MEPDG Traffic Loading Defaults Derived From Traffic Pooled Fund Study

CHAPTER 1-INTRODUCTION

BACKGROUND

The Mechanistic-Empirical Pavement Design Guide (MEPDG) methodology was developed under National Cooperative Highway Research Program (NCHRP) Project 1-37A, and the MEPDG Manual of Practice was balloted affirmatively through the American Association of State Highway and Transportation Officials (AASHTO) in 2008 at the Joint Technical Committee on Pavements annual meeting.^(3,1) Currently, many State highway agencies are evaluating and implementing the new pavement design procedures.

This report focuses on traffic inputs that support the MEPDG method. As a result, the abbreviation "MEPDG" is used extensively throughout this report. This abbreviation does not refer to a specific publication; rather, it is used to define parameters, models, and computational procedures that support the MEPDG method.

The MEPDG requires actual traffic inputs, rather than the number of 18-kip equivalent single-axle loads (ESALs) determined through the use of the AASHTO equivalency factors.⁽⁴⁾ The MEPDG traffic inputs are normalized axle load spectra (NALS) for each truck class and axle group type, number of axles of each type per vehicle class, percentile truck class volume distribution, truck volume, and truck growth.

NALS are percentile distributions of axle counts by load level. Individual NALS are computed for each axle group type and truck class. The axle group types included in the MEPDG design procedure are single, tandem, tridem, and quad.⁽⁴⁾ Truck classes are vehicle classes 4 through 13 based on the Federal Highway Administration (FHWA) vehicle classification scheme F-13. For the design, NALS representative of axle load distribution for a typical day of the month are used. For roads that do not show significant seasonal variations, the same NALS are used for all calendar months.

Out of all MEPDG traffic inputs, NALS are the most challenging to collect due to the high data collection costs, the time needed to collect an adequate quantity of data, and weigh-in-motion (WIM) equipment calibration and maintenance requirements. Therefore, the majority of pavement designs in accordance with the MEPDG rely on regional, agency-wide, or national NALS traffic loading defaults. It is important that default NALS values are as representative of the actual truck loading along the project site as possible.

The default traffic datasets included in the MEPDG were developed during NCHRP Project 1-37A and based on Long-Term Pavement Performance (LTPP) traffic data collected by State highway agencies prior to 1999.⁽³⁾ At that time, these data represented the best available and most comprehensive national set of WIM and automatic vehicle classification (AVC) data. However, some concerns existed regarding the lack of documented quality of the data that were used to determine the truck traffic default values-specifically, the NALS values for each axle group type within a truck class. More importantly, many agency personnel have also questioned the quality of their own truck traffic data.

Since 1999, the LTPP Specific Pavement Study (SPS) Traffic Data Collection transportation pooled fund (TPF) study, TPF-5(004) (SPS TPF study) has generated high-quality traffic loading information for 26 LTPP SPS TPF WIM sites.⁽²⁾ These data were collected using uniform protocols for data quality assurance (QA) and in accordance with the LTPP vehicle classification scheme. These new data provide an opportunity to evaluate the applicability of the MEPDG NALS defaults and to provide alternate NALS.

RESEARCH OBJECTIVES AND OUTCOMES

The primary purpose of this project was to revise axle loading defaults based on data from SPS TPF WIM sites, compare new defaults with the existing MEPDG global NALS defaults, and provide recommendations for use of the new axle loading defaults. This research project addresses the following specific objectives:

• Examine improvements since the original issue of the traffic data for the MEPDG to develop new design defaults for traffic inputs.
  
  
• Develop additional default NALS for the MEPDG using SPS TPF data.
  
  
• Develop a procedure for estimating the reliability of the traffic load data using confidence levels (CLs).

The following are the outcomes and deliverables of this study:

• An alternate set of default NALS for the MEPDG.
  
  
• Practical guidelines for generation and use of MEPDG traffic loading defaults in the form of a stand-alone guide, LTPP Pavement Loading User Guide, known as LTPP-PLUG.⁽⁵⁾
  
  
• A procedure for estimating the reliability of the traffic loading data used in pavement design.
  
  
• A final report summarizing the study findings and outcomes, including a description of the new default traffic datasets for the MEPDG and an assessment of the sensitivity of MEPDG pavement design models to differences in NALS and to the accuracy of WIM data.
  
  
• A software tool (LTPP-PLUG) for the development of NALS input files for the MEPDG and DARWin-ME^TM software using the new LTPP SPS TPF defaults as well as user-supplied defaults or site-specific NALS.⁽⁵⁾
  
  
• The following two new statistical parameters intended to help simplify analyses involving NALS while also allowing pavement analysts to easily summarize and communicate the relative size of traffic loads being experienced by roadways:
  
  
  • Relative pavement performance impact factor (RPPIF): A summary statistic for comparison and grouping of similar NALS. This statistic converts NALS to a single value, considering both the frequency of load applications and the relative effect of different load magnitudes on pavement performance.
    
    
  • Pooled weighted load error (PWLE): A parameter to quantify the potential for error associated with using several NALS to compute a single group NALS. This summary can be used to assess potential errors when computing an average monthly or an average annual NALS at a specific site using different data availability scenarios or when computing a single NALS that is intended to represent a group from the NALS measured at multiple sites within that group. The PWLE statistic, computed for a selected CL, can also be used to assess the reliability of a summary NALS. It takes into consideration both the variation in relative percentage of the traffic at each axle load level and the relative importance of each load level on the MEPDG-based pavement performance predictions.

REPORT ORGANIZATION

This report consists of two parts. Part I (chapters 1 through 7) focuses on a review of the MEPDG traffic loading data requirements, data assessment, and development of data selection criteria. Part II (chapters 8 through 11) focuses on the development of alternate MEPDG traffic loading defaults and recommendations for use of those alternate defaults.