanistic Empirical Pavement Design Guide
(MEPDG) developed under National Cooperative Research Program (NCHRP)
Projects 1-37A and 1-40D has been developed to eliminate some of the
limitations of the empirical method and to provide the highway community with
a state-of-the-practice tool for the design of new and rehabilitated pavement
structures.
Unlike empirical procedure, the mechanistic empirical format of MEPDG
provides a framework for continuous improvements to keep up with changes in
trucking, materials, construction, design concepts, computers, and others.
The Mechanistic part of the procedure refers to the mathematical modeling
that relates performance criteria (e.g. stress and strains of critical locations of
pavement) to distresses such as fatigue or permanent deformation while
empirical calibration factors are needed to relate mentioned equations to
laboratory results. Empirical transfer functions are also needed to develop
relationships between laboratory tests to field results with a specific set of
conditions (pavement materials, traffic in addition to climatic situation).
MEPDG procedure is expected to reduce early failures and increase
pavement longevity [9].
Mechanistic-Empirical pavement design is still in progress and is not
officially announced as the standard procedure for pavement design, there are
some limitations that restrict its application, like development of calibration
factors and transfer functions. However, in the process of developing, there is
an urgent need for validation of results and comparison of outputs with previous
methods of pavement design.
The purpose of this research is to compare the pavement design of AASHTO
empirical procedure with MEPDG. Data of 5 sections which already designed
with AASHTO method are used as inputs for MEPDG software and other data
which are necessary for MEPDG software such as traffic and climate are
specified using level 1 and these two methods are compared.

2
2. LITERATURE REVIEW

There are a number of published work in this field and here some of them are
mentioned.
Carvalho and Schwartz compared AASHTO empirical versus NCHRP
project 1-37A and concluded that also these methods are consistent with each
other, the empirical AASHTO guide underestimates distress for pavements in
warm locations and at high traffic levels [3]. Li et al. compared empirical,
MEPDG and historical performance of Washington state pavements and
concluded that empirical method overdesigns pavement thickness for all
sections [6].
In a research done by Mulandi et all to compare the the design and
performance of MEPDG with AASHTO empirical, it is found that MEPDG
software yielded thinner AC sections in comparison with empirical procedure
[8]. Another research done by Timm on rigid pavement revealed that the slab
thickness resulted from MEPDG software are typically 9 percent thinner than
empirical method [12].

3. METHODOLOGY

For comparison of these two methodologies, data of five corridors with

different climate and traffic volume are gathered. These corridors cover a wide
range of climate and traffic situations.
Initially, using AASHTO empirical pavement design method, as-built
pavement section are controlled for each of these five corridors and then two
approaches are applied for comparison of methods and in each approach
different initial and terminal IRI are used which will be explained later in the
paper.
In first approach, sections are modeled with MEPDG software with the same
thickness of conventionally designed sections and final distresses and damages
to pavements are evaluated. This approach is called “analysis comparison”
because it is an analysis of the section by empirical and MEPDG methods. The
second approach consists of designing sections for a specific design life with
MEPDG method and comparison of the required thickness of asphaltic layers.
This approach is called “design comparison”.
Comparisons are made based on two criteria, International Roughness Index
(IRI) and bottom up fatigue cracking. The other distresses did not show any
significant level of deterioration and are not considered as failure criteria.
It must be noted that because distresses appear on AC layers and varying
these layers thickness change distresses considerably, only the thickness of this
layer is redesigned for all corridors. As there was not any local research on
MEPDG calibration factors, default calibration factors are used in this study.

3
3.1. Performance Criteria
To correlate pavement section responses to traffic and climatic parameters from
these methods, there are two parameters that must be interchanged into one
single parameter. These parameters are Present Serviceability Rating (PSR) and
International Roughness Index (IRI). As determined in empirical procedures,
initial PSR and terminal PSR of main highways are considered 4.2 and 2.5,
respectively. Various correlations have been developed between PSR and IRI.
Two most used of these equations are used in this paper.
A relationship is developed by Paterson in 1986: [10]
(-0.00286*IRI)
PSR= 5e (1)
Another correlation is reported in 1992 at Illinois by Al-Omari and Darter:
[2]
(-0.00413*IRI)
PSR= 5e (2)
Where:
PSR: Present Serviceability Rating (varying between 5 to 0),
IRI: International Roughness Index (in/mile).
Based on the equation (1) the initial and terminal IRI would be 63 and 242
in/mile, respectively. And based on the equation 2 the initial and terminal IRI
would be 42 and 168 in/mile respectively.
Another major criterion is bottom up fatigue cracking known as alligator
cracking. Alligator cracking is a major distress especially in Iran, and that's why
it is selected for evaluating the performance of pavement in addition to IRI.
Alligator cracking distress is evaluated by the percentage of the total lane area
covered with cracks. Default maximum allowable alligator cracking
recommended by MEPDG is 25 percent of total lane area. Besides considering
25 percent alligator cracking, in this paper, another maximum allowable crack
area of 10 percent is used to compare the results of changing alligator cracking.

3.2. MEPDG Software

The software has three main categories including: general Information, inputs
and Results. When defining the inputs for the MEPDG, there are three
hierarchal input levels. The input level used is selected depending on the
amount of information available to the designer and the significance of the
project. These input levels are as follows:
• Level 1: Inputs are provided for the highest level of accuracy and
lowest level of uncertainty.
• Level 2: Inputs are provided from a limited testing program, or could be
estimated through correlation.
• Level 3: Inputs are estimated using typical values chosen from available
source and experience. [9]
In this study, version 1.1 of the software is used. (Release date: August
2009)

4
3.3. Input data
Meteorological data of eight consecutive years (2000 – 2007) is gathered
through contacts with Iran meteorological organization. Traffic data of corridors
including AADT and AHT are put in MEPDG according to Road Maintenance
and Transport Organization (RMTO) data. Vehicle properties are also obtained
from vehicle brochures at website of car manufacturer and the database needed
for specification of as-built layers is gathered through contacts with ministry of
road and transportation of Iran. Table 1 shows general specification of sections
for each project. General climate data of project locations can be seen in table 2,
while design parameters and final thickness for empirical method is presented in
table 3.
General information includes design period, construction month, traffic
opening month, pavement type, failure criteria for each deteriorations and their
reliability. Inputs section has three main subcategories: Traffic, Climate, and
Structure.

Table 1. Specification of pavement projects and relative sections

Origin Andimeshk Hamedan Saveh Sirjan Tehran
Destination Ahvaz Sahneh Salafchegan Bandar Rudehen
Project codes AA HS SS SB TR
Length of Project (Km) 150 140 66 72 32.3
Design Period (year) 20 20 20 20 20
Initial two way AADT 17813 5852 10000 5764 27181
Number of lanes 6 4 6 4 4
Percent of trucks (%) 49 31 55 75 14
Trucks in design direction (%) 50 50 50 50 50
Trucks in design lane (%) 70 80 70 90 90
Annual growth rate of traffic (%) 4 5 3 3 3

Table 2. General climate specification of pavement projects location

Project Codes AA HS SS SB TR
Annual Average Temperature(°c) 26 11 18 26 18
Annual Average Precipitation (mm) 20 27 13 15 20

3.3.1. Traffic
Traffic parameters include traffic volume adjustment factors (monthly
adjustment, vehicle class distribution, hourly truck distribution, and traffic
growth factors), axle load distribution factors and general traffic inputs (number
of axles per truck, axle configuration, and wheelbase).
As described before, traffic volume adjustment factors are gathered from the
website of RMTO. RMTO provides detailed data of traffic for Iran road
network [11]. And general traffic inputs are obtained from vehicle brochures at

5
the website of car manufacturer [5]. Traffic data is entered in MEPDG software
using level 1 of data entry, which is the highest data entry level.

3.3.2. Climate
Climate input is hourly data for five climate parameters. Temperature,
precipitation, wind speed, percentage of sunshine and relative humidity is
necessary to create Integrated Climatic Models (ICM), an input file that must be
put in software.
As mentioned, meteorological data of eight consecutive years (2000 – 2007)
is gathered from Iran meteorological organization, and ICM files for every
section is created and used in MEPDG software.

3.3.3. Structure
Materials type of each layer and its characteristics is specified in accordance
with table 3. Level 1 of data entry is also used for the specifications of
pavement structure.

Table 3. Design specification and thickness of sections

Projects* AA HS SS SB TR
Reliability 94% 85 94% 94% 90%
ZR -1.555 -1.037 -1.555 -1.555 -1.282
S0 0.35 0.35 0.35 0.35 0.35
Wt18(million) 136 25,5 69,7 72,7 66,3
P0 4.2 4.2 4.2 4.2 4.2
Pt 2.5 2.5 2.5 2.5 2.5
Subgrade Mr (psi) 10000 11450 13000 12500 11500
Sub base Mr (psi) 14500 13500 15000 15000 14000
Base Mr (psi) 30000 30000 35000 30000 29000
m2 and m3 1.0 1.0 1.2 1.1 1.2
AC Thickness (mm) 240 170 160 170 180
Base Thickness (mm) 200 150 150 200 150
Sub base Thickness (mm) 300 300 300 300 300
* a1, a2, and a3 are considered 0.42, 0.14, and 0.1 respectively for all sections.

It must be note that reliability of the sections is considered according to the

classification of the corridors. And according to Iranian code for pavement
design, standard deviation (S0) for main highways is suggested to be 0.35 as it is
considered in this research [7].

6
4. RESULTS

4.1. Analysis Comparison

Generally, MEPDG results of sections analyzed predict lower IRI than
conventional method’s prediction at terminal IRI at the end of design life. This
is in agree with works which state AASHTO empirical method overdesigns and
is too conservative [3, 6]. All sections show lower alligator cracking and none
of them meet terminal alligator cracking of 25 percent and 10 percent except
section SB. Furthermore, changing the initial and terminal IRI has no effect on
the result of alligator cracking. So, only one graph for alligator cracking and
two graphs for IRI are shown in figures 1 to 3.
Figure 1 shows the result of alligator cracking. According to this figure all of
the sections have low alligator cracking and according to this criteria all
sections except SB, have life more than twice of the design period, which is 240
months. As shown in figure 2, according to IRI criteria, with initial IRI of 63
and terminal IRI of 242, all section have design life 10 to 20 years more than
the design period of 20, except AA section which has 21 years life. Figure 3
shows the results of IRI over the years based on initial IRI of 42 and terminal
IRI of 168. By using these initial and terminal IRI, the life of sections decreased
10 to 13 years in comparison with initial and terminal IRI of 63, 242. But except
section AA, all sections have longer life than the design life of 20 years. Results
of analyses are shown in table 3.

4.2. Design Comparison

The tables of required thickness and percentage of difference in thickness are
shown based on two different criterions of alligator cracking.IRI is not
controlling criterion except for section AA.

Figure 1. Bottom-up damage for alligator cracking

7
 Figure 2. IRI values (IRIi=63)

Figure 3. IRI values (IRIi=42,)

Results of design comparison are shown in table 4. In this table numbers

between parentheses are to show the change in thickness of design by MEPDG
in comparison with empirical method. It is seen that in Iran, initial and terminal
IRI is closer to results of Paterson equation. Typically, initial and terminal IRI
are 63 and 252 in Iran. So, the required thickness is calculated only based on
Paterson equation.
For all sections the controlling criterion for design is alligator cracking
except section AA which IRI is controlling criterion. According to this table,
there is 6 to 33 percent decrease in asphalt thickness by considering 25 percent
alligator cracking, except section AA and considering 10 percent alligator
cracking, there is 6 percent increase for section SB, and average 9 percent
decrease for other sections except section AA.

8
 Table 3. Time of reaching Performance Criteria (year)
th)
Project AA HS SS SB TR
Bottom Up Maximum Damage (10%) 36 29 34 16 31
Bottom Up Maximum Damage (25%) - - - 26 -
IRI (IRIi=63, IRIt=242) 21 42 40 31 41
IRI (IRIi=42, IRIt=168) 11 31 28 20 28

Table 4. Total AC thicknesses & percentage of difference of empirical and

MEPDG
Project AA HS SS SB TR
25% Alligator
240 (0%) 100 (-33%) 130 (-19%) 160 (-5.9%) 130 - (-28%)
Cracking
10% Alligator
240 (0%) 135 (-10%) 145 (-9%) 180 (+5.9%) 165- (-8.3%)
Cracking

5. CONCLUSION

The results of comparison between empirical AASHTO and MEPDG methods

presented in this paper show that there is a significant difference in analysis and
design of pavement sections, at least when there exist extensive data to specify
the conditions of the pavement. MEPDG thickness required is up to 33 percent
thinner, although in a section the same thickness has been acquired for both
procedures. This difference is logical and explainable, when you have an
empirical method with one basic equation and very little data available, you
prefer to stay on the safe side. However, too many calibration factors and
transfer functions make MEPDG hard to use, at least for developing countries.
Apparently, thinner sections mean more economical pavements and less
material consumption and environmental impacts. This means that MEPDG can
lead to more accurate pavement modeling that can lead to a considerable
amount of cost savings. Some brief notes on conclusions of this study are as
follows:
Generally, in this research asphalt thickness designed with MEPDG are
thinner than empirical method. Results of comparison are also found to be
sensitive to failure criteria chosen. In sections AA, and SB, which have high
truck traffics and temperature the results of two methods are closer to each other
than sections HS, SS, and TR which have low truck traffic and temperature.
Alligator cracking is much higher than other distresses, and other distresses did
not show any significant amount. This is in agreement with the distresses
occurring in Iran, because in Iran, alligator cracking is the important distress in
pavements.

9
ACKNOWLEDGMENT: We wish to sincerely thank Iran Meteorological
Organization, RMTO and especially CDTIC of ministry of road and
transportation for providing data used in this paper.

REFERENCES:

[1] AASHTO, AASHTO Flexible Pavement Structural Design, American

Association of State Highway and Transportation Officials, 1993.
[2] Al-Omari B, and M.I Darter, Relationships between IRI and PSR, Interim
Report, Publication No.UILU-ENG-92-2013, Illinois Department of
Transportation, 1992.
[3] Carvalho, R.L, Schwartz, C.W., Comparisons of Flexible Pavement
Designs: AASHTO Empirical Versus NCHRP Project 1-37A Mechanistic-
Empirical, Transportation Research Record, 0361-1981, vol. 1947, pp.
167-174, 2007.
[4] Huang Y.H, Pavement Analysis and Design.2nd Edition, Upper Saddle
River, New Jersey: Pearson Education, Inc, 2004.
[5] Iran Khodro Company (IKCO), Specification of products, retrieved from
http://www.ikco.com/Fa/Default.aspx, 2010.
[6] Li, J., Uhlmeyer, J.S., Mahoney, J.P., Muench, S.T., Use of the AASHTO
1993 Guide, MEPDG and Historical Performance to Update the WSDOT
Pavement Design Catalog, Transportation Research Board 2010 Annual
Meeting CD-Rom, 2010.
[7] Ministry of road and transportation, Iranian code for pavement design
(234), 2003.
[8] Mulandi J, T. Khanum, M. Hossain, and G. Schieber, Comparison of
Pavement Design Using AASHTO 1993 and NCHRP Mechanistic-
Empirical Pavement Design Guides, Proceedings of the Airfield and
Highway Pavement Specialty Conference, Atlanta, GA, Sponsored by the
Transportation & Development Institute of ASCE, 2006.
[9] National Cooperative Highway Research Program (NCHRP), Guide for
Mechanistic-Empirical Design of New and Rehabilitated Pavement
Structures, Final Report for Project 1-37A. Part1, Chapter1. Washington,
D.C, NCHRP, Transportation Research Board, 2004.
[10] Paterson D.E, Evaluation of Pavement Maintenance Strategies, NCHRP
Synthesis 77, National Research Council, Transportation Research Board,
Washington, D.C., September, 1986.
[11] Road Maintenance and transportation organization (RMTO), Statistical
data of transportation, retrieved from, http://www.rmto.ir/MainE.asp,
2010.
[12] Timm D.H, Design Comparisons Using the New Mechanistic-Empirical
Rigid Pavement Design Guide, Proceedings of the Airfield and Highway
Pavement Specialty Conference, Atlanta, GA, Sponsored by the
Transportation & Development Institute of ASCE, 2006.

10