Investigation of Increase in Roughness Due to Environmental Factors in Flexible Pavements Using Profile Data From Long-Term Pavement Performance Specific Pavement Studies 1 Experiment

CHAPTER 1. INTRODUCTION

This chapter provides the background for this study, including a general description of the Long-Term Pavement Performance (LTPP) Program; information on how pavement profile information has been collected for the program over the years; a discussion of the Specific Pavement Studies (SPS) 1 experiment, the data from which was used in this study; and a description of the objectives of the current study.

LTPP PROGRAM

The LTPP Program is a research program that investigates in-service pavement performance. Started in 1987 as part of the first Strategic Highway Research Program, the LTPP Program has been managed by the Federal Highway Administration (FHWA) since 1992. The primary goal of the LTPP Program is to determine how and why pavements perform as they do. To accomplish this goal, the LTPP Program monitors test sections established on in-service roads by collecting a variety of data on these test sections such as pavement distress, longitudinal profile, and deflection data at regular intervals using standard data collection protocols. The collected data are stored in the Pavement Performance Database (PPDB) and can be used by pavement engineers and researchers worldwide to advance the science of pavement engineering.⁽¹⁾

The LTPP Program consists of two complementary kinds of research: General Pavement Studies (GPS) and SPS experiments. The study of the performance of in-service pavement test sections that were either in their original design phase or in their first overlay phase has been addressed in the GPS experiment. The effect of specific design features on pavement performance has been addressed in the SPS experiment.

PROFILE DATA COLLECTION FOR THE LTPP PROGRAM

From the start of the LTPP Program, the longitudinal profiles along the two wheelpaths at test sections have been collected using an inertial profiler. The LTPP Program has always used four inertial profilers, each operated by a regional contractor, to collect profile data at the test sections. Data at test sections located in a specific geographical area have been collected by a single regional contractor. From the start of the LTPP Program until November 1996, profile data at test sections were collected using K.J. Law DNC 690 inertial profilers. In December 1996, these profilers were replaced with K.J. Law T-6600 inertial profilers. In September 2002, the K.J. Law T-6600 profilers were replaced with International Cybernetics Corporation (ICC) MDR 4086L3 inertial profilers. In April 2013, the ICC profilers were replaced with Ames Engineering Model 8300 inertial profilers.

The K.J. Law DNC 690 inertial profiler only collected profile data along the two wheelpaths. Starting with the K.J. Law T-6600 inertial profiler, all profilers used in the LTPP Program have collected profile data along the center of the lane in addition to collecting data along the two wheelpaths.

When collecting data with an inertial profiler, seven to nine repeat runs have typically been performed at an LTPP section during a site visit. After performing quality control checks on the data, ride quality parameters computed from the profile data and the profile data corresponding to five repeat runs for each site visit have been uploaded to the PPDB. The International Roughness Index (IRI) has been one of the ride quality parameters computed from the profile data that has been uploaded to the PPDB.⁽¹⁾

The K.J. Law DNC 690 profiler collected profile data at 1-inch intervals and then applied a 12‑inch moving average to these data and recorded the profile data at 6-inch intervals. The recorded profile data as well as ride quality parameters computed from these data for the five selected profile runs for each site visit were uploaded to the PPDB.⁽¹⁾

The K.J. Law T-6600 profiler and the ICC profiler recorded profile data at 0.98-inch intervals. The left and right wheelpath profile data collected by these devices were processed by applying an 11.8-inch moving average to the data, and then data at 5.9-inch intervals were extracted. The 5.9‑inch interval profile data for the two wheelpaths and the ride quality indices computed from these data for five selected runs for each site visit were uploaded to the PPDB. In addition, data files conforming to the University of Michigan Transportation Research Institute Engineering Research Division (ERD) file format that contained the 0.98-inch interval profile data were created for all profile runs obtained at a site. These ERD files contained the 0.98-inch interval profile data for the left wheelpath, right wheelpath, and center of the lane and were stored in the LTPP Ancillary Information Management System (AIMS). These ERD files could be requested for analysis through the LTPP Customer Support Service.⁽²⁾

The Ames Engineering profilers also recorded profile data at 0.98-inch intervals. The left wheelpath, right wheelpath, and center of the lane profile data collected by these devices were processed by applying an 11.8-inch moving average to the data and then extracting data at 5.9‑inch intervals. The 5.9-inch interval profile data were used to compute the IRI of the left wheelpath, right wheelpath, and center of the lane. The 5.9-inch interval profile data and the IRI values for the three paths for five profile runs for each site visit were uploaded to the PPDB. In addition, profile data collected at 0.98-inch intervals along the left wheelpath, right wheelpath, and center of the lane for these five runs were also uploaded to the PPDB.

THE LTPP SPS-1 EXPERIMENT

The LTPP SPS-1 experiment was developed to investigate the effect of selected structural factors on the long-term performance of flexible pavements that were constructed on different subgrade types and in different environmental regions. New pavements were constructed for the SPS-1 experiment. In the SPS-1 experiment, 12 test sections were constructed at a project location. Each test section was 500 ft long with a transition area between the test sections. The pavement structure of the test sections in the SPS-1 experiment are shown in table 1. The 12 test sections in an SPS-1 project were either section numbers 1 through 12 or section numbers 13 through 24. These test sections have been referred to as core test sections at an SPS-1 project. In addition to these core test sections, some State transportation departments constructed other test sections referred to as supplemental sections at the SPS-1 project location to study factors that were of interest to the State transportation departments.

The structural factors considered in the SPS-1 experiment have been asphalt thickness, base type, base thickness, and drainability (presence or lack of it as provided by an open-graded permeable asphalt-treated layer and edge drains). Five different base types have been used in this experiment: dense graded aggregate base (DGAB), asphalt-treated base (ATB), ATB over DGAB, permeable asphalt-treated base (PATB) over DGAB, and ATB over PATB. The subgrade types considered in this experiment were classified as fine- and coarse-grained, and the environmental regions considered were the four LTPP environmental regions, which were wet-freeze (WF), wet-no freeze (WNF), dry-freeze (DF) and dry-no freeze (DNF).

Table 1. SPS-1 test sections.

Eighteen SPS-1 projects were constructed for the LTPP Program. Table 2 shows the States where the SPS-1 projects were constructed, the State code, test section numbers constructed for each project, and the date the project was opened to traffic.

Table 2. SPS-1 projects.

The profile data collection at an SPS-1 project has typically been performed by collecting data along the entire SPS-1 project, which means data have been collected over the test sections as well as within the transition areas between the test sections. Thereafter, the data corresponding to each test section have been extracted from the collected data using software that employed the stationing associated with each test section. This process has been referred to as subsectioning in the LTPP Program.

The first profile measurements at an SPS-1 project were performed within a short time after construction. Thereafter, profile data at SPS-1 test sections were collected at regular intervals following the guidelines established by the FHWA. Profile data collection at a test section ended when the test section was taken out of the LTPP study because of rehabilitation. A test section is said to have been deassigned when it is taken out of the LTPP Program.

OBJECTIVES OF THIS STUDY

State transportation departments use the mean IRI (MIRI), which is the average of the left and the right wheelpath IRI, to monitor the roughness of their pavement network. The change in roughness of a pavement segment over a specified time interval can be evaluated by determining the change in MIRI over that period. The change in roughness over time along the wheelpaths of a pavement segment occurs because of the change in the profile of the wheelpaths over time.

As described previously, profile data along the center of the lane as well as along the wheelpaths were collected at the LTPP test sections starting in December 1996. The change in the profile along the center of the lane in flexible pavements was expected to be affected mainly by environmental effects. The only traffic the center of the lane received was when vehicles change lanes, and such maneuvers were expected to apply only minimal traffic to the center of the lane. Environmental effects can cause changes in the moisture content of the subgrade from the as-constructed value, which can cause the subgrade to shrink or swell. This can affect the profile of the pavement and cause a change in roughness. Freezing temperatures can cause frost heave, which can also affect the pavement profile and cause an increase in roughness. Therefore, the interaction between environmental effects and subsurface layers can cause a change in the profile of a pavement, thereby increasing the roughness of a pavement. In flexible pavements, transverse cracking can occur because of thermal movements induced on the AC surface, and this cracking can also increase the roughness. Hence, along the center of the lane, transverse cracking and the interaction between environmental effects and subsurface layers that cause a change in the profile can cause an increase in roughness.

Along the wheelpaths of a flexible pavement, fatigue cracking and rutting caused by traffic can contribute to an increase in roughness. Therefore, the increase in roughness along the wheelpaths can be attributed to the change in the profile along the wheelpaths caused by traffic loadings and environmental effects. When evaluating the changes in roughness that have occurred along the wheelpaths, environmental effects could not be separated from traffic effects because the collected profile showed the consequences of both factors. However, the profile data collected along the center of the lane in a flexible pavement could be used to evaluate the change in roughness due only to environmental effects. On a flexible pavement, the change in roughness along the center of the lane could be compared with the change in roughness along the wheelpaths to evaluate the contribution of environmental factors to the increase in roughness along the wheelpaths. The profile data collected at SPS-1 projects have provided an excellent dataset to perform this type of analysis for flexible pavements because the effect of environmental factors on the change in roughness could be evaluated for a variety of pavement structures.

The goal of this study was to evaluate the changes in IRI along the center of the lane over time and to identify subgrade and environmental parameters that have contributed to an increase in the center lane IRI. The specific objectives of this study were to use the data collected at SPS-1 test sections to do the following:

• Analyze the center of the lane IRI (CLIRI) and MIRI at test sections to evaluate changes over time.
  
  
• Compare the change in CLIRI with the change in MIRI at the test sections.
  
  
• Identify subgrade and environmental parameters that influence the changes in the center of the lane profile that contribute to an increase in CLIRI.
  
  
• Perform a detailed evaluation to identify how CLIRI changes for a select group of test sections and to determine whether wavelengths that contribute to the increase in CLIRI can be identified.
  
  
• Develop models to predict the change in CLIRI at the test sections.

This study was intended to provide information on how environmental conditions interact with subgrade conditions and influence the increase in roughness for the different pavement structures present on an SPS-1 project.