Estimation of Key PCC, Base, Subbase, and Pavement Engineering Properties From Routine Tests and Physical Characteristics

PDF Version (4.44 MB)

PDF files can be viewed with the Acrobat® Reader®

CHAPTER 2. MATERIAL PROPERTIES FOR PREDICTIVE EQUATIONS

The primary engineering material properties considered for indepth evaluation in this study were those required for pavement analysis and design using the MEPDG. The MEPDG considers the effects of a comprehensive set of material properties in the structural design of JPCP and CRCP. Its capability to consider strength, modulus, thermal, and other materials properties is the foundation for designing for performance under traffic loads and climatic conditions. Also, the MEPDG procedure can accommodate various material types and uniquely model each material’s response to load, temperature, and moisture and predict their effects on performance. Therefore, the research approach was to develop a list of material properties that likely could be estimated based on the availability of data in the LTPP database and the needs of the MEPDG performance prediction models.

INPUTS FOR MEPDG

Hierarchical Inputs for MEPDG

The MEPDG adopts a hierarchical input level scheme to accommodate the designer’s knowledge of the input parameter. Inputs can be provided at three different levels. Level 1 inputs represent the greatest knowledge about the input parameter and typically are obtained from a project-specific data collection or test effort. Level 2 represents a moderate level of knowledge of the input parameter and is often calculated from correlations with other site-specific data or a less expensive measure. Level 3 represents the least knowledge of the input parameter and is based on “best-estimated” or default values. For example, level 1 data for concrete flexural strength would involve a flexural beam test, level 2 would be a flexural strength value estimated using a compressive strength test and correlation to flexural strength, and level 3 would be a default value for concrete strength used by a particular SHA. Most agencies have adequate information in their materials and construction quality databases to develop agency-specific default values for immediate implementation of the MEPDG (based on current knowledge and surveys conducted by Rao et al.).⁽¹⁰⁾

During the development of the MEPDG, the need for correlation equations to determine some input values was recognized. Many designs must be created years in advance of construction, and little is known of the exact materials that will be used. However, it is highly desirable to have the best estimates of these inputs possible based on available information. The MEPDG therefore supports the use of level 2 or 3 data in the absence of level 1 laboratory test data. This adaptability is critical to the model types developed in this study and is further discussed in chapter 5 of this report.

Input Categories for the MEPDG

Table 1 provides a list of major material types considered in the MEPDG. Bolded information reflects the material types that are relevant in this project. Each material type requires a variety of material inputs (not all easily available) during local calibration efforts and after the procedure is implemented. As an example, the various PCC material-related inputs considered by the MEPDG are presented in table 2 under the following three categories:

• Material properties required for computing pavement responses.
• Additional material inputs to the distress/transfer functions.
• Additional material inputs required for climatic modeling.

Table 1. Major material types for the MEPDG.⁽²⁾

Note: Bolded information reflects the material types that are relevant in this project.

Table 2 reflects the additional significance of other material engineering properties beyond strength properties for PCC materials in the analysis process. As an example, while concrete modulus of rupture (MR) was the main material input for the AASHTO 1993 rigid pavement design procedure (along with the modulus of elasticity), the MEPDG allows correlations through level 2 inputs with compressive strength and requires other volumetric properties such as shrinkage, CTE, specific heat, and thermal conductivity for analysis.⁽¹²⁾ In addition, strength parameters that are used in the analysis include compressive strength, modulus of elasticity, and tensile strength for CRCP. The modulus of elasticity has a much greater effect on performance with the MEPDG than with the AASHTO 1993 procedure. In other words, the MEPDG offers a framework to optimize mix designs to balance a range of strength, modulus, CTE, shrinkage, and other engineering properties for improved performance.⁽¹³⁾

Table 2. PCC material inputs beyond strength considered by the MEPDG for JPCP and CRCP.

*Estimated from compressive strength, cement type, curing type, cement content, and w/c ratio.

**Estimated from cement content and mean monthly temperatures at the project location.

Likewise, unbound materials are characterized by material parameters that account for the changing stress state in the material with seasonal changes in moisture conditions in a specific location. The gradation of the soil, maximum dry density, and optimum moisture content are key inputs to the procedure.

The MEPDG also requires the input of construction and field-specific parameters that are critical to performance. These construction or site features are not restrictive to a particular material, but they are associated with specific material index properties, climatic conditions, and construction practices. For example, base erodibility is a function of base material properties such as the strength of the base layer, the amount of fines, and site conditions such as the level of precipitation and traffic load repetitions.

Correlations Developed/Adapted for the MEPDG

The calibration of the rigid pavement distress models utilized several inputs from levels 2 and 3 based on the best information available from literature and LTPP databases. The following is a partial list of correlations used in the MEPDG models:

• The default concrete strength gain models in the MEPDG were derived from two decades of Portland Cement Association (PCA) research and the LTPP Specific Pavement Studies (SPS)-2 time series data.
• The ultimate shrinkage calculation model is based on Bazant’s model and was derived as a function of compressive strength, cement type, curing type, cement content, and w/c ratio.(14) Other shrinkage inputs, such as time to 50 percent of ultimate shrinkage, were adopted from the American Concrete Institute (ACI) recommended models.
• CTE defaults by coarse aggregate type were established based on testing and petrography performed under the LTPP program.⁽¹⁵⁾
• Correlations between several of the PCC strength and modulus terms are based on work done under the LTPP program.⁽¹⁶⁾
• Erosion of the base layer under a CRCP was calculated using an empirical model that considers the climatic conditions and the properties of the base layer to determine the extent of erosion and voids under the slab.
• Erodibility Index (EI) for JPCP design is based on the base type and was established during the calibration process.
• Resilient modulus at optimum moisture content, which is a key material input for unbound layers, was determined through an iterative process by matching the subgrade k-value backcalculated from falling weight deflectometer (FWD) test data with the subgrade k-value for the same calendar month as determined by the MEPDG analysis.
• Friction coefficients of the slab/base layer in CRCP (mean friction coefficient for each type of base course) were determined by matching the predicted and field measured transverse crack spacing.
• ^deltaT in rigid pavement design was set at -10 °F for the calibration of the MEPDG distress models. Since it was not possible to determine the actual magnitude of this parameter, it was found that a value of -10 °F optimized the transverse cracking prediction over the numerous calibration sections. The MEPDG recommends -10 °F for this parameter as a default value and is more or less considered a level 3 default.

SELECTION OF MATERIAL PARAMETERS

A preliminary list of material properties was prepared for developing predictive models based on inputs required for the MEPDG and their level of significance in performance prediction as well as their importance during the design, construction, or pavement management phases. The materials are classified broadly as PCC materials, stabilized materials, and unbound materials. Note that unbound materials include both coarse-grained and fine-grained soils, which have different mechanical behavior in response to applied stress states. The preliminary material properties are listed in table 3 through table 5 for PCC, stabilized, and unbound materials. These tables list all input variables, identify the conventional source of data for SHAs, and indicate whether a predictive model can be developed for the parameter.

Table 3. PCC material properties and rigid pavement design features considered for generating predictive models.

N/A = Not applicable. *Parameter was not selected for model development. **Considered a design feature input to rigid pavement design process.

Table 4. Chemically stabilized materials properties considered for generating predictive models.

Table 5. Unbound material properties considered for generating predictive models.

N/A = Not applicable. *Parameter was not selected for model development.

In selecting the material properties that require predictive models, the following factors were considered:

• The selected parameters are not index properties of the material or they are not part of mix design information and hence would require a predictive model to determine its value. In other words, these parameters are not independent variables for the material, and their values depend on the individual properties of constituent materials.
• Routine test procedures are not sufficient to determine the value of these parameters, and they require either expensive or time-consuming tests to determine their values. CTE and resilient modulus of unbound base materials are examples of such parameters.
• The material parameter exists in the LTPP database, and the required data are available for developing the model. For example, PCC shrinkage and modulus of stabilized base materials were not part of the database. In fact, they were never intended to be included in the LTPP tests when the experiments were designed.
• The engineering property is of significance to pavement performance.
• The material property is not affected significantly by construction factors or by parameters that cannot be determined prior to the design stage. For example, the unit weight of a material is an important parameter for pavement performance. It contributes to the total stresses developed in the PCC slab and is an indicator of the quality of consolidation achieved during construction that has a direct impact on the strength and durability characteristics. However, unit weight cannot be predicted using index properties, as it is also a function of aggregate gradation and packing achieved in the field.
• The likelihood of actual value of the material property deviating from the assumed defaults and its implications on design or performance is significant. For example, Poisson’s ratio of PCC or specific heat of PCC would not qualify as an engineering property requiring a predictive relationship because these parameters do not vary significantly from the recommended/typical default values. Additionally, within a realistic range, their effects on performance are not large.

Accordingly, several variables in table 3 to table 5 were not selected for model development.

The selected parameters from table 3 to table 5 as well as predictive models to estimate their values were the focus of the literature review performed in this study. The literature review sought to identify independent variables (index properties) that could be used to determine the values of the selected material parameters. A list of index properties that can serve as independent variables in the prediction models is provided at the end of chapter 3.