Developing Pavement Thickness Design incorporating the Catalogue and Mechanistic Approach

New approaches are being taken to improve the design and durability of road pavement. Photo credit: ADB

Share on:           


A preliminary study in Sri Lanka provides important insights into mechanistic-empirical pavement thickness and overlay design for roadway networks.


During the last nine years, the Sri Lankan economy has grown rapidly. The government strives to, and encourages, continued growth. A prerequisite for this is an improved transport system and investments in the road network have been intensified.

For the last 9 years more than 10,000 km of roads were rehabilitated and around 150 km of new expressways were built. The government plans to build another 200 km length of expressways and to rehabilitate around 6,000 km of road within the next five years. The Road Development Authority of Sri Lanka is responsible for building, operating and managing national roads and expressways and provincial road agencies and local authorities are responsible for provincial and rural roads.

The current road pavement designs are based on empirical methods, based on Road Note 31, published by the Transport Research Laboratory about 30 years ago.  The empirical methods restrict the material properties to specified limits and as a result, most of the soil and aggregates available in the country cannot be used for the road construction in an optimum way of usage. This situation has caused shortage of materials such as aggregates and as a solution the government expects to adopt the mechanistic-empirical approach for a road pavement thickness design.

Road networks in Sri Lanka are structurally designed according to Road Note 31, which restricts the indirect material property to a specified-limit of the California Bearing Ratio value. As a result, a large amount of in-situ soil and aggregates available in the country have not been used as roadway construction materials in the context of optimum quantity. This may also result in over-estimated or under-estimated catalogues of the current pavement design guide without any consideration of the structural capacity of a pavement system.

This situation has caused a shortage or overuse of materials. To solve this problem, the Sri Lankan government expects to adopt the mechanistic-empirical design system for new or overlay thickness design of pavements, especially for low-volume roads. Mechanistic-empirical design can provide flexibility in using a wide range of material properties because the pavement thickness design can be defined not by the indirect material property such as the California Bearing Ratio value but by mechanical responses, such as strain or stress in a critical location of a layered pavement system.

In January 2017, the Asian Development Bank and the Road Development Authority of Sri Lanka started a technical assistance project to develop the mechanistic-empirical design concept and procedure for a new road construction and the rehabilitation of existing pavements and to help the Road Development Authority prepare a comprehensive database system for pavement management and analysis in the near future.

Project Snapshot

  • January 2017 : Approval Date
  • August 2017 : Completion Date

  • $138,000 : Cost

  • Executing agency :
    • Korea Institute of Civil Engineering and Building Technology, South Korea
  • Financing :
  • Financing :
    • Road Development Authority, Sri Lanka

The Road Note 31 is a catalogue pavement design guide, which is in the category of the first generation of the pavement design method, such as the empirical approach. The drawbacks in utilizing the empirical catalogue design method may be the lack of flexibility to introduce a new type of pavement system using in-situ materials in view of optimum amount and non-availability of mechanistic–empirical material property database.

Currently, the pavement design in Sri Lanka is carried out considering the traffic loading for a selected design period in terms of Equivalent Single Axle Load. While the asphalt surface is not treated as a structural layer especially for the double bituminous surface treatment (DBST), the structural capacities of unbound layers in base, subbase, and subgrade are generally characterized by California Bearing Ratio value with consideration of drainage capacity without and the calibration of seasonal variation.

A mechanistic analysis engine, including well-defined functions for material behavior, was needed to calculate the pavement responses. The pavement responses under the specified equivalent single axle load can be used as an independent variable to estimate the pavement performance life. The mechanistic-empirical design approach can provide flexibility in using a wide range of materials but is not optimum because the pavement thickness design can be defined not by the indirect material property of the California Bearing Ratio value but by the mechanistic responses such as strain or stress in a critical location of a layered pavement system.

Besides, an overlay design system was another challenge which is widely used in Sri Lanka but Road Note 31 doesn’t cover anything about overlay design. Another challenge was to shift asphalt binder testing standards from Penetration grading to Superpave performance grading and a comprehensive rheological analysis for construction of Mastercurve for Sri Lankan asphalt binder.

For the successful implementation of aforementioned tools and lab tests, an overall roadmap explaining the time schedule and deliverables was proposed and roadmaps for individual systems were presented as a result of this project.


Mechanistic-empirical design requires short term and long term implementation plans. Short term and long term plans detail the time schedule for developing the mechanistic-empirical design engine for pavement response calculation, thickness design algorithm, and mechanistic-empirical design software for pre and post processes of the design and limited lab test and database preparation. The only difference between the two is that long term planning details are an additional feature of overlay design tools and database management systems.

The mechanistic-empirical design engine is supposed to calculate pavement responses due to axle loading. The engine can be developed by the layered elastic theory or continuum theory in the context of finite element approach. The concept of the design software is to verify the current catalogue of Road Note 31 based upon a mechanistic approach and have a function of a new structural pavement design also. A user can retrieve the graphical analysis results for all catalogues by a user query with respect to a location and traffic level. The design algorithm can be used for designing a new pavement structure in addition to the current catalogues.

The conceptual and logical database design can be done during the short-term plan. However, the physical database only includes a limited lab test, field test, and field survey data. The database should be interrelated with the mechanistic-empirical design tool, overlay design tool, and the pavement management system in the long run. A national specification for roadway construction can be effectively updated annually or else using the database system to decide a number of calibration factors after doing a multivariable regression analysis.

User requirement analysis, conceptual, logical, and physical database design can be done during the long-term plan. All the data will be interrelated by the entity-relation diagram and data flow diagram. A graphic user interface will be designed and implemented to manage all the data.

The mechanistic-empirical design of asphalt overlay pavements requires an iterative and trial and error approach. A designer must select a proposed trial overlay design and then analyze a design in detail to determine whether it meets applicable performance criteria (i.e., rutting and fatigue cracking) limited in a specification. If a particular trial overlay design does not meet the performance criteria, the design must be modified and reanalyzed until it meets the criteria. The design that meet the applicable performance criteria is then considered feasible from a structural and functional viewpoint and can be further considered for cost efficiency.

The mastercurve of an asphalt binder provides a relationship between the binder stiffness and reduced frequency over a range of temperatures and frequencies. For this purpose, complex modulus of asphalt binder at multiple temperatures and frequencies was measured using Dynamic Shear Rheometer. The data from laboratory test was then fitted into a viscoelastic model to construct mastercurve.


Figure 1: Implementation Roadmap for Mechanistic-Empirical Pavement Design in Sri Lanka. Photo credit: KICT

All pavement catalogues in the current Road Note 31 should be verified by the mechanistic analysis in the context of finite element analysis. The RoadMap is focused on developing the Mechanistic-Empirical design software. It includes calculation engine design and coding, imbedding material models, calibration, and trial application. The finite element engine was developed in the context of a 2D axisymmetric domain equipped with pre- and post-processors. For the verification purpose, the commercial three-dimensional finite element analysis software could be utilized to comparatively analyze the results from the 2D engine.

The material model for the asphalt binder and asphalt mixture will be based on a time-temperature dependent models, such as the Generalized Maxwell Model and Generalized Kelvin Model defined by the lab tests, such as creep test and dynamic modulus test. For the unbound materials, such aggregate base and subgrade, the resilient modulus can be utilized as material properties in view of elastic material behavior.

Figure 2: Implementation Roadmap for Overlay Design in Sri Lanka. Photo credit: KICT

The mechanistic-empirical design of asphalt overlay pavements requires an iterative and trial-error approach. The designer must select a proposed trial overlay design and then analyze the design in detail to determine whether it meets the applicable performance criteria (i.e., rutting and fatigue cracking) established by the specification. The mechanistic-empirical overlay design system developed should be validated and calibrated for accurate prediction of pavement performance using the field monitoring data. The overlay design program consists of input characterization, backcalculation code, link process with mechanistic-empirical design, and overlay thickness determination. The graphic user interface will be compatible with Sri Lankan mechanistic-empirical design system.

Figure 3: Implementation Roadmap for Database Management System in Sri Lanka. Photo credit: KICT

The roadmap for developing database system for Sri Lanka consists of data collection, schema and database design, and Implementation. The database management system for Sri Lanka will be carried out considering following key points;

  1. Data input and output modules
  2. Data generation module
  3. Data report module
  4. RoadNote pavement design and overlay link module


Figure 4: Comparison of Sri Lankan Binder Mastercurves with standard PG 64-22 at 40oC. Photo credit: KICT

Frequency sweeps test data at different temperatures was obtained using Dynamic Shear Rheometer. The data was further analyzed for predicted complex modulus using mathematical models for linear viscoelasticity of asphalt binder. Error between experimental data and predicted complex modulus values was then calculated. Sigmoidal Curve model was employed to fit the data to experimental values and a mastercurve was produced.


Structural Analysis with Different Modules

A structural Analysis was conducted to illustrate the effect of various parameters on pavement responses. Due to the complex interactions among the large number of parameters (loads, tire pressure, moduli, layer thicknesses, Poisson’s ratio etc.), the best approach is to fix all other parameters at their most reasonable values while varying the parameter in question to show its effect.

A series of mechanistic analysis was performed with variations of HMA thickness, Base thickness, Subbase thickness, and Base layer modulus, for the current iRoad design by the Road Note 31, 1993 AASHTO design for T3 traffic level, and TA recommended-design by the mechanistic analysis software. The major conclusion drawn from the results of the sensitivity analysis is that increasing the moduli of the base, i.e., using an asphalt treated base; in the design significantly reduces the critical responses between the pavement systems.

Figure 5: Comparison of Pavement Layers Thicknesses using different Design Approaches. Photo credit: KICT

2D and 3D Finite Element Analysis

Figure 6: Comparison of Mechanistic Analysis with 3D Finite Element Analysis. Photo credit: KICT

All the mechanistic analysis was performed using the KICTPAVE and KPRP, the KICTPAVE is a two-dimensional FE analysis program developed by Korea Institute of Civil engineering and building Technology and the KPRP is a mechanistic pavement design software developed by Korean government.

The iRoad design was defined by the Road Note 31 specification, while the AASHTO design was defined using the 1993 AASHTO design guide aforementioned. KPRP design is based on the layered elastic design approach, while the recommended thickness designs bu KICTPAVE, such as ReCom 1 and ReCom 2 were mechanically decided what the critical responses were highly diminished being compared to others.

The modulus for the surface HMA was fixed in the 400,000 psi and the stiffness of the remaining materials was varied by the relative stiffness of adjacent layers. The main issue of this study is that the mechanistic comparison between the pavement structure with an aggregate base with relatively thick layer (ABC, 10cm~18cm) and the pavement layer with a stiff treated-base (Asphalt Treated Base, 7~10cm).

The influential damage area of the pavement design having a relatively loose aggregate course, such as the iRoad design, AASHTO design, and the KPRP software was the largest because of the strain energy may dissipate into strain damage.

This means that the pavement design with a stiff but thin base course may be an optimal design in view of structural soundness.

In conclusion, the mechanistic-empirical approach can draw an analytical result supporting a mechanistic backup to reduce the thickness of a specific layer such as aforementioned ABC layer, even though the recommended was for a relatively thin asphalt treated base instead of the ABC layer in this study.

Figure 7: Comparison of 2D Critical Influential Area with Different Designs. Photo credit: KICT

Pyeong Jun Yoo
Research Fellow, Korea Institute of Civil Engineering and Building Technology

Pyeong Jun Yoo has been developing and implementing pavement design and management systems for roadways for 20 years. He earned his BSc and MSc in Civil Engineering from Konkuk University in Seoul and his PhD program in finite element analysis from the University of Illinois at Urbana-Champaign in 2007.

Hee Mun Park
Research Fellow, Korea Institute of Civil Engineering and Building Technology

Hee Mun Park, a Research Fellow, Highway Research Division, Korea Institute of Civil Engineering and Building Technology, has been worked in the areas of pavement design, maintenance, rehabilitation, management for 15 years.  He received the BSc degree in civil engineering department of Hanyang University in Seoul.  He earned his master degree in geotechnical engineering from Texas A&M University, USA.  He also received the PhD degree in pavement engineering from North Carolina State University in 2001 and his dissertation focused on the condition evaluation of asphalt pavements.

Korea Institute of Civil Engineering and Building Technology (KICT)

The Korea Institute of Civil Engineering and Building Technology contributes to the development of the Korean construction industry, improves quality of life standards, furthers national economic growth, and improves social welfare. It promotes original technology in the fields of land, infrastructure, and construction.

Leave your question or comment in the section below:

The views expressed on this website are those of the authors and do not necessarily reflect the views and policies of the Asian Development Bank (ADB) or its Board of Governors or the governments they represent. ADB does not guarantee the accuracy of the data included in this publication and accepts no responsibility for any consequence of their use. By making any designation of or reference to a particular territory or geographic area, or by using the term “country” in this document, ADB does not intend to make any judgments as to the legal or other status of any territory or area.