Verification, Refinement, and Applicability of Long-Term Pavement Performance Vehicle Classification Rules

Chapter 1. Introduction

Project Background

Pavement analyses depend on accurate and consistent load estimates derived from traffic data. To meet this need, the Long-Term Pavement Performance (LTPP) project requested that all States submit traffic data according to the Federal Highway Administration (FHWA) 13-category vehicle classification rule set. The practical reality, however, is that the FHWA 13-category vehicle classification rule set is a visual description, and the rules and algorithms used to convert axle spacing and weight data into those visual classifications vary considerably from State to State and vendor to vendor.

The Transportation Research Board (TRB) Expert Task Group (ETG) on LTPP Traffic Data Collection and Analysis (or Traffic ETG) identified the inconsistencies in the classification data as problematic. Therefore, under the LTPP Specific Pavement Study (SPS) Transportation Pooled Fund Study (TPF), TPF-5(004), the Traffic ETG developed a prototype set of classification rules for use in an effort to bring uniformity to the SPS traffic data collection. The developers of the LTPP classification rules recognized the challenges of using the same rules at different locations because many vehicles are specific to regions or States. Now that the LTPP rules have been deployed, insight was sought to determine how well they were functioning and gain a better understanding of the implications of their use for pavement design.

Consequently, this project was developed to examine the operational performance of the LTPP classification rules, determine the size and nature of errors in traffic loading estimates that inconsistencies between State and LTPP vehicle classification approaches create, and determine the effects those traffic-loading errors have on pavement analysis outputs. Based on those findings, this project then developed recommendations on how load spectra collected at the LTPP TPF sites could be used within various pavement analyses conducted at non-TPF test sites. In addition, minor changes to the LTPP rules were recommended to improve their performance. These changes were implemented at three pilot SPS weight-in-motion (WIM) sites. The results of pilot field tests of the recommended revised classification rules are included in this report.

Project Objectives

More specifically, this project had the broad objectives of answering the following questions:

• How well are the LTPP classification rules performing across the country? (For example are they misclassifying specific types of vehicles found in some States but not others?)
• Do the LTPP classification rules have to accurately classify all vehicle types in order for them to be used for pavement research?
• What changes are recommended (if any) to the LTPP classification rules to make them more accurate and universal?
• What is the field of applicability of the LTPP classification rules, i.e., are certain vehicle classes unambiguous and the same throughout the study area, even though all classes are not?

As part of answering these questions, the project investigated the sensitivity of pavement design to the use of different vehicle classification rules. Particularly emphasized was the case where the load spectra used in the pavement design process were developed based on volume by classification data collected using one set of classification rules while the truck volume count information used for that design were collected using a different set of vehicle classification rules.

Project Outcomes

The primary outcomes of this project are as follows:

• A refined set of LTPP vehicle classification rules.
• Mathematical estimates of the size of traffic loading errors caused by the use of different classification rules and algorithms at sites where load spectra are developed versus where truck volumes are counted.
• Mathematical estimates of the effects those traffic-loading errors can have on pavement analysis outputs.
• Procedures for selecting SPS TPF load spectra for use at LTPP test sites where site-specific load spectra data are not available.
• Identification of the traffic-loading conditions for which the use of SPS TPF load spectra data for pavement analysis is not appropriate.
• Recommended changes to the LTPP classification rules to improve their performance

Report Overview

This report is divided into three distinct sections. Part I describes the results of tests that verified the applicability of the “LTPP classification rule set” given the wide variety of traffic data collection locations and the diversity of vehicles encountered across the nation. It describes the size and nature of differences in the volume of vehicles by vehicle classification that are reported from data collection systems simply because different rules are in use to perform those classifications. Finally, part I discusses the effects on the load spectra computed when these different classification rules are used and what effect these differences have on the total traffic load computed for use in pavement analysis.

Part II examines the effect these differences in traffic load have on pavement design and analysis. The primary goal of these analyses is to determine the sensitivity of key pavement analyses to the errors that occur in traffic load estimates when the vehicle volume by classification estimates are produced from equipment using a State-supplied vehicle classification rule set and the load spectra used to produce that traffic load estimate employ the LTPP classification rules. Based on the analysis findings, the report provides recommendations on when load spectra collected as part of the SPS TPF on traffic data collection can be used at LTPP test sections that do not have valid site-specific load spectra.

Part III of the report describes recommended changes to the initially deployed LTPP classification rules. These refinements are designed to improve the ability of those rules to correctly classify some vehicles that tests show are not being correctly identified. Finally, part III presents the results of field tests of the refined LTPP classification rule set.

PART I. COMPARISON OF LTPP VEHICLE CLASSIFICATION Rules WITH Rules Used By OTHER STATES

Part I of this report introduces the FHWA vehicle classification system and describes how modern traffic data collection equipment converts available sensor outputs into estimates of traffic volume by FHWA vehicle classification. The report describes the differences in the rules to be used by LTPP for classifying vehicles and a number of other rule sets currently used by States. The report then examines the size and significance of the differences in volume counts by class of truck that result from the use of these different classification rule sets. It also examines how load spectra tables change as a result of how specific vehicle configurations are classified by different classification rule sets.

The report then combines these effects to gain an understanding of the size of traffic loading errors for pavement design that are created when the vehicle classification system used to generate the load spectra is different than the classification system used to collect the truck volume estimate used in the pavement analysis.

Finally, this section describes the recommended Mechanistic Empirical Pavement Design Guide (MEPDG) sensitivity tests (carried out in Part II of this report) that are needed to determine when the differences in load due to inconsistent classification data and/or missing site-specific load spectra are important for pavement analysis.