Local Calibration of the MEPDG Using Pavement Management Systems
Chapter 5. Preliminary Framework
Introduction
This chapter identifies the types of information a SHA needs to support its efforts to locally calibrate the MEPDG models using a data contained within a pavement management system. Also included are guidelines for the development of the MEPDG calibration database for storing needed MEPDG inputs. It is envisioned that the developed MEPDG calibration database will not duplicate the information contained in an existing database, but will establish a link for retrieving needed MEPDG input data.
The preliminary framework includes the following five steps:
- Identify the information that can be extracted from the pavement management system, as well as the types of design and as-built information (e.g. thicknesses, material types, as-constructed properties) that are required for calibration activities. Identify sources of information not provided through pavement management.
- Analyze and implement the data storage and backup methodology. The calibration of the MEPDG models will require the collection of additional data that is not typically included in a State pavement management system. A simple relational database table is recommended for storing the additional needed data.
- Link the created MEPDG calibration database with the State pavement management system database.
- Link the created MEPDG calibration database with other SHA databases.
- Outline how missing data related to traffic, climate, materials, and performance parameters could be obtained to support the local calibration effort of a single State.
The application of the framework to a SHA requires consideration of the following factors:
- Based on the results of previous research, the preliminary framework builds on the recommendation to develop a satellite database that combines pavement management and pavement design information on sections that are designed and constructed using the MEPDG. This approach is illustrated in figure 1. The framework identifies information that is expected to be extracted from a pavement management system, as well as the types of design and as-built information that should be obtained from other sources for calibration activities. Specific data requirements for annual measurements (including supplemental materials evaluation testing, actual climate data, maintenance histories, and observed traffic volumes) are also outlined in the preliminary framework.
- A data storage and backup scheme is required. Of all the data required for MEPDG calibration, only a portion of the data is typically stored within a State pavement management system database. These include, but are not limited to, the county, route, milepost, pavement layer descriptions (pavement types and thicknesses), treatment histories, and pavement condition survey data. The rest of the data, such as the construction related (e.g. air voids, compressive strength) data, materials and mix design data, and climatic data, are not typically contained within a State’s pavement management system database. To simplify the process of calibration, it is preferable to combine the various data into one MEPDG calibration database or establish a process for linking them together. For some types of relational databases, database links can be created so that the various databases can work like a single logic database.
Figure 1. Supplemental database approach for MEPDG calibration activities (FHWA 2006a).
- An approach for linking the created MEPDG calibration database with the State pavement management system database must be included. Generally, there are two different approaches to combine one relational database with another relational database or spreadsheet. The first method is to import the data from the other database or spreadsheets files. The second method is to link the data without importing them. The advantages and disadvantages of each approach are described later in this section. In general, databases are more useful for linking data and for retrieving records. However, many individuals are more comfortable using spreadsheets, which are especially useful for numeric computations. A disadvantage to the use of spreadsheets is that they can only handle simple data relationships.
- The approach to link the created MEPDG calibration database with other SHA databases (e.g., materials databases) needs to be analyzed and the most effective strategy determined. Because some of the required data may be stored as flat text files, a front-end application may be needed to process and import any data into the MEPDG calibration database.
- Guidelines for standard database management and maintenance techniques (e.g. quality control of data inputs, security, backups) are needed. In addition, since the MEPDG is an evolving software program, existing models may be modified and new models may be added. Guidelines for database modification to incorporate future models (and potential changes in data inputs), enhancements, and additions will be necessary.
- The use of a common referencing system will be critical for obtaining applicable data across multiple databases. It is recognized that the various departments within a SHA (such as pavement management, traffic, construction) may maintain their records according to different referencing systems. The ability to relate the various referencing systems to a single referencing system will be essential in the calibration process, thereby insuring that all data relates to the same roadway location on the State highway network.
Project Summary Module
The project-specific information used in a typical MEPDG run is presented in table 4. This information is not used directly in the calibration process but is necessary to define performance parameters (e.g., distresses) and reliability levels. The majority of this information should be available within a typical State pavement management system; however, some information (e.g. traffic opening date, design life) may need to be obtained from alternate sources.
Table 4. Project summary information.
| Description |
Variable |
HMA |
PCC JPCP CRCP |
Typical Data |
| Design properties |
Project name and description |
X |
X |
X |
Yes |
| Design life (years) |
X |
X |
X |
Assumed |
| Base/subgrade construction (date) |
X |
X |
X |
Maybe |
| Restoration/Overlay |
|
|
|
|
| Existing pavement construction (date) |
X |
X |
X |
Yes |
| Pavement restoration/overlay (date) |
X |
X |
X |
Yes |
| Traffic opening (date) |
X |
X |
X |
No |
| Site/project identification |
Location |
X |
X |
X |
Yes |
| Project ID |
X |
X |
X |
Yes |
| Section ID |
X |
X |
X |
Yes |
| Stationing (format, beginning and end) |
X |
X |
X |
Yes |
| Traffic direction |
X |
X |
X |
Yes |
| Analysis parameters (limit and reliability) |
Initial IRI (in/mi) |
X |
X |
X |
Yes |
| Terminal IRI (in/mi) |
X |
X |
X |
Yes |
| AC surface down cracking (ft/mi) |
X |
|
|
No |
| AC bottom up cracking (%) |
X |
|
|
Yes |
| AC thermal fracture (ft/mi) |
X |
|
|
Yes |
| Chemically stabilized layer fatigue fracture (%) |
X |
X |
|
No |
| Permanent deformation – total (in) |
X |
X |
|
No |
| Permanent deformation – AC only (in) |
X |
X |
|
Yes |
| Transverse cracking (% slabs cracked) |
|
X |
|
Yes |
| Mean joint faulting (in) |
|
X |
|
Yes |
| Existing punchouts |
|
|
X |
Yes |
| Maximum crack width (in) |
|
|
X |
No |
| Minimum crack load transfer efficiency (%) |
|
|
X |
No |
| Minimum crack spacing (ft) |
|
|
X |
No |
| Maximum crack spacing (ft) |
|
|
X |
No |
Traffic Module
The MEPDG utilizes axle load spectra as an input to the analysis process. The axle load spectra represent the hourly, daily, monthly, and seasonal distributions of the traffic with respect to axle type/load of various vehicle classes. This represents a major departure from the equivalent single axle loads (ESAL) concept that was used in previous American Association of State Highway and Transportation Officials (AASHTO) methodologies. Table 5 lists the required MEPDG traffic inputs and the availability of this information in a typical pavement management system database.
Table 5. Traffic data inputs.
| Description |
Variable |
HMA |
PCC1 |
Typical Data |
| Design properties |
Initial two-way average annual daily truck traffic (AADTT) |
X |
X |
No |
| Number lanes in design direction |
X |
X |
Yes |
| Trucks in the design direction (%) |
X |
X |
No |
| Trucks in the design lane (%) |
X |
X |
No |
| Operational speed |
X |
X |
No |
| Traffic volume adjustment factors |
Monthly adjustment factors |
X |
X |
No |
| Vehicle class distribution (%) |
X |
X |
No |
| Truck hourly distribution factors (%) |
X |
X |
No |
| Traffic growth factors (%) |
X |
X |
No |
| Axle load distribution factors |
Axle load distribution factors by axle type |
X |
X |
No |
| General traffic inputs |
Mean wheel location (inches from lane marking) |
X |
X |
No |
| Traffic wander standard deviation (in) |
X |
X |
No |
| Design lane width (in) |
X |
X |
Yes |
| Number axles per truck class |
X |
X |
No |
| Axle configuration (axle width, dual tire spacing, tire pressure, axle spacing) |
X |
X |
No |
| Wheel base distribution (axle spacing and percent of trucks) |
X |
X |
No |
1 Data required for both JPCP and CRCP
Based on the way that traffic is characterized in the MEPDG, most pavement management databases will not contain the needed traffic information. Therefore, most of this information will need to be provided by other sources within the SHA. Table 6 provides additional details regarding the data collection or measurement requirements for each of the MEPDG input levels. Additional (more detailed) information on the traffic module and axle load spectra is available in the NCHRP report, Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures, Final Report, Part 2, Chapter 4, Traffic (NCHRP 2004).
Table 6. Traffic data estimation.
| Variable |
Level |
How to acquire and/or measure |
| Initial two-way AADTT |
1 |
Site specific WIM, AVC or traffic forecasting models |
| 2 |
Regional WIM, AVC, vehicle counts or traffic forecasting models |
| 3 |
National WIM, AVC, vehicle counts or traffic forecasting models |
| Trucks in the design direction |
1 |
Site specific WIM, AVC or vehicle counts |
| 2 |
Regional WIM, AVC or vehicle counts |
| 3 |
National WIM, AVC or local vehicle counts/experience |
| Trucks in the design lane |
1 |
Site specific WIM, AVC or vehicle counts |
| 2 |
Regional WIM, AVC or vehicle counts |
| 3 |
National WIM, AVC or local vehicle counts/experience |
| Operational speed |
N/A |
Direct measurement of site specific segment or calculate based on Highway Capacity Manual |
| Monthly adjustment |
1 |
Site specific WIM or AVC |
| 2 |
Regional WIM or AVC |
| 3 |
National WIM or AVC |
| Vehicle class distribution |
1 |
Site specific WIM, AVC or vehicle counts |
| 2 |
Regional WIM, AVC or vehicle counts |
| 3 |
National WIM, AVC or local vehicle counts/experience |
| Hourly distribution |
1 |
Site specific WIM, AVC or vehicle counts |
| 2 |
Regional WIM, AVC or vehicle counts |
| 3 |
National WIM, AVC or local vehicle counts/experience |
| Traffic growth rate |
N/A |
Continuous or short duration AADTT counts |
| Axle load distribution factors |
1 |
Site specific WIM or AVC |
| 2 |
Regional WIM or AVC |
| 3 |
National WIM or AVC |
| Mean wheel location |
1 |
Direct measurement of site specific segment |
| 2 |
Regional/statewide average |
| 3 |
National average or local experience |
| Traffic wander standard deviation |
1 |
Direct measurement of site specific segment |
| 2 |
Regional/statewide average |
| 3 |
National average or local experience |
| Design lane width |
N/A |
Direct measurement of site specific segment |
| Number of axles per truck |
1 |
Site specific WIM, AVC or vehicle counts |
| 2 |
Regional WIM, AVC or vehicle counts |
| 3 |
National WIM, AVC or local vehicle counts/experience |
| Axle configuration |
N/A |
Measure directly, obtain information from manufacturers, national average or local experience |
Environmental/Climatic Model
The MEPDG uses detailed climatic information in the analysis of pavement performance by predicting distress quantities over time for each of the different pavement types. This is a significant enhancement over the previous approach, which merely specifies a climatic region. The MEPDG considers the impacts of seasonal, daily, and hourly moisture and temperature distributions on pavement performance. The climatic data used in the MEPDG is shown in table
Table 7. Environment/climatic parameters.
| Description |
Variable |
HMA |
PCC1 |
Typical Data |
| Design properties |
Climatic data file2 |
X |
X |
No |
| Latitude (degrees, minutes) |
X |
X |
No |
| Longitude (degrees, minutes) |
X |
X |
No |
| Elevation (ft) |
X |
X |
No |
| Depth of water table (ft) |
X |
X |
No |
1 Data required for both JPCP and CRCP
2 Climatic data file can be imported (previously generated) or generated
The climatic data files developed for use with the MEPDG are available for downloaded at the NCHRP website (http://www.trb.org/mepdg/climatic_state.htm). The downloaded climatic files can be supplemented with files developed by the SHA based on available weather station data using hourly temperature, wind speed, percent sunshine, precipitation, and relative humidity. Other needed climatic information (as shown in table 7) is typically available from design personnel or other internal sources. In order to model thermal and moisture conditions within the pavement structure, numerous data inputs are required. Detailed information on the necessary inputs is available in the NCHRP report, Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures, Final Report, Part 2, Chapter 3, Environmental Effects (NCHRP 2004).
Pavement Structure Model
The pavement structure module allows for the creation of a basic pavement structure (HMA or PCC) for new or rehabilitation design and analysis. In addition, the pavement structure module requires detailed material properties data. A summary of the necessary basic pavement structure information used with the MEPDG is provided in table 8.
Table 8. Pavement structure summary.
| Description |
Variable |
New |
HMA Overlay |
New |
PCC1 Overlay |
Typical Data |
| Structure properties |
Layer type |
X |
|
X |
X |
|
X |
Maybe |
| Layer material |
X |
|
X |
X |
|
X |
Yes |
| Layer thickness (in) |
X |
|
X |
X |
|
X |
Maybe |
| Rehabilitation level |
|
|
X |
|
|
|
Yes |
| Milled thickness (in) |
|
|
X |
|
|
X |
Maybe |
| Pavement rating |
|
|
X |
|
|
X |
Yes |
| Total rutting (in) |
|
|
X |
|
|
|
Yes |
| Surface short-wave absorptivity |
X |
|
X |
X |
|
X |
No |
1 Data required for both JPCP and CRCP
The information required for the basic pavement structure section is typically available within a pavement management system with a few exceptions. For instance, the surface short-wave absorptivity of the pavement surface is not typically included in a pavement management database. Table 9 identifies how the surface short-wave absorptivity can be estimated for various MEPDG levels. Other information that may be missing from a pavement management database, such as layer type and layer thickness, may be obtained from cores or from design records.
Table 9. Determining surface short-wave absorptivity.
| Variable |
Level |
How to acquire and/or measure |
| Surface short-wave absorptivity |
1 |
Estimate through laboratory testing |
| 2 |
N/A |
| 3 |
Default values |
Material Characterization
Significant material characterization is required to support the MEPDG, especially at Level 1 and Level 2. The required HMA, PCC, chemically stabilized, unbound, and bedrock material input parameters are presented in table 10 through table 14. These input parameters are typically not found in most pavement management systems. However, this information may be obtained from records in a SHA materials laboratory, construction records, and from field cores.
Table 10. HMA layer characterization.
| Description |
Variable |
HMA |
Typical Data |
|---|
| New |
Overlay1 |
| Design properties | HMA E* predictive model | X | X | No |
| HMA rutting model coefficients | X | X | No |
| Fatigue analysis endurance limit | X | X | No |
| Include reflective cracking in analysis | |
X | N/A |
| Mix properties | Aggregate gradation (% retained, % passing) | X | X | No |
| Asphalt binder type | X | X | No |
| Asphalt binder grade | X | X | No |
| General properties | Reference temperature (oF) | X | X | No |
| Effective binder content (%) | X | X | No |
| Air voids (%) | X | X | No |
| Total unit weight (pcf) | X | X | No |
| Poisson's ratio | X | X | No |
| Thermal properties | Thermal conductivity (BTU/hr ft oF) | X | X | No |
| Heat capacity (BTU/lf oF) | X | X | No |
| Average tensile strength at 14oF (psi) | X | X | No |
| Creep compliance (1/psi) | X | X | No |
| Coefficient of thermal contraction (in/in/oF) | X | X | No |
| Rehabilitation (HMA overlay of PCC) | Poisson's ratio of PCC | |
X | No |
| Elastic resilient modulus of fractured slab | |
X | No |
| Type of slab fracture | |
X | No |
| Thermal conductivity of PCC slab | |
X | No |
| Heat capacity of PCC slab | | | |
| Slabs with transverse crack before restoration (%) | |
X | Yes |
| Repaired slabs after restoration (%) | |
X | Yes |
| Dynamic modulus of subgrade reaction (psi/in) | |
X | Yes |
Month measured | |
X | Yes |
1 HMA overlays include: overlays of HMA, and overlays of JPCP and fractured JPCP 1 JPCP/CRCP overlays include: bonded and unbonded overlays and overlays of flexible pavements
Table 11. PCC layer properties.
| Description |
Variable |
JPCP |
CRCP |
Typical Data |
|---|
| New |
Overlay1 |
New |
Overlay1 |
|---|
| Design properties | Permanent curl/warp effective temperature difference (oF) | X | X | X | X | No |
| Joint spacing (ft) | X | X | | |
Yes |
| Sealant type | X | X | | |
No |
| Dowel diameter and joint spacing | X | X | | |
No |
| Edge support - tied PCC (% LTE) | X | X | X | X | No |
| Edge support - widened slab (ft) | X | X | | |
No |
| PCC-base interface | X | X | | |
No |
| Base erodibility index | X | X | X | X | No |
| Steel reinforcement (%) | | |
X | X | No |
| Diameter of steel reinforcement (in) | | |
X | X | No |
| Depth of steel reinforcement (in) | | |
X | X | No |
| Base/slab friction coefficient | | |
X | X | No |
| Crack spacing (in) | | |
X | X | No |
| General
properties | Layer thickness (in) | X | X | X | X | Maybe |
| Unit weight (pcf) | X | X | X | X | No |
| Poisson's ratio | X | X | X | X | No |
| Thermal properties | Coefficient of thermal expansiono -6(per Fx10 ) | X | X | X | X | No |
| Thermal conductivity (BTU/hr ft oF) | X | X | X | X | No |
| Heat capacity (BTU/lf oF) | X | X | X | X | No |
| Mix properties | Cement type | X | X | X | X | No |
| Cementitious material content (lb/yr3) | X | X | X | X | No |
| Water/cement ratio | X | X | X | X | No |
| Aggregate type | X | X | X | X | No |
| PCC zero-stress temperature | X | X | X | X | No |
| Ultimate shrinkage at 40% R.H. (microstrain) | X | X | X | X | No |
| Reversible shrinkage(% of ultimate shrinkage) | X | X | X | X | No |
| Time to develop 50% of ultimate shrinkage | X | X | X | X | No |
| Curing method | X | X | X | X | No |
| Strength properties | 28-day Elastic modulus (psi) | X | X | X | X | No |
| 28-day Modulus of rupture (psi) | X | X | X | X | No |
| Compressive strength (psi) | X | X | X | X | No |
| Splitting tensile strength (psi) | | |
X | X | No |
| Rehabilitation | Slabs with transverse cracks before restoration (%)3 | |
X | X | X | Yes |
| Repaired slabs after restoration (%) | |
X | X | X | Yes |
| CRCP existing punchouts (per mi) | | |
X | X | Yes |
| Dynamic modulus of subgrade reaction(psi/in) | |
X | X | X | No |
| Month measured | |
X | X | X | No |
Table 12. Stabilized layer inputs.
| Description | Variable |
Typical Data |
|---|
| General properties | Material type (cement and lime alternatives) | Yes |
| Layer thickness (in) | Maybe |
| Unit weight (pcf) | No |
| Poisson's ratio | No |
| Strength
properties | Elastic/resilient modulus (psi) | No |
| Minimum elastic/resilient modulus (psi) | No |
| Modulus of rupture (psi) | No |
| Thermal
properties | Thermal conductivity (BTU/hr ft oF) | No |
| Heat capacity (BTU/lf oF) | No |
Table 13. Unbound layer inputs.
| Description |
Variable |
Typical Data |
| General properties | Material type | Yes |
| Layer thickness (in) | Maybe |
| Poisson's ratio | No |
| Coefficient of lateral pressure | No |
| Strength
properties1 | Modulus (psi) | No |
| CBR | No |
| R-value | No |
| Layer coefficient (ai) | No |
| Penetration DCP | No |
| Plasticity index and gradation | No |
| ICM
properties | Gradation (% passing) | No |
| Plasticity index | No |
| Liquid limit | No |
| Compacted layer (Yes/No) | No |
Table 14. Bedrock layer inputs
| Description |
Variable |
Typical Data |
| General properties |
Material type |
Yes |
| Layer thickness (in) Unit weight (pcf) Poisson's ratio |
Maybe No No |
| Resilient modulus (psi) |
No |
As indicated previously, most of the required materials input data is not typically contained in a SHA pavement management database. However, other sources, such as construction records, materials laboratories, or other SHA databases should be explored. A coring program may also be used to obtain missing pavement layer type, thickness, and material characterization information. Missing materials information can be addressed following the MEPDG proposed guidelines included below in tables 15 through 19.
Table 15. Estimating HMA layer parameters.
| Variable |
Level |
How to acquire and/or measure |
| Dynamic Modulus | 1 | AASHTO TP62 |
| 2 | Predictive equation using G*-D Ai-VTSi calculated values |
| 3 | Predictive equation using typical Ai-VTSi values |
| Aggregate gradation | 1 | AASHTO T27 |
| 2 | N/A |
| 3 | N/A |
| Effective binder content | 1 | AASHTO R35 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Air voids | 1 | AASHTO 269 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Total unit weight | 1 | AASHTO T166 and AASHTO T209 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Poisson's ratio | 1 | N/A |
| 2 | Regression equation based on 'a' and 'b' values |
| 3 | Agency historical data or typical values |
| Thermal conductivity | 1 | ASTM E1952 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Heat capacity | 1 | ASTM D2766 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Average tensile strength | 1 | AASHTO T322 |
| 2 | N/A |
| 3 | Regression equation based on NCHRP 1-37a |
| Creep compliance | 1 | AASHTO T322 |
| 2 | AASHTO T322 |
| 3 | Regression equation based on NCHRP 1-37a |
| Coefficient of thermal contraction | 1 | N/A |
| 2 | Correlation based on HMA volumetric properties |
| 3 | N/A |
| Dynamic modulus of subgrade reaction | 1 | AASHTO T307 |
| 2 | Correlation based on CBR, R-value, ai, and DCP |
| 3 | Agency historical data or typical values |
Table 16. Determining PCC layer values.
| Variable |
Level |
How to acquire and/or measure |
|---|
| Unit
weight | 1 | AASHTO T121 or T271 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Poisson's
ratio | 1 | ASTM C469 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Coefficient
of thermal expansion | 1 | AASHTO TP60 |
| 2 | Correlation based on aggregate and paste CTE values |
| 3 | Agency historical data or typical values |
| Thermal
conductivity | 1 | ASTM E1952 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Heat
capacity | 1 | ASTM D2766 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Ultimate
shrinkage | 1 | AASHTO T160 |
| 2 | Correlation based on PCC mix parameters |
| 3 | Level 2 correlation |
| Reversible
shrinkage | 1 | AASHTO T160 |
| 2 | As per Level 1 |
| 3 | As per Level 1 |
| Elastic
modulus | 1 | ASTM C469 |
| 2 | Correlation based on compressive strength |
| 3 | ASTM C469, historical data, or typical values |
| Modulus
of rupture | 1 | AASHTO T97 |
| 2 | Correlation based on compressive strength |
| 3 | AASHTO T97, historical data, or typical values |
| Splitting
tensile strength | 1 | AASHTO T198 |
| 2 | Correlation based on compressive strength |
| 3 | AASHTO T198, historical data, or typical values |
| Compressive
strength | 1 | AASHTO T22 |
| 2 | N/A |
| 3 | AASHTO T22, historical data, or typical values |
Table 17. Characterizing stabilized layer inputs.
| Variable |
Level |
How to acquire and/or measure |
|---|
| Unit
weight | 1 | AASHTO T121 or T271 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Poisson's
Ratio | 1 | N/A |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Elastic/resilient
modulus1(PCC surface) | 1 | ASTM C469 and AASHTO T307 |
| 2 | Correlation based on strength |
| 3 | Agency historical data or typical values |
| Elastic/resilient
modulus1 (HMA surface) | 1 | AASHTO T307 and ASTM D3497 |
| 2 | Correlation based on strength |
| 3 | Agency historical data or typical values |
| Thermal
conductivity | 1 | ASTM E1952 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Heat
capacity | 1 | ASTM D2766 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
Table 18. Characterizing unbound layer inputs.
| Variable |
Level |
How to acquire and/or measure |
|---|
| Poisson's
ratio | 1 | N/A |
| 2 | Correlation based on local knowledge and experience |
| 3 | Agency historical data or typical values |
| Coefficient
of lateral pressure | 1 | N/A |
| 2 | Correlation based on material properties |
| 3 | Agency historical data or typical values |
| Modulus | 1 | AASHTO T307 |
| 2 | Correlation based on CBR, R-value, ai, and DCP |
| 3 | Agency historical data or typical values |
| CBR | 1 | AASHTO T193 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| R-value | 1 | AASHTO T190 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Layer
coefficient | 1 | AASHTO Guide for the Design of Pavement Structures |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| PenetrationDCP | 1 | ASTM D6951 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Gradation | 1 | AASHTO T27 |
| 2 | N/A |
| 3 | N/A |
| Plasticity
index | 1 | AASHTO T90 |
| 2 | N/A |
| 3 | N/A |
| Liquid
limit | 1 | AASHTO T89 |
| 2 | N/A |
| 3 | N/A |
Table 19. Characterizing bedrock layer inputs.
| Variable |
Level |
How to acquire and/or measure |
|---|
| Unit
weight | 1 | AASHTO T121 |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Poisson'sRatio | 1 | N/A |
| 2 | N/A |
| 3 | Agency historical data or typical values |
| Resilient
modulusus | 1 | AASHTO T307 |
| 2 | Correlation based on strength |
| 3 | Agency historical data or typical values |
1 Test method depends on type of stabilized base
It is important to note that additional State-specific material characterization data may also be extracted from the LTPP database to supplement missing information that cannot be obtained through direct testing or an agency specific historical/typical values. State research reports may also be an excellent source for historical/typical data.
Detailed information on materials characterization requirements for MEPDG Level 1(data associated with specified test protocols), Level 2 (correlation equations), and Level 3 (typical default values) are available in the NCHRP report, Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures, Final Report, Part 2, Chapter 2, Material Characterization (NCHRP 2004).
Pavement Distress Prediction and Measurements
The key pavement performance indicators used by the MEPDG are summarized in table 20. These performance indicators (pavement distresses), associated limits, and reliability levels are used to predict the performance of a typical pavement design using the MEPDG models. The MEPDG models were calibrated using national data; however, each agency should consider calibration of the distress models to local State conditions. There are expected to be some difficulties in calibrating the models, since some distress (such as HMA top-down cracking) are not included in most network-level condition surveys conducted as part of an agency’s pavement management activities. Therefore, each agency must determine which models will or will not be calibrated with the pavement management survey information.
The majority of the pavement performance indicators (pavement distresses) specified in table 20 are available in a typical pavement management system. However, performance indicators that are not currently collected can be obtained using automated or visual distress surveys. Using caution will ensure compatible performance indicators (definition and format) between the MEPDG and the pavement management system. The condition definitions used in the MEPDG are based on the LTPP Distress Identification Manual. Differences between SHA pavement condition definitions and those identified in the LTPP Distress Identification Manualwill need to be considered in the MEPDG calibration process. One option to consider is incorporating these differences as part of the calibration process. A second option is to calibrate the MEPDG
to each State’s LTPP sites and then calibrate the MEPDG results to the State’s pavement
condition survey. For those States with limited LTPP sites, it may be advisable to identify appropriate pavement sections, conduct the pavement condition survey according to the LTPP Distress Identification Manual, and calibrate the MEPDG accordingly. In addition, occasionally one or more distress types are combined under a single pavement management system classification; therefore, it will be important to identify the distress classification and measurement units in a State pavement management system before a local calibration effort is attempted. During the calibration process, States are highly recommended to review the Recommended Practice for Local Calibration of the ME Pavement Design Guide(anticipated AASHTO publication in 2010) for details related to selection of calibration sections, estimation of needed sample size, and determination of standard error and bias.
Database Development Framework
The development of a simple database or series of spreadsheets is required to store additional MEPDG related/specific traffic, climate, material, and pavement performance data that currently does not exist within a State pavement management system. MS Access® was selected by the research team to create the MEPDG calibration database and associated tables to support the local calibration of MEPDG models by a SHA. A database system is proposed instead of a spreadsheet-based system due to the distinct benefits of database systems. The comparison between databases and spreadsheets is shown in table 21.
Table 21. Differences between databases and spreadsheets.
| Database |
Spreadsheet |
- Easier to store, organize and retrieve data
- Data links to minimize redundancies, which results in smaller file sizes, faster speeds for data access and reduced errors
- Supports complex searches
- Supports multiple user access
|
- Easy to use and familiar to many users
- Good for numerical computations and developing graphs
- Strength is in calculations and not organizing records
- Can only handle simple data relationships
- Challenging to manipulate large quantity of records
|
The differences between databases and spreadsheets demonstrate that the use of a database is much more versatile and functional for capturing data from existing databases and incorporating additional information needed for MEPDG operation and calibration. The selection of a database system is recommended for implementation by SHAs interested in performing local calibration for MEPDG models. The preliminary framework includes a series of MS Excel® files that might be partially or fully developed by a SHA for some MEPDG related inputs and are linked together by the engine of the MS Access® database program. The MS Access® database program is user-friendly, does not require extensive training, and the associated database tables are simple to develop, as highlighted in the next section.
Figure 2 illustrates that the general structure of the MEPDG calibration database and the tables proposed in the preliminary framework closely follow the structure of the MEPDG software program. This framework allows the MEPDG inputs to be populated in a logical manner for MEPDG design and analysis runs and the subsequent local calibration of MEPDG models. The proposed MEPDG calibration database consists of five main modules:
- The Project Module contains the project summary input information and is also used to link one or more modules together. This table is referred to as the "master table."
- The Traffic Module contains all MEPDG traffic input data.
- The Climate Module contains all necessary MEPDG environmental related inputs data.
- The Material Module contains both structure (thickness and material types) and material characterization inputs data.
- The Performance Module contains all distress measurement data and the distress limits for each distress type or trigger values for rehabilitation design.
All proposed input data is specified and required for MEPDG design, analysis, and subsequent local calibration of the MEPDG models. The proposed MEPDG calibration database structure will allow SHAs, over time, to develop a catalogue of agency typical/specific design input values when site-specific information is not readily available. This will allow for an improved characterization of the local conditions/environment resulting in increased accuracy of the performance models.
Figure 2. General MEPDG calibration database structure.
The following sections further describe each of the proposed database elements, which are presented in more detail in appendix A.
Project
The project element of the database serves as the "master table" and is linked to the other elements by project and site specific information. This table contains the MEPDG project summary information (design properties, project/site identification) and spatial coordinates for each site. Analysis limit and associated reliability parameters are stored within the performance tables.
Traffic
The traffic element contains nine tables for storing the required data for the MEPDG. These tables cover traffic design properties, traffic volume adjustments, axle load distribution based on axle type, and general traffic inputs as outlined in the NCHRP report, Guide for Mechanistic-Empirical Design of New and Rehabilitated Pavement Structures, Final Report, Part 2, Chapter 4, Traffic (NCHRP 2004). Traffic files can be compiled as detailed in Chapter 3 of this report.
Climate
The climate element contains two tables that store site-specific weather station data necessary to create hourly climatic database files and other environmental related information. Similarly, climatic information can be compiled as detailed in Chapter 4 of this report.
Material
The material element contains 17 tables to describe the material properties for each pavement layer. These tables are broken down into five material areas: HMA, PCC, stabilized, unbound, and bedrock. Within each of these five main areas, additional tables are required to describe design properties, mix properties, thermal properties, thermal cracking, and materials strength properties. The information contained in these tables is discussed in Chapter 5 of this report.
Performance
The performance element contains six tables to summarize the observed distresses for each pavement type, analysis trigger limits, and rehabilitation overlay data (including project history data for use with rehabilitation projects).
Other Tables
Those items not readily available in the pavement management system, but available in other State maintained databases or files (e.g. construction history, traffic data, GPS referencing system).
Future Enhancements
The MEPDG is an evolving pavement analysis tool. It is fully anticipated that future modifications will be made to the existing models, as well as the potential for the inclusion of entirely new models and design features, all of which may require additional sources of data (e.g. performance prediction and material characteristics). As these additions come to fruition, comparable modifications to the MEPDG calibration database will need to occur. Though this study will not resolve the issues (such as quantifiable performance data, material characterization, and impact of the existing pavement condition on preventive maintenance treatment performance, and so on) surrounding the incorporation of preventive maintenance treatments into a pavement design/analysis procedure, table 22 lists the potential data needs (assuming pavement performance prediction models have been developed) to analyze these activities.
Table 22. Example of data needs for preventive maintenance treatments.
| Description |
Variable |
Typical Data |
| Design properties |
Project name and description |
No |
| Design life (years) |
No |
| Traffic opening (date) |
No |
| Site/project identification |
Location |
No |
| Project ID |
No |
| Section ID |
No |
| Stationing (format, beginning and end) |
No |
| Traffic direction |
No |
| Analysis parameters (limit and reliability) |
Initial IRI (in/mi) |
Yes |
| Terminal IRI (in/mi) |
Yes |
| AC surface down cracking (ft/mi) |
No |
| AC bottom up cracking (%) |
Yes |
| AC thermal fracture (ft/mi) |
Yes |
| Permanent deformation -total pavement (in) |
No |
| Permanent deformation -AC only (in) |
Yes |
| Transverse cracking (% slabs cracked) |
Yes |
| Mean joint faulting (in) |
Yes |
| Existing punchouts |
Yes |
| Maximum crack width (in) |
No |
| Minimum crack load transfer efficiency (%) |
No |
| Minimum crack spacing (ft) |
No |
| Maximum crack spacing (ft) |
No |
| Structure properties |
Treatment type |
No |
| Layer material |
No |
| Layer thickness (in) |
No |
Summary
Local calibration is an integral part of the implementation of the MEPDG for any SHA. This is necessary because the default MEPDG calibration coefficients are based on national information and may not accurately describe the local traffic conditions, climatic environment, materials, and construction/maintenance practices.
State pavement management system databases will be able to provide basic input parameters required to support the local calibration of the MEPDG. However, there is a need to look outside the pavement management system to as-built construction records, material testing databases or records, and other SHA databases for the necessary traffic, climate, material characterization, and performance/distress measurements. Identifying how the missing data requirements can be obtained will allow SHAs to focus their resources to successfully calibrate the MEPDG to local conditions.
The developed MEPDG calibration database structure will allow for the storage of necessary MEPDG inputs that are not currently in the pavement management system. Having this information in a centralized location, SHAs can effectively extract the necessary data for MEPDG implementation and identify areas that need further characterization and development to better model local traffic, environment, and material conditions.
