Structural Factors of Jointed Plain Concrete Pavements: SPS-2—Initial Evaluation and Analysis

Chapter 7. Summary, Conclusions, and Recommendations

The SPS-2 project, Strategic Study of Structural Factors for Rigid Pavement of Jointed Plain Concrete Pavements, is one of the key experiments in the LTPP program. The main objective is to determine the relative influence and long-term effectiveness of the JPC pavement strategic factors that affect its performance. There are some concerns about the ability of the SPS-2 experiment to meet those expectations, given that several SPS-2 sites were not constructed. In addition, at the SPS-2 sites that were constructed, some construction deviations and data collection deficiencies exist.

This study provides the first comprehensive review and evaluation of the SPS-2 experiment. Issues of experimental design, construction quality, data availability and completeness, and early performance trends are addressed in this report. The key findings are summarized in this chapter. This chapter also presents the research team's recommendations on improving the SPS-2 experiment and its data availability, expectations of the SPS-2 experiment, and future data collection and analysis topics.

SPS-2 Experimental Site Status

As of January 2000, 13 SPS-2 sites have been constructed throughout the United States. An additional site in California has been nominated and is currently under construction. The full factorial of the original experimental design is not completely satisfied by the constructed sites, leaving a portion (31 percent) of the desired inference space (subgrade and climate) with no experimental sites.

These 14 sites fill in 11 of the 16 SPS-2 experiment factorial cells and provide good coverage of major areas of the United States. Five sites (31 percent) are missing from the original experiment design, as listed below:

• Two sites in a wet no-freeze climate (southeast United States) with a coarse-grained subgrade. Data from each site will fill half of the design factorial.
• One site in a dry freeze climate (northwest United States) with a fine-grained subgrade. Data from each site will fill half of the design factorial.
• Two sites in a dry no-freeze climate (southwest United States), one on a fine-grained subgrade and the other on a coarse-grained subgrade. Data from each site will fill half of the design factorial.

Five additional SPS-2 sites (60 sections total) are needed to complete the design factorial. While it is impossible to determine the exact effects at this time, the lack of data from these missing sites will limit the results obtainable from the SPS-2 experiment, as summarized below:

• There will be no performance data and, thus, no performance findings from the missing sites.
• The missing site corresponding to the site in Arizona (dry/no-freeze and coarse subgrade) will make it impossible to determine (without confounding factors) the main effects and interactions for the dry/no-freeze climate.
• The missing section at the Nevada site will reduce the findings for that site and corresponding cell (although the Washington site appears to be a replicate).
• Adequate SPS-2 sites exist in wet-freeze climates, making a full inference space of performance data available. All main effects and interactions in this climate should be ascertainable.
• The number of SPS-2 sites in no-freeze climates, both wet and dry areas, is deficient. There will be difficulties in determining the main effects and interactions in these climates.

Some of these deficiencies can be overcome through use of mechanistic data analysis. However, there is no mechanistic analysis method that considers all factors involved, and the missing data cells will always limit in the verification and calibration of any performance models.

In addition to the 12 core sections at each site required by the SPS-2 experiment, SHAs have built a total of 40 supplemental sections with SPS-2 projects. The main value of the supplemental sections will be as a direct comparison to the core sections. For example, one supplemental section in Washington did not have dowels, and this will provide a direct comparison to a similar design with dowel bars. Various other comparisons are possible, including skewed joints, base types, subdrainage, slab thickness, asphalt concrete pavement, jointed reinforced concrete, special dowels, and variable slab thickness. These sections are valuable to the States, and efforts should be made to ensure that their construction and monitoring data are collected and stored.

Data Availability and Completeness

The SPS-2 project is an extremely valuable source of performance data for modern concrete pavements. The performance of these sections will be of great interest to all highway agencies building concrete pavements, and of course to the concrete industry as well. This report will be the major reference document for all future SPS-2 research studies, and will be the major reference document for the 14 SHAs that have constructed SPS-2 project sites.

Data elements that were considered to be essential to the SPS-2 experiment analyses have been assessed. The data availability and completeness for the SPS-2 sites are considered good overall. A high percentage of the SPS-2 data is at level E-greater than 82 percent for all data types, and greater than 99 percent for many.

However, a significant amount of data is still missing, especially traffic, distress and faulting surveys, and key materials testing data. These deficiencies need to be addressed before serious analysis of the SPS-2 experiment can occur.

The SPS-2 data deficiencies are summarized below:

• Wisconsin-newly constructed, data processing under way and all data expected to be complete.
• Arizona, Arkansas, and North Carolina-late initial survey for most monitoring types. Backcasting of IRI and distress data will be required.
• Colorado and North Dakota-late initial survey for one monitoring collection activity, either longitudinal profile measurements, deflection testing, faulting, or distress data.
• Kansas SPS-2-very deficient faulting data. Faulting measurements currently under way.
• Traffic data are very deficient for 5 of 13 sites (40 percent).
• Joint faulting data are not being collected as required by LTPP, and this will limit the analyses that can be conducted.
• Arkansas, Kansas, North Carolina, and Wisconsin are missing significant PCC materials testing data.

Note that the LTPP program is embarking on a systemwide effort to resolve all missing data. Some missing data have already been obtained. This effort will greatly improve the data availability for future analysis.

Experimental Design versus Actual Construction

Required experimental design factors were compared with the actual constructed values stored in the IMS database. This database includes both the site condition factors and pavement design features. Most SPS-2 sections follow the experiment design for the large majority of the design factors. Most deviations from the experiment design are found in the concrete slab thickness and 14-day flexural strength. A number of sections were more than 12 mm too thick or too thin, or were more than 10 percent too high or too low in flexural strength.

Out of the 13 SPS-2 project locations, only the recently constructed Wisconsin site does not have enough data in the IMS database to be evaluated. These data are currently in the pipeline. Of course, data from the California site under construction also are not in the database at this time.

Eight projects can be characterized as good to excellent when comparing designed versus constructed values, while the remaining four projects are considered poor to fair.

In addition to the comparison of the designed versus constructed values from the IMS database, construction reports of from all 13 SPS-2 reports sites were reviewed to identify construction deviations and difficulties. Four projects were found to have experienced more than minor construction difficulties or deviations.

The following summarizes significant deviations when comparing the designed versus constructed factorial factors, and more than minor construction deviations:

• Delaware-significant deviations in both slab thickness and flexural strength. This project is also considered to have moderate construction deviations.
• Iowa-significant deviations in flexural strength, slab thickness, and lane width. The lane width deviation may be a data entry error. This project is also considered to have moderate construction deviations.
• Michigan-moderate construction deviations.
• Nevada-significant deviations in subgrade type, flexural strength, and slab thickness. This project is also considered to have moderate construction deviations. Extensive cracking occurred early, and one section has been taken out of service.
• Washington-significant deviations in flexural strength and slab thickness. The traffic level for 1997 appears to be erroneous.

Early Performance Trends

The SPS-2 sections are relatively young, and a large majority show little distress. As of January 2000, only 43 of 155 sections (28 percent) are showing any noticeable distresses. Ninety-five percent of the SPS-2 sections have less than 1 mm of edge joint faulting. Eighty-seven percent of the SPS-2 sections show zero transverse cracking, and 78 percent of the sections have zero longitudinal cracking.

Based on the initial statistical analyses and comparisons, the following initial performance trends are noted (note that long-term performance may be different from short-term performance):

• The initial IRI of SPS-2 sections after placement ranged from 0.76 to 2.19 m/km with a mean of 1.30 m/km.
• JPCP constructed on coarse-grained soil were smoother (lower initial IRI) than those constructed on fine-grained soils.
• JPCP constructed on PATB were smoother than sections constructed on LCB or untreated aggregate base.
• The IRI trend over time depends heavily on the initial IRI, the traffic loadings, and the extent of joint faulting.
• Sections with PATB show the lowest total longitudinal cracking levels, while the sections with LCB show the highest longitudinal cracking.
• Thinner (203 mm) slabs show more longitudinal cracks. Sections with a thinner slab and widened slab show the highest level of longitudinal cracking.
• Sections with PATB show the lowest percentage of slabs cracked transversely, while the sections with an LCB show the highest transverse cracking.
• Thinner (203 mm) slabs show more transverse cracks than thicker slabs. Sections with a thinner slab and a widened slab show the highest level of transverse cracking.
• Sections with aggregate base show the highest joint faulting level. Sections with LCB and PATB have the lowest joint faulting.
• Widened slab sections show less faulting than conventional width slabs.

States' Expectations from the States for the SPS-2 Experiment

Two national workshops were held where input was received from the States on the SPS-2 project. The meetings were held on November 2-3, 1999, in Columbus, OH, and on April 27, 2000, in Newport, RI. The research team made presentations at both conferences about the status of SPS-2 data collection, analysis, and availability, and the near- and long-term LTPP products. Several participating States made presentations on the status and analyses of their SPS-2 projects, as well as the States' expectations of the SPS-2 experiment. There were many discussions of the future directions of the SPS-2 experiment and the analyses of the SPS-2 data at both conferences. Those discussions are summarized below.

In general, the States are satisfied with the SPS-2 experiment and fully expect to get valuable information about different design features from the project. Many States have been conducting or planning their own analyses of their SPS-2 projects. Some analyses have already yielded useful results. The States would like to see a focus on implementation of SPS-2 findings as they evolve over time.

First and foremost, what the States want to know the effect on pavement performance and cost-effectiveness of the SPS-2 experimental factors, such as:

• Drainage and base type.
• Widened lanes.
• Slab thickness.
• Concrete strength.

In addition to the structural design features, the States also would like to know what major site condition factors influence performance of concrete pavement, including:

• Climate.
• Traffic volume and loading.
• Subgrade type properties.
• Embankment.

Other specific expectations from the States include:

• Evaluation of existing performance prediction equations (i.e., AASHTO formulas).
• Validation of pavement analysis models.
• Mechanistic/empirical calibration of lab materials properties, backcalculated materials properties, and WIM data.
• Effects of soil type, base type, drainage, and climate on long-term subgrade moisture gradients.
• Cost-effectiveness of drainable bases, underdrains, high-strength concrete, different base types, and other features.
• Dynamic load response of the concrete pavements (response of PCC to single, tandem, and tridem axles; effect of vehicle speed on dynamic response of PCC).
• Using stiffness rather than density for subgrade acceptance.

As to the future analysis of SPS-2 data, the States believe that it is worthwhile to complete the missing data (backcast if necessary) to obtain traffic and materials data. Participants believed that many fundamental studies can be conducted to see how SPS-2 sections are responding to load and environmental stresses. It was also suggested that an integrated analysis plan is needed for future research.

Can SPS-2 Meet Expectations?

The specific experiment expectation of the SPS-2 project was to determine the main effects and interactions of the following key design features:

• Slab thickness.
• Concrete strength (14 day).
• Base type including PATB, LCB, and untreated aggregate base.
• Lane width.

These main effects and interactions were to be determined for each of the following subgrade and climatic conditions:

• Fine-grained and coarse-grained subgrade soils.
• Wet freeze, wet no-freeze, dry freeze, and dry no-freeze climates.

This evaluation has identified several significant problems that will limit the results that can be obtained from the SPS-2. Specifically, SPS-2 project sites are missing for certain subgrade-climate combinations. Some SPS-2 sites had construction deviations. Significant materials data and traffic data are missing from some sites or sections. The missing traffic data and key materials data must be obtained or forecasted before meaningful global analysis can be performed. The Nevada site has excessive early cracking that will limit its usefulness.

However, these problems do not mean that many important and useful findings and results cannot be obtained from SPS-2 experiments. Some interesting and important early trends have already been identified that will be useful in the design and construction of JPCP, even though the sections are only a maximum of 7.5 years old. As time and traffic loadings accumulate on the SPS-2 sites, additional valuable performance data will be obtained.

Because of FHWA's intense ongoing effort to obtain missing data (construction, materials, traffic, and monitoring), valuable results can be obtained from the SPS-2 sites. It is further believed that even more results can be obtained if a concerted effort is made to perform proper analyses of the data.

Recommendations for the SPS-2 Experiment

Missing SPS-2 Sites. To complete all the SPS-2 factorial cells, construction of the following sites is recommended:

• Two sites are needed to complete the factorial in the dry no-freeze climate: one site with a fine-grained subgrade, and the other site with a coarse-grained subgrade (i.e., Arizona, California in the southwest United States).
• Two sites are needed to complete the factorial and match sites in the wet no-freeze climate with a coarse-grained subgrade (southeast United States).
• One site is needed to complete the factorial and match a dry freeze climate with a fine-grained subgrade (north central United States).

Missing SPS-2 Data. Significant effort is recommended to obtain the following missing data:

• Materials-PCC strength.
• Traffic-5 sites completely missing traffic data, 11 sites missing continuous WIM data.
• Faulting-collect faulting data from several SPS-2 sites immediately.

Expectations from SPS-2. The overall objective is for the SPS-2 project's results to provide SHAs with documented findings to help them improve their management, design, construction, and materials procedures related to JPCP. The following specific information is expected to be gained from the SPS-2 project:

Specific design, subgrade, climate, and traffic effects

• Effect of subdrainage on performance (faulting, transverse cracking, longitudinal cracking, IRI).
• Effect of widened slab on performance (faulting, transverse cracking, longitudinal cracking, IRI).
• Effect of base type (lean concrete, permeable asphalt, dense aggregate) on performance (faulting, transverse cracking, longitudinal cracking, IRI).
• Effect of concrete slab thickness on performance (faulting, transverse cracking, longitudinal cracking, IRI).
• Effect of 14-day concrete strength on performance (faulting, transverse cracking, longitudinal cracking, IRI).
• Effect of climatic region on performance (precipitation, temperature).
• Effect of subgrade soil on performance (fine-grained, coarse-grained).
• Interactive effect of subdrainage, widened slab, base type, slab thickness, concrete strength, climatic region, and subgrade soil on performance.
• Effect of traffic loading on performance of various design treatments.

Data for use in calibration of mechanistic-empirical distress models

• 2002 Design Guide distress models.
• Subsequent improvement in future versions of the guide over time.

Data for use in empirical performance modeling (for pavement management)

Data for use in a variety of mechanistic modeling (backcalculation, structural analysis)

Data for use in a variety of cost/benefit analyses

• SPS-2 performance data are ideally suited for use in cost and benefit studies to determine the relative cost-effectiveness of each design feature in various climates and subgrades.

Future Data Collection. It is recommended that the following areas receive special emphasis in SPS-2 data collection:

• Routine data collection.
  • WIM and AVC traffic monitoring: ensure that LTPP guidelines are followed.
  • Joint faulting: follow LTPP guidelines closely.
  • Resolve irregular distress measurements over time for each SPS-2 section (variations of distress quantities over time).
  • Thermal coefficient of expansion of concrete: ensure that all SPS-2 sites are tested in 2000.
• Collect new data required for 2002 Design Guide calibration.
  • Slab curvature measurements: measure slab curvature when the thermal gradients are zero. Measure this during two seasons of year, wet and dry.
  • Conduct video surveys of edge drains to ensure they are working, and schedule maintenance if needed.
  • Cores along the cracks in JPCP to determine the initiation of the crack and the direction of its propagation. In other words, where did the crack initiate: top-down or bottom-up?

Recommended Future Analyses for SPS-2 Experiment

As stated previously, a very small percentage of the SPS-2 test sections currently have significant levels of distress, and only a few have been taken out service. The real benefit from this experiment will occur over the next 15 years, as more and more test sections exhibit higher levels of distress, magnifying the effect of the experimental and other structural factors on performance.

This report focuses on the quality and completeness of the SPS-2 construction and monitoring data and on the adequacy of the experiment to achieve the original expectations and objectives. Detailed analysis of the effect of different design factors on performance was outside the scope of work for this study. Thus, future studies using the SPS-2 experimental data should be planned and prioritized so they can be initiated as the SPS-2 projects exhibit higher levels of distress.

These future studies should be planned for in two stages that focus on local and national expectations from the experiment. The first stage is to conduct a detailed assessment or case study on each experimental cell in the project (companion SPS-2 sites that define a full factorial experiment) to ensure data adequacy, assess construction deficiencies, and support local interests and expectations. The second stage evaluates the effect of different structural features across the entire national experiment. Both analysis stages are briefly discussed in the following sections. After the sections are 15 to 20 years of age, a third-stage analysis will ultimately be needed to fully reap the benefits of the SPS-2 experiment.

Initial Stage-Analysis of Individual Factorial Cells

Each major cell in the SPS-2 experiment consists of at least two companion projects. One of these companion projects contains experimental sections 1 through 12, and the other contains sections 13 through 24. These companion SPS-2 sites constitute a full factorial of design factors and make it possible to evaluate the main effects and interactions of each experimental factor for those site conditions. A detailed evaluation of the companion projects within each major cell should be completed as soon as possible to ensure that all of the data exist and are acceptable. The purposes of the case studies in the first stage are listed below:

• Resolve construction and monitoring data anomalies and experimental cell differences for those projects that changed cell locations from the original experiment design, as they relate to the specific cell in the experiment.
• Conduct comparative analyses of the individual test sections at each site, including the supplemental test sections, to identify differences in pavement performance and response. These comparative studies should include performance measures, material properties, and as-built conditions.
• Determine the effect of any construction difficulties and problems and material noncompliance issues with the SPS-2 project specifications, if any, on pavement performance and response at each site.
• Develop findings regarding comparisons made between the companion projects and test sections and prepare a case study report that will be useful for the SHAs involved. Such information will also be useful for the national studies.

This first-stage analysis is considered absolutely essential prior to initiation of the second-stage analyses.

Second Stage-Analysis of Experimental Findings

The second-stage analyses should not be pursued until the first-stage analysis has been completed. It is expected that the analyses performed at this stage will be coordinated with the Strategic Plan for LTPP Data Analysis. The SPS-2 experiment can contribute to the following specific analyses outlined in the strategic plan:

• Develop relationships to enable interchangeable use of laboratory- and field-derived material parameters (Strategic Plan No. 2B).
• Establish procedures for determining as-built material properties (2C).
• Identify quantitative information on the performance impact of different levels of material variability and quality (2D).
• Estimate material design parameters from other materials data (2E).
• Quantify information as to the relationship between as-designed and as-built material characteristics (2F).
• Develop recommendations for climatic data collection to adequately predict pavement performance (3D).
• Develop models relating functional and structural performance (4C).
• Calibrate relationships (transfer functions) between pavement response and individual distress types (5C).
• Identify quantitative information on the impact of design features on measured pavement responses (deflections, load-transfer, strains, etc.) (7A).
• Identify quantitative information on the impact of design features on pavement distress (7B).
• Develop guidelines for the selection of pavement design features (7C).

In summary, the following future analysis objectives are recommended for the SPS-2 experiment. These analysis topics are discussed in more detail in figures 25 through 31.

1. Perform site-by-site analyses of SPS-2 projects to resolve data problems and gain understanding of performance of individual test sections (figure 25).
2. Determine the effect of the SPS-2 experimental factors on the performance of the jointed plain concrete pavements (figure 26).
3. Determine the optimum JPCP design features for specific site conditions and traffic loading (figure 27).
4. Determine the effect of concrete slab thickness variations on LTPP and initial ride quality (figure 28).
5. Calibrate and validate relationships (transfer functions) between pavement structural response and individual distress types (figure 29).
6. Conduct mechanistic analyses of SPS-2 sites (particularly Ohio and North Carolina) to gain knowledge of critical stresses and deflections to explain their performance in terms of joint faulting and slab transverse and longitudinal cracking (figure 30).
7. Conduct cost-benefit analyses of SPS-2 data to determine the cost-effectiveness of various design features (figure 31).

The full results from the SPS-2 experiment will require 20 years of monitoring for the majority of sections. Additional studies beyond these proposed will be required.