KLK572 Calibration of the AASHTOWare Pavement ME Design Performance Models for Flexible Pavements in Idaho
ITD Research Project 235; Task Order Number UI-15-04
Principal Investigators
- Fouad Bayomy
ITD Project Manager
- Mike Santi
FHWA Project Advisor
- Kyle Holman
Subaward
- Haifang Wen, WSU
Research Problem Statement
The Mechanistic-Empirical Pavement Design Guide (MEPDG) is a comprehensive tool for the analysis and design of new and rehabilitated flexible and rigid pavement structures based on mechanistic-empirical principles. In 2009, ITD initiated a major effort toward the implementation of the MEPDG (ITD Project RP193). The main focus of the implementation project was to establish a comprehensive material, traffic, and climatic database for the MEPDG. Later, ITD contracted with ARA to develop an implementation plan and a design guide to help ITD personnel to implement the new design guide. The MEPDG research version software has been modified and is now transferred as an AASHTOWare software known as the “AASHTOWare Pavement ME Design”. For short, we will refer to it in this task order as the “ME software”. It is to be noted that the MEPDG software and for that matter the ME Software predicts pavement performance based on nationally calibrated predictive models. The national calibration is based on field data from LTPP sections distributed all over North America. Thus, for successful implementation of the AASHTOWare Pavement ME Design software in Idaho, these models need to be locally calibrated based on Idaho conditions.
Project Objectives
The main objective of this research project is to develop local calibration (adjustment) factors for the ME Pavement Design performance models for flexible pavement design in Idaho. This is an important part for a successful implementation of the ME Design in Idaho.
Project Tasks
The project team will consist of researchers from the University of Idaho (U of I) and Washington State University (WSU). The U of I research team will be led by the principal investigator (PI), Dr. Fouad Bayomy and the WSU team will be led by Dr. Haifang Wen as a project Co-PI. WSU team will essentially be responsible of upgrading the current material database as described in Task 4 below.
Following is a description of the project tasks:
Task 1: Review of the ME software distress prediction models for flexible pavements — The new AASHTOWare software was released in summer 2013. ITD shall be responsible for acquiring the software and documentation. The distress and roughness (expressed as International Roughness Index, IRI) prediction models in the ME software as well as the global calibration factors shall be reviewed. Trial runs will be performed with the most recent version of the software and compared with the current version to get familiar with the software and investigate the changes.
Task 2: Evaluate the design inputs required for the ME software — In this task, the research team will study and evaluate the inputs required to run the latest version of the ME software. In addition, the level of input for each required parameter will be determined based on previous literature studies as well as ITD available data.
Task 3: Identify and select the pavement sections for calibration —In addition to the LTPP projects available in Idaho, the research team in coordination with ITD will identify and select pavement projects representative of the different districts in Idaho. These projects shall cover a reasonable range of environmental conditions, traffic levels, and subgrade strength. The selected projects shall have all sufficient inputs required by the software as well as sufficient field measured performance data. For the selected projects, ITD shall provide project data required to run the software as well as time series of measured field performance of these projects. These data shall include but not limited to:
- Project location (latitude and longitude).
- Construction year and month.
- As-built pavement structure (layer type and thickness of each layer).
- AC/Base/Subbase/Subgrade material properties required by the software.
- Ground water table level.
- Traffic volume and axle load spectra data in the required MEPDG format.
- Performance data (rutting, longitudinal cracking, alligator fatigue cracking, transverse cracking, and IRI) at different points of time.
- Maintenance history.
Based on the experience of the research team with the data available for in-service pavement sections in Idaho, level 1 M-E input data is difficult to obtain for most of the required inputs. Thus levels 2 and 3 input data will be used if level 1 data is unavailable. Previous outcomes of project RP193 (KLK557) will play a vital role in the characterization of the material, traffic, and climate of the selected projects.
In order to have a reasonable performance record, pavement projects with more than 5 years in service would be recommended. Pavements that are near major rehab period will be better candidates to show various distresses. However, projects need to be restricted to those built with Superpave HMA mixes to be able to utilize the materials database that was developed under RP 193 for Superpave mixes.
The total number of required projects in addition to Idaho LTPP flexible pavement sections required for the local calibration of the distress/IRI models will be determined by the research team in coordination with ITD.
Task 4: Conduct Creep Compliance and IDT Strength Testing at WSU — Thermal cracking is one of dominant distresses in Northern States in the U.S. Based the NCHRP 01-40, the thermal cracking prediction by the Pavement ME is very sensitive (highest category) to the IDT creep compliance and IDT strength. Similar to the case of dynamic modulus of asphalt mix, a material library and a calibration of prediction models of IDT creep compliance and strength need to be completed before a meaningful calibration of Pavement ME can be conducted based on local materials. Tentatively, all six classes of ITD mixes are included in this study. Each class will include three mixes, pending the availability of mixes. It is desired that the three mixes of each class have different performance grades of binders. In total, there will be 18 mixes included in this study. Field cores will be delivered to WSU. The cores shall be taken from new paving projects. Indirect tensile (IDT) strength and IDT creep compliance tests will be conducted, in accordance with AASHTO T322.
Task 5: Develop a performance database for the ME calibration for the identified pavement sections — The performance database required for the local calibration of the performance models shall come from two sources. For LTPP sections, the data will be recruited from the LTPP database. Since the distress/roughness models in the software were nationally (globally) calibrated based on 90+ LTPP sections distributed throughout the U. S., the LTPP performance data is consistent with the software requirements. For the pavement projects in Idaho, performance data will be recruited from ITD Pavement Performance Management Information System (PPMIS). This data will be first evaluated for accuracy, reasonableness and to identify any outliers and anomalies.
Review of ITD cracking (alligator fatigue cracking, longitudinal cracking, and transverse cracking) measurement methods (which are measured by severity and extent for each type of crack) indicated that ITD cracking data are collected and measured differently from the LTPP cracking evaluation method. The research team in cooperation with construction and materials engineer at ITD will study the feasibility of processing and converting ITD cracking to be consistent with the LTPP cracking data. Thus, cracking prediction models in the ME software can be evaluated and calibrated using both LTPP as well as ITD data.
The ME software predicts rutting within each layer in the pavement separately. The total predicted rutting is then the sum of the rutting from each individual layer. On the other hand, ITD as well as LTPP data only measure the total pavement rutting at the surface of the AC layer. Thus, for the calibration of rutting models, the focus will be on the total rut-depth. Percentages of the sub-layer rutting, based on engineering reasonableness as well as the original national calibration of the rutting models, will be assumed.
Task 6: Run the software (using the nationally calibrated models) with the assembled database — Once all projects database are assembled, each project data will be entered into the software. The software will be run and predicted performance (using the nationally calibrated models) will be compared with the field measured performance. Precision and bias in the predictions of the nationally calibrated performance models will be assessed. This will warrant the models that need to be locally calibrated.
Task 7: Develop Idaho calibration factors — In this task, using the outcomes of Task 5, the software will be run using different trial sets of calibration coefficients for each distress model separately to find the best combination of calibration factors. The set of calibration factors that will produce a higher precision and lower bias for each distress model compared to the nationally calibrated models will be selected. The IRI model will then be recalibrated after the distress models as it is dependent on the predicted rutting and cracking as well as other factors. Finally, the resulting goodness of fit and bias of the calibrated models will be statistically validated.
Task 8: Summary of findings and recommendations — Before, completing the final report and developing the training workshop, a summary of findings and draft calibration factors will be developed and submitted to ITD for review. ITD review comments and recommendations will be incorporated in the development of final report and the training workshop.
Task 9: Prepare and conduct training workshop — The Idaho design manual that was prepared by ARA will be reviewed and updated to reflect the new findings from this project. A workshop to demonstrate and train ITD personnel on the new software will be performed at a suitable location where access to the software will be available.
Task 10: Final report — This task is divided in 7 subtasks as detailed in the Gantt chart. A draft report will be edited by a professional editor and reviewed by an external expert. The edited and reviewed draft report will be submitted to ITD for comments before submission of the final report.
Project Communication Schedule
In addition to the monthly progress report on ITD 771 form, regular contact with ITD project oversight committee will occur as needed. Project meetings, as requested by the research team or the technical advisory committee, will be held to update ITD project team on the status. The meetings may also be held via telephone or video conference depending on the scheduling. A face-to-face meeting will be scheduled following submission of the final report outline to discuss the report structure and key research findings, conclusions, and proposed recommendations prior to writing the final report
Needs and Requirements
Full cooperation with ITD is essential to the success of this project. The above tasks identify the information that shall be furnished and / or coordinated by ITD. In summary:
- ITD will provide the research team with access to the latest ME software version and manual.
- ITD will provide detailed input data required by the software as well as measured performance for the selected projects. Task 3 lists some of the important data needed for the selected projects. For Task 4, ITD will provide field cores extracted from new paving projects for various SP classes (SP1-6). This task will be coordinated to cover all possible mixes that match those tested in the previous project (RP193).
Required Outputs
Required project output will be as follows:
- Final report describing purpose, methodology, findings, and recommendations. The report shall include a list of the developed calibration factors.
- Training workshop on the use of the latest version of the ME software and the local calibration factors.
Implementation
The project deliverables shall aid ITD engineers to implement the new AASHTO design procedure in the state. ITD Engineers will be able to use the results directly to run the software and conduct the various design tasks for flexible pavement design in Idaho. It is important, however, to note that the project scope is limited to the flexible pavements.
Expected Outcomes/Savings to ITD
Outcomes:
- Locally calibrated distress/IRI models specific to Idaho conditions.
- Full implementation of the AASHTOWare Pavement ME Design method in Idaho.
- Training on the use of the latest version of the software with the local calibration factors.
Savings:
- This project allows the pavement design engineers to realistically assess pavement distress and design the pavement accordingly. It also ties material and pavement design together.
- It is expected that this will lead to optimal design of the pavement structure.
- Implementation of ME Design method shall result in huge savings in pavement construction cost. The current method has high factor of safety to account for the uncertainty impeded in the design process and for the lack of knowing the level of performance at any point of the pavement service life.
Monthly Progress Reports (ITD 771) and Task Progress Charts
Draft report due:
7/31/17
Project status:
Active
Final report due date:
12/31/17