U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
REPORT |
This report is an archived publication and may contain dated technical, contact, and link information |
Publication Number: FHWA-HRT-12-030 Date: August 2012 |
Publication Number: FHWA-HRT-12-030 Date: August 2012 |
PDF Version (4.44 MB)
PDF files can be viewed with the Acrobat® Reader®
Material characterization is vital to pavement analyses and has received increasing focus as it forms a critical component in recent improvements to engineering practices. This pertains to all aspects of pavement engineering—analysis, design, construction, QC/QA, pavement management, and rehabilitation. At each stage during the life of a project, the influence of several fundamental engineering material parameters on the long-term performance of the pavement has been recognized. There is a greater emphasis for optimizing the performance of concrete pavements, which involves a detailed understanding of the variables that affect pavement behavior and the properties of concrete that correspond to the desired performance.
Consequently, there is a need for more information about material properties so that they can be characterized accurately for predicting performance or for verifying their quality during the construction phase. With limited resources for performing laboratory and field tests to determine material properties, the need for a secondary means to obtain these material property values (i.e., through correlations or predictive models based on data from routine or less expensive tests) is obvious. Additionally, the American Association of State Highway and Transportation Officials (AASHTO) MEPDG offers users the option of using inputs obtained through correlations.(2,4,1) The MEPDG defines level 2 inputs as those obtained from correlations between the primary inputs (level 1 measured) and other parameters that are material-specific or are measured through simpler tests. The LTPP database provides an excellent source of information to develop these correlations using material properties of field sections.(5)
The current report addresses critical data needs for design, construction, and pavement management operations under the LTPP Data Analysis Technical Support contract. This project focuses on developing predictive models to estimate PCC and unbound material properties using LTPP data.
Material property data needs in the context of this study are grouped into the following three categories:
In both empirical and mechanistic-empirical (M-E) design systems, material property inputs are essential to characterize pavement behavior and to predict pavement responses, such as the magnitudes of stress, strain, and displacement, when subjected to applied traffic loads and environmental conditions. Furthermore, major pavement distresses are associated directly with the material properties of a component (or layer) of the pavement structure. For example, in JPCP, transverse cracking is influenced by PCC flexural strength. Faulting can be related to the erodibility of the underlying base/subbase material. Punchout development in CRCP can be related to PCC tensile strength.
The MEPDG, developed under National Cooperative Highway Research Program (NCHRP) Project 1-37A and subsequently improved under NCHRP 1-40D, allows users to model the effects of project-specific climate, traffic loads, materials, design features, and construction practices mechanistically to predict pavement performance based on distress models calibrated with LTPP field sections.(2,4) The MEPDG is considered a significant improvement over current pavement design procedures, and in November 2007, it received the status of an AASHTO interim standard.(1)The publication User Manual and Local Calibration Guide for the Mechanistic-Empirical Pavement Design Guide and Software developed under NCHRP 1-40B provides guidance to State highway agencies (SHAs) that are considering implementing the MEPDG.(6) It is expected that SHAs will adopt locally calibrated distress models that are representative of their specific materials and design conditions.
The need for a variety of material inputs is being recognized as agencies evaluate the MEPDG and streamline efforts for implementation. They continue to face challenges in estimating material parameter inputs and understanding their impact on pavement performance. For example, agencies do not have measured test data or access to databases and the necessary engineering expertise to develop correlations for their needs. Furthermore, due to a lack of familiarity with several input categories, they have come to rely on default values to characterize their typical materials. These default parameters are often a gross approximation of the true value, which may lead to erroneous distress and International Roughness Index (IRI) predictions. As another example, the permanent curl/warp gradient in the national calibration was set at -10 °F through the slab, as it was not possible to obtain an accurate value for this parameter, which depends on construction-related conditions. Analysis of selected LTPP data made it possible to derive an improved way to estimate this important input for design.
This study provides much needed procedures to obtain several inputs and provide correlations to determine the whole range of material properties based on routine test results and physical characteristics. These correlations will supplement the User Manual and Local Calibration Guide for the Mechanistic-Empirical Pavement Design Guide and Software to support MEPDG implementation efforts.(6)
Pavement construction practices are being enhanced continually for faster and more efficient processes. In addition, new materials and material types are being introduced. For example, cement compositions and cement types have changed considerably over the years, resulting in PCC properties and durability characteristics that are different from the past.
The focus of QC/QA procedures is now on identifying more reliable and faster QC/QA tests and determining material properties that are related directly to performance. The MEPDG enables performance prediction of the as-built pavement in addition to that of the as-designed pavement, as long as deviations from design assumptions (material properties or construction practices such as curing or temperature during construction) are identified during the construction process (see figure 1). For example, although the density of an unbound material is a good indicator of construction quality, the more fundamental resilient modulus is an indicator of performance and is a key input to the MEPDG. The ability to predict resilient modulus from index properties measured during construction will make the QA process address both construction quality and pavement performance issues. Note that in figure 1, material properties measured during construction can be used to predict performance in the field and might be different from the design/target performance.
Also, performance-related specifications (PRSs) for concrete pavements have been developed in recent years. Irick et al., under a Federal Highway Administration (FHWA) study to demonstrate PRS system for rigid pavement construction, considered three key performance indicators: PCC strength, PCC slab thickness, and initial serviceability.(7) Several relationships for the prediction of PCC properties were evaluated under this study. Irick et al. provide a comprehensive literature summary of models to predict concrete strength parameters. PRSs have also been implemented on several projects that required many correlations between pavement properties and performance.(8,9)
Figure 1. Illustration. MEPDG performance prediction during the design and construction stage.
This study will bridge the gaps in current knowledge regarding the estimation of more fundamental material parameters that influence performance based on index properties or other commonly measured properties during construction.
One of the key needs in managing pavements is an estimation of remaining life. Several SHAs use this parameter to program rehabilitation treatments. Various models (including the MEPDG models) are useful in that they can be utilized to predict the remaining life until critical levels
of each distress and IRI are reached. Also, agencies are now considering the integration of construction quality databases with pavement management databases to track the effect of design and construction quality on long-term performance.(10) Such efforts lend themselves to more accurate performance predictions, whereby the performance of the as-constructed pavement can be used for scheduling maintenance and rehabilitation programs (see figure 1). However, many model inputs are needed related to the existing pavement, including inputs to characterize materials accurately. Index properties from construction QA data can be used to predict fundamental material properties that are related to performance.
In summary, the MEPDG provides a tool to specify material characteristics during the design and construction processes to achieve desired performance. The same models used in the MEPDG for design and construction analyses can be used in the future management of the pavement to estimate its remaining structural and functional life. For example, the inputs for a 10-year-old pavement could be measured from the existing pavement and estimated from the MEPDG models to project future slab cracking. The curve can be adjusted to match today’s actual performance to improve the prediction. The slab cracking curve can be projected into the future to determine when it reaches a critical value to estimate its remaining life. The same could be done with joint faulting and IRI. Therefore, the design, construction, and pavement management stages share a common need for determining a variety of material properties based on correlations from index properties and/or properties determined from more routine test procedures. This practice has been used in past AASHTO pavement design procedures and likely will increase in the future due to the more complex fundamental inputs required for the MEPDG procedure.
This study utilized LTPP data to develop correlations for SHAs to characterize PCC, subbase, and subgrade materials as necessary for design, QC/QA, and pavement management. It is expected that the findings from this study will assist an agency’s materials selection procedures, materials specification, pavement design, and pavement management practices.
To accomplish this overall goal, this research sought to address the following three objectives:
1. Identify a set of material engineering properties for which predictive relationships would be useful in pavement design, construction QC/QA, and pavement management applications.
2. Establish and/or validate relationships between these engineering properties and routine test results, index properties, and/or other readily available information.
3. Develop a practical guide accompanied by user-friendly software for applying the results of the aforementioned tasks in pavement design, construction QC/QA, and pavement management.
This project provides the necessary guidance for agencies to use more accurate values for material properties in design, construction, and pavement management. The correlations developed as part of this study are based on actual data from LTPP sections and, therefore, are more reliable than default or typical values currently being used. The full potential of the MEPDG to predict performance accurately can be realized by providing more accurate input values to the procedure. This also supports improvements in material specifications for use in pavement construction, particularly for PRS. Eventually, these models can be implemented into the PaveSpec PRS software and used for construction.(11) In particular, the following major benefits will be obtained from this study:
Most of the data used in the development of prediction models to estimate material properties were obtained from the LTPP Standard Data Release 23.0.(3)
This report documents the work performed under this project and presents the models developed to characterize materials and estimate material inputs. The report consists of five chapters. Chapter 2 describes the selection procedure to identify material properties that require predictive models. Chapter 3 provides a summary of literature reviewed for this study and concludes with a list of index properties (independent variables) used to characterize the material properties identified in chapter 2. Chapter 4 explains the data analyses procedures and discusses the developed models. Chapter 5 provides a summary of the report and presents the conclusions.