Local Calibration of the MEPDG Using Pavement Management Systems
Chapter 6. Final Framework
Introduction
The final framework itemizes specific activities (e.g., links to other data sources, establishing roles and responsibilities) that are needed prior to populating the MEPDG calibration database. The following summarizes the data and process for finalizing the framework for integrating state pavement management data for calibration of the MEPDG. The majority of the MEPDG input data were extracted from the NCDOT pavement management system. Input data associated with materials and traffic was acquired from other NCDOT files and/or databases.
NCDOT is interested in calibrating the MEPDG to Level 2 only. There are several reasons for this. Typically, NCDOT pavement designs occur one to two years prior to letting and the ability to obtain material specific data is not a reality until the project has been awarded. In addition, NCDOT is unable to justify the expense for collecting data according to Level 1 standards; data collection to Level 2 inputs is more justifiable and realistic at this time.
The NCDOT highway network is comprised of primarily HMA pavements, with a lower percentage of PCC pavements. The NCDOT PMS contains sufficient pavement sections, construction history, and performance data for the HMA pavements that no major issues are anticipated for calibrating the MEPDG. For PCC pavements (of which NCDOT only constructs JPCP), the quality and existence of construction and performance data is similar to that of HMA pavements; however, the number of PCC pavement sections and the length of in-service life is considerably less. At a minimum, the framework will provide NCDOT with the step-by-step calibration process as additional pavement history is obtained on PCC pavements.
As a first step, data contained in the NCDOT pavement management system was identified (table 23 through table 31) for applicability with the preliminary framework. In addition, where appropriate, MEPDG default values were established (e.g. analysis parameters, surface shortwave absorptivity, coefficient of thermal expansion).
Table 23. Project summary information.
| Description |
Variable |
HMA |
PCC |
| Design properties |
Project name and description |
PMS1 |
PMS1 |
| Design life (years)2 |
20 |
30 |
| Base/subgrade construction (date)3 |
Assumed |
Assumed |
| Restoration/overlay |
|
|
| Existing pavement construction (date) |
PMS1 |
PMS1 |
| Pavement restoration/overlay (date) |
PMS1 |
PMS1 |
| Traffic opening (date) |
PMS1 |
PMS1 |
| Site/project identification |
Location |
PMS1 |
PMS1 |
| Project ID |
PMS1 |
PMS1 |
| Section ID |
PMS1 |
PMS1 |
| Stationing (format, beginning and end)4 |
PMS1 |
PMS1 |
| Traffic direction |
PMS1 |
PMS1 |
| Analysis parameters6 |
Initial IRI (in/mi) |
60 |
75 |
| Terminal IRI (in/mi) |
170 |
170 |
| AC surface down cracking (ft/mi)5 |
n/a |
n/a |
| AC bottom up cracking (%)5 |
10 |
n/a |
| AC thermal fracture (ft/mi)6 |
n/a |
n/a |
| Chemically stabilized layer fatigue fracture (%)7 |
n/a |
n/a |
| Permanent deformation – total (in)8 |
n/a |
n/a |
| Permanent deformation – AC only (in)8 |
¾ |
n/a |
| Transverse cracking (% slabs cracked)9 |
n/a |
10 |
| Mean joint faulting (in) |
n/a |
¾ |
1Data contained within the NCDOT PMS.
2 Based on current NCDOT pavement design practice.
3 Data is typically not collected and will be assumed to be equivalent to the opening to traffic date.
4 NCDOT PMS uses a referencing system based on milepost. Latitudes and longitudes will also be determined for each project for locating weather and soils data.
5 NCDOT does not distinguish between surface down (longitudinal) and bottom up (alligator) cracking. Distress is collected by NCDOT as alligator cracking.
6 Distress not present on NCDOT highways.
7 Distress not collected by NCDOT.
8 Based on NCDOT pavement investigations’, rutting is almost exclusively confined to the top or second lift of the HMA surface. Therefore, total pavement rutting will default to the rut depth of the HMA surfaces.
9 NCDOT does not distinguish between various forms of PCCP cracking.
Table 24. Traffic data.
| Description |
Variable |
Data Location |
| Design properties |
Initial two-way AADTT |
PMS1/ , Project Plans |
| Number lanes in design direction |
PMS1 |
| Trucks in the design direction (%) |
Project Plans |
| Trucks in the design lane (%)2 |
Project Plans, Variable |
| Operational speed |
DOT Universe/, Project Plans, PMS1 |
| Traffic volume adjustment factors |
Monthly adjustment factors |
Traffic3 |
| Vehicle class distribution (%) |
Traffic3 |
| Truck hourly distribution factors (%) |
Traffic3 |
| Traffic growth factors (%) |
Traffic3 |
| Axle load distribution factors |
Axle load distribution factors by axle type |
Traffic3 |
| General traffic inputs |
Mean wheel location (inches from lane marking) |
Default4 |
| Traffic wander standard deviation (in) |
Default4 |
| Design lane width (in) |
Default4 |
| Number axles per truck class |
Default4 |
| Axle configuration |
Default4 |
| Axle width (ft) |
Default4 |
| Dual tire spacing (in) |
Default4 |
| Tire pressure (psi) |
Default4 |
| Axle spacing (ft) |
Default4 |
| Wheel base distribution |
Default4 |
1Data contained within the NCDOT PMS.
2Based on current AASHTO lane distribution factors.
3Data contained within the NCDOT Traffic database.
4MEPDG default values (level 3, where applicable).
Table 25. Existing pavement structure.
| Description |
Variable |
Data Location HMA PCC New Overlay New Overlay |
| Structure properties |
Layer type |
PMS1 |
PMS1 |
PMS1 |
PMS1 |
| Layer material |
PMS1 |
PMS1 |
PMS1 |
PMS1 |
| Layer thickness (in) |
PMS1 |
PMS1 |
PMS1 |
PMS1 |
| Milled thickness (in) |
n/a |
PMS1 |
n/a |
PMS1 |
| Pavement rating2 |
n/a |
PMS1 |
n/a |
PMS1 |
| Total rutting (in) |
n/a |
PMS1 |
n/a |
PMS1 |
| Surface short-wave absorptivity3 |
Default |
Default |
Default |
Default |
1Data contained within the NCDOT PMS.
2NCDOT PMS rating based on a scale of 0 – 100; the MEPDG condition categories, ranging from very poor to excellent, will be defined in increments of 20 points.
3MEPDG default values (Aged PCC: 0.70-0.90; weathered asphalt: 0.80-0.90; new asphalt: 0.90-0.98).
Table 26. HMA layer characterization.
| Description |
Variable |
Data Location |
| New |
Overlay1 |
| Design properties |
HMA E* predictive model |
MATS2 |
MATS2 |
| HMA rutting model coefficients |
MATS2 |
MATS2 |
| Fatigue analysis endurance limit |
MATS2 |
MATS2 |
| Include reflective cracking in analysis |
n/a |
MATS2 |
| Mix properties |
Aggregate gradation (% retained, % passing) |
MATS2 |
MATS2 |
| Asphalt binder type |
MATS2 |
MATS2 |
| Asphalt binder grade |
MATS2 |
MATS2 |
| General properties |
Reference temperature (oF) |
MATS2 |
MATS2 |
| Effective binder content (%) |
MATS2 |
MATS2 |
| Air voids (%) |
MATS2 |
MATS2 |
| Total unit weight (pcf) |
MATS2 |
MATS2 |
| Poisson's ratio |
MATS2 |
MATS2 |
| Thermal properties |
Thermal conductivity (BTU/hr ft oF) |
MATS2 |
MATS2 |
| Heat capacity (BTU/lf oF) |
MATS2 |
MATS2 |
| Thermal cracking |
Average tensile strength at 14oF (psi) |
MATS2 |
MATS2 |
| Creep compliance (1/psi) |
MATS2 |
MATS2 |
| Coefficient of thermal contraction (in/in/oF) |
MATS2,
Project files |
MATS2,
Project files |
| Rehabilitation (HMA overlay of PCC)3 |
Poisson’s ratio of PCC |
n/a |
n/a |
| Elastic resilient modulus of fractured slab |
n/a |
n/a |
| Type of slab fracture |
n/a |
n/a |
| Thermal conductivity of PCC slab |
n/a |
n/a |
| Heat capacity of PCC slab |
n/a |
n/a |
| Slabs with transverse crack before restoration (%) |
n/a |
n/a |
| Repaired slabs after restoration (%) |
n/a |
n/a |
| Dynamic modulus of subgrade reaction (psi/in) |
n/a |
n/a |
| Month measured |
n/a |
n/a |
1HMA overlays include: overlays of HMA, JPCP, and fractured JPCP.
2Data contained within the materials database developed by North Carolina State University.
3Due to incomplete data contained within the NCDOT PMS, this rehabilitation treatment will not be included in the calibration process.
Table 27. PCC layer properties
| Description |
Variable |
JPCP |
| New |
Overlay |
| Design properties |
Permanent curl/warp effective temperature difference (oF) |
Default1 |
Default1 |
| Joint spacing (ft) |
15 |
15 |
| Sealant type |
Silicone |
Silicone |
| Dowel diameter (in) |
CD2 |
CD2 |
| Dowel bar spacing (in) |
12 |
12 |
| Edge support - tied PCC (% LTE)3 |
n/a |
n/a |
| Edge support - widened slab (ft)4 |
n/a |
n/a |
| PCC-base interface |
Full |
Full |
| Base erodibility index |
Resistant |
Resistant |
| Loss of full friction (age in months) |
360 |
360 |
| General properties |
Layer thickness (in) |
PMS5 |
PMS5 |
| Unit weight (pcf) |
150 |
150 |
| Poisson's ratio |
0.20 |
0.20 |
| Thermal properties |
Coefficient of thermal expansion (per oFx10-6) |
Project
Files |
Project
Files |
| Thermal conductivity (BTU/hr ft oF) |
MATS6 |
MATS6 |
| Heat capacity (BTU/lf oF) |
MATS6 |
MATS6 |
| Mix properties |
Cement type |
Type II |
Type II |
| Cementitious material content (lb/yd3) |
526 |
526 |
| Water/cement ratio |
0.559 |
0.559 |
| Aggregate type |
? |
? |
| PCC zero-stress temperature |
1
Default |
Default1 |
| Ultimate shrinkage at 40% R.H. (microstrain) |
Default1 |
Default1 |
| Reversible shrinkage (% of ultimate shrinkage) |
Default1 |
Default1 |
| Time to develop 50% of ultimate shrinkage |
Default1 |
Default1 |
| Curing method |
Compound |
Compound |
| Strength properties7 |
28-day Elastic modulus (psi) |
n/a |
n/a |
| 28-day Modulus of rupture (psi) |
n/a |
n/a |
| Compressive strength (psi) |
4500 |
4500 |
| Splitting tensile strength (psi) |
n/a |
n/a |
| Rehabilitation8 |
Slabs with transverse cracks before restoration (%)3 |
n/a |
n/a |
| Repaired slabs after restoration (%) |
n/a |
n/a |
| Dynamic modulus of subgrade reaction (psi/in) |
n/a |
n/a |
| Month measured |
n/a |
n/a |
1MEPDG default values.
2 Dowel bar diameter varies with pavement thickness, use construction drawings for selecting dowel bar diameter.
3NCDOT does not use tied shoulders.
4NCDOT does not use a widened slab.
5Data contained within NCDOT PMS.
6 Data contained within the materials database developed by North Carolina State University.
7 Only one strength property is required.
8 Due to the relatively young age of NCDOT PCC pavements, rehabilitation has not been conducted and will not be included in the calibration process.
Table 28. Stabilized layer inputs.
| Description |
Variable |
Data Location |
| General properties |
Material type (cement and lime alternatives) |
PMS1, Project Files |
| Layer thickness (in) |
PMS1 |
| Unit weight (pcf) |
MATS2 |
| Poisson's ratio |
MATS2 |
| Strength properties |
Elastic/resilient modulus (psi) |
MATS2 |
| Minimum elastic/resilient modulus (psi) |
MATS2 |
| Modulus of rupture (psi) |
MATS2 |
| Thermal properties |
Thermal conductivity (BTU/hr ft oF) |
MATS2 |
| Heat capacity (BTU/lf oF) |
MATS2 ,
Project Files |
1Data contained within NCDOT PMS.
2 Data contained within the materials database developed by North Carolina State University.
Table 29. Unbound layer inputs.
| Description |
Variable |
Data Location |
| General properties | Material type |
MATS1 |
| Layer thickness (in) |
MATS1 |
| Poisson's ratio |
MATS1 |
| Coefficient of lateral pressure |
MATS1 |
| Strength properties2 | Modulus (psi) |
MATS1 |
| CBR |
MATS1 |
| R-value |
MATS1 |
| Layer coefficient (ai) |
MATS1 |
| Penetration DCP |
MATS1 |
| Plasticity index and gradation |
MATS1 |
| ICM properties | Gradation (% passing) |
MATS1 |
| Plasticity index |
MATS1 |
| Liquid limit |
MATS1 |
| Compacted layer (Yes/No) |
Yes |
1 Data contained within the materials database developed by North Carolina State University.
2 Only one strength property is required.
Table 30. Bedrock layer inputs.
| Description |
Variable |
Data Location |
| General properties | Material type |
MATS1 |
| Layer thickness (in) |
MATS1 |
| Unit weight (pcf) |
MATS1 |
| Poisson's ratio |
MATS1 |
| Resilient modulus (psi) |
MATS1 |
1 Data contained within the materials database developed by North Carolina State University.
The key pavement performance indicators, and source of NCDOT data, used by the MEPDG for HMA and PCC are summarized in table 31 and 32, respectively.
Table 31. HMA pavement performance indicators.
| Description |
MEPDG Variable |
Data Location |
Comment |
| Performance Indicator | IRI (in/mi) |
PMS |
Collected and summarized in 0.1 mile increments. |
| Surface down cracking (ft/mi) |
PMS |
Collected as alligator cracking in accordance with severity level (light, moderate, and severe) and as a percentage of roadway area – conversion to feet of cracking per mile will be necessary. |
| Bottom up cracking (%) |
PMS |
Distress not specifically collected by NCDOT. |
| Thermal fracture (ft/mi) |
-- |
Distress generally not present on NCDOT highways. |
| Chemically stabilized layer fatigue fracture (%) |
PMS |
Distress not collected by NCDOT. |
| Permanent deformation -total pavement (in) |
-- |
Based on NCDOT pavement investigations’, rutting is almost exclusively confined to the top or second lift of the HMA surface. Therefore, total pavement rutting will default to the rut depth of the HMA surfaces. |
| Permanent deformation -AC layers only (in) |
PMS |
Collected with three point laser and summarized on 0.1 mile increments. |
Table 32. PCC pavement performance indicators.
| Description |
MEPDG Variable |
Data Location |
Comment |
| Performance Indicator | Terminal IRI (in/mi) |
PMS |
|
| Transverse cracking (% slabs cracked) |
PMS |
Cracking (all types) is collected according to severity level (light, moderate, and severe). Calculation as a percent of slabs will be necessary. |
| Mean joint faulting (in) |
PMS |
|
Integration of Input Data into the MEPDG Calibration Database
The following work was conducted by NCDOT Pavement Management staff and the APTech project team to integrate the NCDOT PMS and other database(s) information into the MEPDG calibration database:
- Customize the relational MEPDG calibration database to meet the data definition (e.g. integer, decimal) format of NCDOT data fields;
- Provide step-by-step details for integrating MEPDG input data into the MEPDG calibration database which is not currently contained within the PMS (e.g. traffic, materials)
- Document the process for collecting data to populate the MEPDG calibration database for at least one pavement section. The allowed NCDOT and the research team the ability to evaluate and ensure that proper definitions were being used for all input values, to determine if there was any missing or needed data, and to provide insight for integration of the larger data set.
- Populate the MEPDG calibration database with all intended NCDOT pavement sections.
Summary
This section documented the implementation of the final framework using data contained within a single state highway agency. This demonstration showed that the majority of the data can be obtained from a pavement management system; however, some of the needed input data, such as material properties and construction data will require an interface with other data sources. The importance of having a common referencing system also became evident during this activity. A common referencing system becomes important when data is being retrieved from specific project locations amongst the various data sources. Additionally, the availability of a database to store the information that would be used for calibration proved beneficial.
