National Academies Press: OpenBook

Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006 (2006)

Chapter: Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006

Page 1
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 1
Page 2
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 2
Page 3
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 3
Page 4
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 4
Page 5
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 5
Page 6
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 6
Page 7
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 7
Page 8
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 8
Page 9
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 9
Page 10
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 10
Page 11
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 11
Page 12
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 12
Page 13
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 13
Page 14
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 14
Page 15
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 15
Page 16
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 16
Page 17
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 17
Page 18
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 18
Page 19
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 19
Page 20
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 20
Page 21
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 21
Page 22
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 22
Page 23
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 23
Page 24
Suggested Citation:"Changes to the Mechanistic-Empirical Pavement Design Guide Software Through Version 0.900, July 2006." National Academies of Sciences, Engineering, and Medicine. 2006. Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006. Washington, DC: The National Academies Press. doi: 10.17226/23218.
×
Page 24

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Research Results Digest 308 September 2006 INTRODUCTION This digest summarizes numerous changes made under NCHRP Project 1-40D to the original Version 0.7 of the Mechanistic-Empirical Pavement Design Guide (MEPDG) software delivered under NCHRP Project 1-37A in July 2004. Specif- ically, the digest describes the corrections incorporated in Version 0.8 (released in November 2005) and the mainly technical improvements and enhancements included in Version 0.900 (released in July 2006). Changes in Version 0.900 include soft- ware changes in general (including changes to traffic and other general topics), as well as changes in the integrated climatic model, in flexible pavement design and analysis, and in rigid pavement design and analysis. These changes reflect the recommenda- tions of the NCHRP 1-40A independent review team, the NCHRP 1-40 panel, the general design community, various other re- searchers, and the Project 1-40D team itself. Most of these changes have been tracked and summarized in the “Bug Tracker” sys- tem at www.ara-tracker.com (the item num- ber for each change is provided in parenthe- ses). Some changes were more technical, including definition changes, and were not included in the Bug Tracker system. Users of Version 0.900 should recog- nize the following: • The climatic database used in the lat- est calibration and available for de- sign is now substantially larger and enhanced. Old databases downloaded with Versions 0.7 and 0.8 should be eliminated. • Design solutions and associated files generated with Versions 0.7 and 0.8 require some additional inputs and may experience some problems if simply reloaded and run. It is rec- ommended that the user verify each input (by clicking “ok”) before re- running an old file. It is especially important to review the inputs for the unbound materials (i.e., base, CHANGES TO THE MECHANISTIC-EMPIRICAL PAVEMENT DESIGN GUIDE SOFTWARE THROUGH VERSION 0.900, JULY 2006 This digest summarizes key findings from NCHRP Project 1-40D, “Techni- cal Assistance to NCHRP and NCHRP Project 1-40A: Versions 0.900 and 1.0 of the M-E Pavement Design Software,” conducted by Applied Re- search Associates, Inc., and Arizona State University. The digest was pre- pared by Michael I. Darter, Jag Mallela, Leslie Titus-Glover, Chetana Rao, Gregg Larson, Alex Gotlif, and Harold Von Quintus, Applied Research Associates, Inc.; Lev Khazanovich, University of Minnesota; and Matthew Witczak, Mohamed El-Basyouny, Sherif El-Badawy, Aleksander Zborowski, and Claudia Zapata, Arizona State University. Subject Area: IIB Pavement Design, Management, and Performance Responsible Senior Program Officer: Edward T. Harrigan NATIONAL COOPERATIVE HIGHWAY RESEARCH PROGRAM C O N T E N T S Introduction, 1 Software in General (Including Traffic), 2 Integrated Climatic Model, 3 Flexible Pavements, 3 Rigid Pavements, 15

subbase, and subgrade) because several un- bound materials require new inputs. If a run still experiences problems, it is recommended that the user re-enter the project inputs as a “new” file and then re-run. The research team recommends that problems of interest be run again with the new version for comparison with earlier solutions. Version 0.900 of the Mechanistic-Empirical Pavement Design Guide software may be downloaded for evaluation at www.trb.org/mepdg/. A final Ver- sion 1.0, incorporating additional improvements and enhancements, is scheduled for release by the end of 2006. The AASHTO Joint Technical Committee on Pavements plan to use Versions 0.900 and 1.0 to ad- vance the Mechanistic-Empirical Pavement Design Guide through the AASHTO balloting process. SOFTWARE IN GENERAL (INCLUDING TRAFFIC) The following general changes were made to the software, including the traffic module: 1. (#430) Add CRCP and JPCP plots to ACC/PCC outputs. Plots for LTE, Punch- out, Crack Width, Cracking, and Cracking Damage were provided. 2. (#427) Import/Export capability for un- bound material grain size distribution, PI, LL. Import/export capability for un- bound materials so that a default grain size and LL/PI values for soils of interest can be created and used when needed. 3. (#426) Calculate ESALs for flexible pave- ment. The panel wanted ESALs calculated and available to the user, but given that the procedure is NOT based on ESALs, they would only appear in intermediate files and not on the official output for a project. Calcu- lation of ESALs for JPCP has been included in an intermediate file. A similar calculation for flexible pavements was completed. Cal- culated ESALs are found in the intermediate files under the project folder. 4. (#414) Include proper default values for grain size distribution and Atterberg lim- its for unbound materials. Unbound ma- terials data from several hundred LTPP sections were obtained for all available AASHTO soil classifications. Mean and standard deviations were computed for the range of grain size distribution and Atterberg limits and other properties and incorporated into the defaults of the Design Guide. 5. (#259) When reducing the number of years, model traffic still calculated for a longer period, which affected run time negatively, perhaps by 20 to 30 seconds. This problem was resolved. 6. (#252) Change default design period start dates for all analyses. September 2006 was used as the start date for all new analyses. 7. (#244) Create traffic export/import capa- bilities. The user is now allowed to import/ export all of the data needed for the traffic files within the interface. 8. (#232) Multiple file select for the batch mode. Made it possible to use multiple file select to add files to batch mode. 9. (#231) Increase the width of batch mode file entry screen. The batch mode screen was widened. 10. (#125) Double clicking on cone file (*.dgp) crashes. Opening an MEPDG project file by clicking on the file did not work properly in earlier versions. This problem was fixed. 11. (#116) Incorporate NCHRP 1-39 Traf- Load files into the MEPDG. The code was modified to allow the direct use of the out- puts of 1-39 software in the MEPDG. 12. (#79) Batch Mode Option of Software. Runs in batch mode are now allowed. 13. (#78) Inadequate amount of climatic data for given file. An improved warning mes- sage was provided for climatic files with less than 12 months of climatic data. Also, several procedures to estimate missing data were included in the software. 14. (#77) Base type default values. Work was completed to change the default values when the base type is changed. 15. (#76) Problems with “Save-As” option. On a few occasions, two input files containing the same data produced different outputs. This occurred when the file was changed and then saved using the Save-As command in the file menu of the software, and then re-run. The changed inputs might actually not be saved. This problem was corrected. 16. (#70) Summation of Axle Load Distribu- tion Factors. In earlier versions, the Axle Load Distribution Factors in the traffic analy- 2

sis did not appear to count in the last load level in the summation column. This prob- lem was fixed. 17. (#66) Remove single tire configuration for traffic input. In earlier versions, the General Traffic Input screen showed both dual and single tire pressure options. The single tire option was removed, given that the software only uses dual tire axles. 18. (#65) Special axle configuration problems. Several problems were uncovered with the Special Axle Configuration portion of Ver- sion 0.8: • Considerable information for the Special Vehicle Help option was added. • The asphalt.Td file was corrected to en- sure that the correct wheel load coordi- nate locations are input into the file. In earlier versions, the program allowed the user to input wheel load coordinate loca- tions. Unfortunately, the program read all coordinates at an x=0; y=0 location for each tire of the special vehicle. • When the Special Vehicle Option is se- lected the General Traffic Output sum- mary information was replaced with the key input traffic wheel load properties of the special vehicle. • A repetitive error message was elimi- nated by correcting the code. • A problem with the frequency input asso- ciated with the use of the Special Vehicle in a Rehabilitation scenario was corrected. 19. (#50) Changing pavement types after completed run. Changing the type of pave- ment analysis once the run was completed caused the program to hang (e.g., changing CRCP to HMA Overlay of CRCP). The prob- lem was fixed. INTEGRATED CLIMATIC MODEL Extensive changes were made to the integrated climatic model and the state climatic files to improve the predictive capabilities of the climatic model. These were as follows: 1. Addition of more climatic data for each weather station. A new set of weather station files with up to 9 years of hourly data for 851 stations has been provided. It is recom- mended that old weather stations be deleted and not used with the new version. The new weather stations should be downloaded and used with Version 0.900. 2. (#360) Review and correct Hourly Climatic Database. The Version 0.8 Hourly Climatic Database had some errors in the precipitation. This was reviewed and repaired in December 2005. Unfortunately, the fixed files were not widely disseminated. 3. (#276) Failure in climatic model building. When trying to build a climatic model from weather stations around Delaware, Iowa, every combination of the six weather stations near the site (42°28′ N 91°21′W, elev = 1053′) reported missing data for month 200101. This problem was fixed. 4. (#253) Update ICM to allow use of the NCHRP 9-23 Models. A grain-size distribu- tion plot was created. A grain-size distribution to index property correlation and a base mois- ture model to the Thornthwaite Moisture Index (TMI) method were updated. 5. Defaults provided for all unbound materi- als based on measured properties from several hundred LTPP sections across the United States. 6. (#229) Failed ICM stability check. After pressing Run Analysis, the Traffic module completes, then the Climatic module tries to load, but stops with the error “Failed ICM stability check.” The problem was fixed. 7. (#122) ICM crashes with file and directory names of over 80 characters. The problem was fixed. 8. (#74) The Enhanced Integrated Climatic Model (EICM) has been modified and en- hanced, based on the results of the NCHRP 9-23 project. See the Flexible Pavements section. FLEXIBLE PAVEMENTS The following changes were made to the soft- ware to enhance the design of flexible pavements: 1. (#413) Granular layer placed over stabi- lized base crashes for flexible pavement. The modulus for the granular layer was not being output in _space.dat. The problem was fixed. This makes it possible to include, for example, a lime-stabilized layer below an un- bound granular base/subbase layer. 3

2. (#74) The EICM has been modified and en- hanced, based on the results of the NCHRP 9-23 project. Several models have been re- vised and incorporated into the MEPDG code. These models include the following: • New Suction Models; • New ASU—TMI Models; • New SWCC Models; • Moisture Content Models with p200-w%; • Revise Compaction Models; • Revised Specific Gravity Models; and • New Saturated Hydraulic Conductivity (ksat) Model. 3. Version 0.900 will run in batch mode. 4. (#124) The analysis period (design life) is increased from 25 to 50 years. 5. The run time of the program is decreased. 6. The calibration models in the settings screens are updated. 7. The national calibration factors of the dis- tress models are updated based on the en- hancements done to the models. In addi- tion, these calibration factors are based on the most up-to-date database. The following changes were made to improve or correct problems in the software for the design of new and rehabilitated flexible pavements: 1. (#56) Reflective Cracking was not reported correctly to the output summary sheet. The problem was corrected and reflective crack- ing is now reported correctly under the col- umn “Reflective Cracking from the Existing Layers.” 2. The methodology used to report area cracking in the HMA rehabilitation design is based on Reflective Cracking reflected from the existing pavement through the overlay to the surface of the pavement in addition to the amount of fatigue crack- ing occurring in the new layer. The re- sults should be interpreted by looking at the time the cracking reached the surface, not by the amount of cracking at the end of the design life. 3. (#57) In the distress summary output sheet, for the Rehab Analysis, the “Alli- gator Cracking” output column name is changed to “Alligator Cracking in New Overlay.” 4. (#58) A new output column named “Total Cracking at Surface” is added to the out- put. This column represents the sum of the “Alligator Cracking in New Overlay” and the “Reflective Cracking from the Existing Layers.” 5. (#99) In some cases of AC over AC Rehab Analysis, the software does not accept more than one AC layer. The problem was corrected and the program now accepts up to three new AC layers over the existing flexible pavement. 6. The software wrongly allows the user to input any material over the existing lay- ers. The problem was corrected and the software is now allowing the user to input either AC materials or base courses (un- bound or CTB). 7. (#54) In Level 1 Rehab Analysis, if the user did not input the FWD modulus, the software crashed. The problem was cor- rected by incorporating a warning to the user to input the FWD modulus so that the software would not crash. 8. (#64) In Level 1 and Level 2 Rehab Analy- ses, if the user did not input the actual rut- ting and cracking for the existing pave- ment, the software crashed or yielded unreasonable results. The problem was corrected by incorporating a warning to the user to input the actual rutting and cracking for the existing pavement so that the soft- ware would not crash or yield unreasonable results. 9. (#55) The damage function in Level 2 Rehab yields results incompatible with Level 3. The problem was corrected. 10. The input summary worksheet in the out- put file for the Rehab Analyses Levels 1, 2, and 3 is missing a lot of information. The problem was corrected. 11. Use of any new or overlay AC sublayers with thicknesses greater than 9 in. may lead to sublayering yielding AC thick- nesses greater than the original input thickness of the layer. The problem was corrected. 12. (#158) The program crashed if the user has a limited access account. The problem was corrected. 4

The following changes and corrections were made in the treatment of E* in the design of flexible pavements: 1. (#67) An error message usually appears when using Level 1 E* characterization in HMA analysis. The problem was corrected. 2. Several Level 1 E* input data for AC mixtures cause program crashes. The problem was corrected. The seed values for the optimization process used for develop- ing the E* master curve coefficients (Time- Temperature Superposition Optimization) were reviewed and updated based on the his- torical database in order to prevent the program from crashing. 3. (#72) In Level 1 input analysis for HMA, inputting E* values at a small range of tem- peratures may result in an incorrect mas- ter curve, which may cause the program to crash or yield very incorrect results. The problem was corrected. The program re- stricted the E* input values to be at least at three temperatures in a wide range (Temper- ature > 125°F, Temperature between 60 and 90°F, and Temperature < 45°F). 4. In Level 1 input analysis for HMA, limits were imposed on the temperature for the G* testing to provide a reasonable range of temperatures. The G* temperature cannot be less than 40°F or more than 300°F. The following errors that led to discontinuities in the calculation of rutting in the flexible pavement design were corrected: 1. (#68) A discontinuity occurs in the AC rut depth with very small changes in AC layer thickness due to the sublayering scheme used in the software. This discontinuity error was sometimes as high as 20 to 25 percent. The problem was corrected and the error was reduced to less than 10 percent. 2. (#68) A discontinuity occurs in the sub- grade rut depth with small changes in the AC layer thickness. The problem was cor- rected and the error was totally eliminated. The following improvements and corrections in the treatment of unbound materials were made: 1. The typical default values and ranges for the unbound materials resilient modulus at optimum moisture condition have some wrong values. The problem was corrected. Table 1 provides new guidelines for use with the program for flexible pavements. 2. The software mistakenly uses the default values for the unbound materials resilient modulus at optimum moisture condition 5 AASHTO Soil Classification Mean Modulus (psi)* Standard Deviation (psi) Corrected Mean Modulus (psi) Standard Deviation (psi) Recommended Resilient Modulus at Optimum (psi)** A-1-a 44,471 22,970 29,650 15,315 29,500 A-1-b 39,965 19,428 26,646 12,953 26,500 A-3 37,041 17,853 24,697 11,903 24,500 A-2-4 32,013 19,807 21,344 13,206 21,500 A-2-5 -- -- -- -- 21,000 A-2-6 30,832 18,443 20,556 12,297 20,500 A-2-7 24,373 6,897 16,250 4,598 16,500 A-4 29,797 18,442 16,429 12,296 16,500 A-5 -- -- -- -- 15,500 A-6 26,313 13,657 14,508 9,106 14,500 A-7-5 23,586 19,595 13,004 13,065 13,000 A-7-6 21,159 11,801 11,666 7,868 11,500 * Results are based on 594 back-calculated values extracted from the MON_FLX_BACKCAL_SECT table found in the Long-Term Pavement Performance (LTPP) database. ** Information obtained after correcting the NDT values to reflect laboratory results at optimum conditions. TABLE 1 Recommended resilient modulus input (at optimum density and moisture) for subgrades under flexible pavements and rehabilitation of flexible pavements

when the user chooses Representative Value (Design Value). The problem was corrected and a new set of typical default values and ranges for the Representative Value (Design Value) for the unbound materials resilient modulus has been incorporated. 3. (#157) The software crashes when run- ning an analysis with unbound materials Level 2 seasonal modulus. The problem was corrected. The following improvements and corrections in the treatment of thermal fracture were made: 1. (#71) In the Level 3 predictive system for thermal fracture, as Vbeff (effective bitu- men content) increases, prediction of thermal cracking increases. The problem was corrected and the Level 3 models for the creep compliance and the tensile strength prediction were modified based on a larger database. Separate creep compliance pre- dictive equations were developed for every test temperature (0, −10, −20°C) to replace a universal equation for all temperatures used in the old version of the Level 3 analy- sis. These new models provide thermal frac- ture predictions that meet the test of engi- neering reasonableness. Details of the new models for thermal fracture are provided in Figures 1 and 2. The following issues related to special axle con- figurations were resolved: 1. (#65-2) Asphalt.td file does not read the wheel load coordinates input by the user. Instead, it sets all the coordinates to zeros. The problem was corrected. 2. (#65-3) The output summary for the spe- cial axle configuration shows the wrong data. The problem was corrected. 3. (#65-4) An error message appears when per- forming an analysis using the special axle configuration. The problem was corrected. 4. (#65-5) In Rehab Analysis, the software re- quires the user to input the frequency for the cases of AC, granular base and sub- grade layers; however, the frequency should be a required input only for the AC layers. The problem was corrected. 5. (#65-5) The program does not save the fre- quency values input by the user. In other words, once the user inputs the frequency and clicks “OK,” then returns to the same screen, he/she will find that the previously input frequency values disappeared. The prob- lem was corrected. 6 y = 0.4668x + 229.19 R 2 = 0.4668 Se/Sy = 0.730 n = 42 100 200 300 400 500 600 700 800 100 200 300 400 500 600 700 800 Measured Tensile Strength [psi] Pr ed ic te d Te ns ile S tre ng th [p si] S t = 4976.34 -42.49*Va -2.73*Va 2 -80.61*VFA +0.465*VFA 2 +174.35*log(Pen77) -1,217.54*log(A RTFO ) Figure 1 Revised Level 3 prediction model for tensile strength at −10°C.

7 Temp = -20C y = 0.2583x 0.9101 R 2 = 0.5637 R 2 adj = 0.6204 Se/Sy = 0.624 n = 313 p = 9 Temp = -10C y = 0.2869x 0.9095 R 2 = 0.6397 R 2 adj = 0.5775 Se/Sy = 0.659 n = 313 p = 9 Temp = 0C y = 0.9468x 0.9913 R 2 = 0.7768 R 2 adj = 0.7737 Se/Sy = 0.482 n = 313 p = 9 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-07 1.0E-06 1.0E-05 1.0E-04 Measured D(t) [1/psi] Pr ed ic te d D( t) [1 /ps i] log(D 1 ) -20C = -11.92540+1.52206*log(Va)+4.49876*log(VFA)-3.81320*log(ARTFO) m -20C = -1.75987+1.78187*Va0.02030 +0.00089*Pen770.96870 D(t) = D 1 *t m log(D 1 ) -10C = -10.76560+1.51960*log(Va)+3.49983*log(VFA)-2.99870*log(ARTFO) log(D 1 ) 0C = - 9.80627+1.50845*log(Va)+2.99000*log(VFA)-2.90157*log(ARTFO) m -10C = -1.82690+1.94218*Va0.01600 +0.00098*Pen770.96857 m 0C = -2.41043+2.59093*Va0.01547 +0.00199*Pen770.97247 All Temps y = 1.0059x 0.9983 R 2 = 0.8034 R 2 adj = 0.7998 Se/Sy = 0.449 n = 939 1.0E-07 1.0E-06 1.0E-05 1.0E-04 1.0E-07 1.0E-06 1.0E-05 1.0E-04 Measured D(t) [1/psi] Pr ed ic te d D( t) [1 /ps i] (a) (b) Figure 2 Prediction models for IDT creep compliance D(t) at 0, −10, and −20°C.

The following traffic-related issues in flexible pavement design were addressed: 1. (#60) Potentially significant problem may occur with the use of the current “Repre- sentative Load Analysis Procedure,” espe- cially when evaluating overloaded vehicular traffic scenarios. The problem was corrected and a fixed tire load of 4,500 lb is used as a representative load instead of the variable 95th percentile. This new procedure elimi- nated the potential problems in the distress prediction that might occur with the use of the old procedure. 2. (#70) Axle load distribution factors in traf- fic analysis do not count in the last load level in summation (Total) column. The problem was corrected. 3. (#66) General traffic input screen shows “Dual and Single” tire pressure options; however, it should only show the Dual tire pressure option. The problem was corrected. The following items summarize the major re- calibration of the flexible pavement distress models carried out: 1. A major effort was expended to recalibrate all of the flexible pavement models, including bottom-up fatigue cracking, top-down fatigue cracking, permanent deformation, transverse cracking, and IRI. The final newly calibrated models had a lower model error, reasonable sensitivity to changes in inputs, and better re- liability than those originally developed under NCHRP 1-37A. Major steps in this effort included the following: • Improving the database – Updating all existing sections with 4 to 5 additional years of performance data, traffic data, materials data, climatic data, and rehabilitation data. – Incorporating the weather stations that included 9 years of historic hourly data. – Establishing the proper input subgrade resilient modulus through an iterative process that included back-calculation of in situ moduli for the sections. • Re-establishing the model coefficients for all of the models using the expanded data- base. Model coefficients were selected that minimized the residual error of prediction. • Re-establishing the reliability model co- efficients for all of the models using the expanded database. • Conducting limited sensitivity analyses to validate the software and changes to the software. • Documenting work accomplished (to be published). 2. Revised calibration curves, relevant statistics, and revised models are shown in Figures 3 through 9. 3. Recalibration of the thermal fracture models. Modification of the thermal fracture predictive equations was followed by the necessary recalibration of the Level 3 thermal cracking model. Given that no new sections were available, the 32 sections from the original calibration developed under NCHRP Project 1-37A were used for the recalibration. The final newly calibrated Level 3 model had, overall, much lower residual error of prediction compared with the old model, and, more important, led to thermal fracture predictions that meet the test of engineer- ing reasonableness. A comparison of predicted versus observed ther- mal cracking for newly calibrated Level 3 models and a comparison of standard deviation for new and old models are presented in Figures 10 and 11. The new thermal fracture Level 3 calibration factor and the corresponding standard deviation are Bt3 = 6.0 Std.Dev (Thermal) = 0.0869 ∗ Thermal + 453.98 Given that no new sections were available when compared with the original calibration of the thermal fracture models for Level 1 and Level 2 analyses, the main objective of the new process was to calibrate the thermal cracking models with an expanded (9 years of hourly data) EICM climatic database. The original calibration of the thermal fracture models developed under NCHRP Project 1-37A was done using the real climatic data corresponding to the time in service of the test sections (in some cases, 20 years). However, this calibration was done with- out the use of EICM and its database (because of the limited amount of data) that is the integral part of the MEPDG software. Level 1 and 2 factors from the original calibration task should be used in the foren- sic studies when “true” climatic data are available. However, it was more reasonable to calibrate the models using the same climatic database that will be further used in the performance prediction. Whenever 8

9 SeFCBottom = 1.13+13/(1+e 7.57-15.5*logD ) New fatigue cracking calibration factors are Bf1 = 0.00432 / 0.571 = 0.007566 Bf2 and Bf3 stay the same. 0 10 20 30 40 50 60 70 80 90 100 -4 -3 -2 -1 0 1 2 3 Log Damage (%) Al lig at or C ra ck in g (% of To ta l L an e Ar ea ) Se = 5.01% Se/Sy = 0.815 N = 405 R2 = 0.275 0 1000 2000 3000 4000 5000 6000 7000 0 1000 2000 3000 4000 5000 6000 7000 Measured Cracking (ft/mile) Pr ed ict ed C ra ck in g (ft/ m ile ) R2 = 0.544 Se = 582.8 ft /mile Se/Sy = 0.688 N = 312 Figure 3 Bottom-up fatigue (alligator) cracking calibration and model. Figure 4 Top-down fatigue cracking calibration, measured vs. predicted.

EICM is used to predict the thermal fracture perfor- mance of the pavement, the new calibration factors should be used. It was observed that recalibration re- sulted in higher standard deviations when compared with the original calibration; these higher standard deviations are due to the fact that EICM data were not as precise as the “true” climatic data used in the orig- inal calibration. A comparison of predicted versus observed ther- mal cracking for recalibrated Level 1 and 2 models and a comparison of standard deviation for new and old models are presented in Figures 12 through 15. 10 SeFCtop = 165.68*Log D + 542.53 Damage in % and limit Se to 700 ft/mile 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 -5 -4 -3 -2 -1 0 1 2 3 log Damage (%) Lo ng itu di na l C ra ck in g (ft/ mi le) Measured Longitudinal Cracking Predicted Top-Down Cracking Asphalt Layer - Rutting 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 Average Estimated Measured AC Rutting (in) Pr ed ict ed A C Ru tti ng (in ) Predicted vs Ave. Estimated Measured AC Rutting Equality Line R2 = 0.64 N = 334 Se = 0.045 Se/Sy = 0.713 Figure 5 Top-down fatigue cracking calibration and model. Figure 6 Permanent deformation calibration—rutting in the asphalt layer.

The new thermal fracture Level 1 and Level 2 calibration factors and corresponding standard devi- ations are Bt1 = 1.0 Std.Dev (Thermal) = −0.0899 ∗ Thermal + 636.97 Bt2 = 0.5 Std.Dev (Thermal) = −0.0169 ∗ Thermal + 654.86 Recalibration permitted the following updating of the flexible pavement IRI models: 1. Considerable additional flexible pavement dis- tress and IRI data have become available since the original model was developed in 1999. Such data were extracted from LTPP and used in the analysis. 2. Two models were developed: • New flexible pavements and overlaid flexi- ble pavements combined. • HMA overlays of jointed concrete pave- ments. 11 Granular Base - Rutting 0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3 0.4 0.5 Average Estimated Measured GB Rutting (in) Pr ed ict ed G B Ru tti ng (in ) Predicted vs Est. Measured GB Rutting Equality Line R2 = 0.785 N = 334 Se = 0.026 Se/Sy = 0.502 Subgrade - Rutting 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.1 0.2 0.3 0.4 0.5 0.6 Average Estimated Measured SG Rutting (in) Pr ed ic te d SG R ut tin g (in ) Predicted vs Est. Measured SG Rutting Equality Line R2 = 0.708 N = 334 Se = 0.045 Se/Sy = 0.576 Figure 8 Permanent deformation calibration—rutting in the subgrade. Figure 7 Permanent deformation calibration—rutting in the granular base.

3. The updated model included the key M-E pre- dicted distress types as follows: • Fatigue cracking: bottom up and top down, all severities. • Permanent deformation, mean values. • Transverse cracking (all severities). 4. The updated models also included the fol- lowing site factors: mean annual air temper- ature; freezing index; annual precipitation; percentages of silt, clay, and sand in the sub- grade soil; and age of the section. 5. The re-calibrated IRI models were an im- provement on the existing models because • The functional form is more similar to IRI for rigid pavements. • Problematic non-predicted distress models were removed. • They are based on a large number of addi- tional LTPP sections. • They are based on sections with additional distress and IRI development over time (1998–2005). • They directly consider the effect of each M-E-predicted distress on IRI: – Permanent deformation – Wheelpath cracking: Alligator and Longitudinal fatigue – Transverse cracking • They consider the effect of site conditions: – Subgrade: percent fine sand, silt, clay, and PI – Climate: freezing index, precipitation – Age: represents cycles of hot/cold, wet/ dry, freeze/thaw 12 Total Pavement - Rutting 0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2 Average Measured Total Rutting (in) Pr ed ict ed T ot al R ut tin g (in ) Predicted vs Measured Total Rutting Equality Line R2 = 0.577 N = 334 Se = 0.107 Se/Sy = 0.818 SeRDAC = 0 .24 ACrut^0.8026 limit to a max value of 0.1 SeRDGB = 0.1477 GBrut^0.6711 limit to a max value of 0.055 SeRDSG = 0.1235 SGrut^0.5012 limit to a max value of 0.065 New Rutting calibration factors are Br1 = -3.35412, Br2 and Br3 stay the same. BrGB = 2.03, BrSG = 1.67 Figure 9 Permanent deformation calibration—total pavement rutting.

13 y = 0.2279x + 1330.8 R 2 = 0.064 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Observed Cracking [ft/mile] Pr ed ic te d Cr ac ki ng [f t/m ile ] Figure 10 Thermal cracking model calibration DG2002 Level 3— calibration factor = 6.0. 36 9 70 9 10 50 13 90 57 8 66 2 74 7 49 3 0.0 200.0 400.0 600.0 800.0 1,000.0 1,200.0 1,400.0 1,600.0 500 1,000 1,500 2,000 Thermal Cracking [ft/mile] St d. D ev ia tio n Old Calib New Calib Figure 11 Level 3 thermal cracking model—old vs. new calibration. 13 4 25 8 3 82 50 5 5 92 54 7 50 2 45 7 0.0 200.0 400.0 600.0 800.0 1,000.0 1,200.0 1,400.0 1,600.0 500 1,000 1,500 2,000 Thermal Cracking [ft/mile] St d. D ev ia tio n Old Calib New Calib Figure 12 Thermal cracking Level 1 model—old vs. new calibration.

• Sensitivity shows the reasonableness of effects of distress on IRI. • There is a far larger database of sections: – Flexible pavement IRI: 1950 vs. 1978 – Five years of additional distress devel- opment (to better determine the effect on IRI) – Standard error of prediction slightly lower – 18.9 vs. 24.5 in/mi 6. Variance models were developed for use in reliability design with IRI. They produce re- alistic results and are similar to those devel- oped for IRI of rigid pavements. 14 Calibration Factor = 1.0 y = 0.7001x + 73.142 R 2 = 0.405 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Observed Cracking [ft/mile] Pr ed ic te d Cr ac ki ng [f t/m ile ] Figure 13 Thermal cracking Level 1 model calibration, observed vs. predicted. Calibration Factor = 0.5 y = 0.5202x + 314.64 R 2 = 0.2219 0 500 1000 1500 2000 2500 0 500 1000 1500 2000 2500 Observed Cracking [ft/mile] Pr ed ic te d Cr ac ki ng [f t/m ile ] Figure 14 Thermal cracking Level 2 model calibration, observed vs. predicted.

7. Sensitivity analyses were run to verify the reasonableness of the IRI models. 8. Figures 16 and 17 show the predicted versus measured IRI for new flexible pavements, HMA over flexible pavements, and HMA over rigid pavements. RIGID PAVEMENTS The following modifications were made in re- sponse to items identified in the Bug Tracker data- base (www.ara-tracker.com): 1. (#434) CPR model for slab cracking. The CPR algorithm for damage and slab crack- ing (after CPR) was found deficient because it uses percent cracked/replaced slabs to ob- tain an estimate of overall accumulated fa- tigue damage. This damage was neither top nor bottom fatigue but a composite of both that was not very accurate. Top-down and bottom-up damage needs to be directly cor- related with measured cracking and an im- proved methodology installed. This upgrade was completed and verified. The CPR crack- ing model is now more theoretically correct and provides a better projection of future cracking after diamond grinding and other repairs. 2. (#433) Widened slab computation defi- ciency. A deficiency in the fatigue damage algorithm was identified for JPCP. The pro- gram does not provide for proper damage calculation for widening less than 24 inches. The problem was fixed so that user may enter 3, 6, 12, etc., up to 24 inches and the program will calculate fatigue damage properly. After approximately 15 inches, the critical fatigue location shifts from the outer slab edge of the widened portion to the inner slab edge near the lane-to-lane longitudinal joint. The algo- rithm makes this switch properly. 3. (#428) HMA overlay of JPCP and CRCP includes several major deficiencies that re- quire updating. The modeling of HMA over JPCP and CRCP in the existing version has serious deficiencies in how the overlay and concrete slab and base course are transformed into an equivalent section for stress calcula- tion purposes. Major modifications are re- quired for both HMA over JPCP and HMA over CRCP to make this a more effective overlay design procedure. [Note: these mod- ifications have not been completed in Ver- sion 0.900 yet. It is recommended that this overlay design procedure not be used until this is completed in late July 2006.] 4. (#425) Level 1 inputs for PCC incorrect for HMA over JPCP and HMA over CRCP. The program is using the xx year (e.g., 30 year) flexural strength instead of the input (long term) existing strength. The problem was corrected. 5. (#393) CRCP shoulder load transfer fac- tor (Js) reporting incorrectly. The problem was corrected. 6. (#386) When designing the AC overlay of JPCP, the design screen displays the AC overlay instead of the PCC thickness. The problem was corrected. 7. (#385) CRCP crashes on exit for analysis of more than 500 months. The problem was corrected. 15 18 3 35 2 52 0 68 9 63 8 63 0 62 1 64 6 0.0 200.0 400.0 600.0 800.0 1,000.0 1,200.0 1,400.0 1,600.0 500 1,000 1,500 2,000 Thermal Cracking [ft/mile] Old Calib New Calib St d. D ev ia tio n Figure 15 Thermal cracking Level 2 model, old vs. new calibration. 0 50 100 150 200 0 50 100 150 200 Measured IRI, in/mi Pr ed ic te d IR I, in /m i N = 1926 R 2 = 56 percent SEE = 18.9 in/mi Figure 16 IRI for new HMA and HMA over HMA flexible pavements.

8. (#382) Open to traffic date and con- struction dates that differ by more than 11 months cause crashes for JPCP. The problem was corrected. 9. (#364) Remove drainage inputs from in- terface. Use of the TMI model makes entry of drainage path and infiltration unnecessary. 10. (#362) Erosion models for CRCP were re- vised to show more effect on performance per review comments. Major revisions were made to the models and incorporated into the program. Tests indicated further bugs that were fixed. 11. (#359) JPCP Cracking model stopped at 40 Years. The problem was corrected. De- sign life can be over 80 years. 12. (#345) Addition of algorithm to com- pute equivalent temperature gradient for “bonded” slab/base conditions. “Bonded” herein means full-contact friction with no slippage between layers at the interface. The existing JPCP model includes a proce- dure for computing the equivalent linear temperature gradient through the concrete slab given a non-linear gradient. The pro- cedure assumes that the slab and base are unbonded. A similar procedure is needed to compute an equivalent linear temperature gradient through the concrete slab for bonded (no slippage, full-contact friction) condi- tions. The significance of the bonding con- dition has been determined to be very high. This change required a change to the lin- earization and damage calculation algo- rithms of the MEPDG cracking model. Note: It was discovered that an error ex- isted in the 2004 Version 0.7 of the Design Guide wherein bonding of slab and base only lasted for 12 months, regardless of the input by the user. This was fixed under #343 and contact friction or bonding now works properly. 13. (#344) Modification of JPCP computa- tional algorithm that includes “design pe- riods” due to various problems. JPCP com- putational algorithm includes periods over which fatigue damage is being computed throughout the design analysis period. These were introduced to reduce the run time. An error was identified in conjunction with the “bonding” slab/base algorithm and extra- polation procedures over the design periods that are not easily solved. The design period procedure may not have as much effect on computer run time as previously thought and it was modified so that the damage accumu- lation will be done month by month and year by year using exact strength, modulus, con- tact friction, k-values, and so forth. This ap- proach will also help the implementation of other features into the JPCP design proce- dure. CRCP already works in this way. 16 0 25 50 75 100 125 150 0 25 50 75 100 125 150 Measured IRI, in/mi Pr ed ic te d IR I, in /m i N = 244 R2 = 51 percent SEE = 9.6 in/mi Figure 17 IRI for HMA overlaid PCC pavements.

14. (#343) Contact friction between concrete slab and base course (commonly called “bonding”) has an error in computatio- nal algorithm. The user can select the num- ber of months over which the slab and base will remain “bonded” (this is really contact friction). Some sensitivity analysis indicated this input to have small effect. Investigation showed that an error in the coding is causing this result. Subsequent runs and the calibra- tion process indicated that most slabs and base courses show more full-contract fric- tion over their lives than previously be- lieved. Design input recommendations were revised. 15. (#241) Difference in CRCP results be- tween estimated crack spacing and user input crack spacing. There was a differ- ence in CRCP output results when the pro- gram uses the internal model to predict crack spacing and when the user inputs the exact same crack spacing. For example, if the pro- gram model predicts 31.7 in, and this is in- put and the program re-run, a different out- put may result. The problem was fixed. 16. (#240) Permanent curl/warp input needs further examination to determine im- proved estimation procedures. Currently, a −10°F is recommended for design. This is inadequate as the permanent curl/warp is known to depend on several key factors. De- velop procedures to estimate the permanent curl/warp input for JPCP and CRCP sepa- rately. Through the calibration process, at- tempt to identify values or relationships that will minimize error of prediction for all JPCP and CRCP distress models. Time and re- sources were insufficient to solve this prob- lem at this time. It is recommended that it be addressed in the next program version. 17. (#239) Base erosion index does not appear to have adequate sensitivity to perfor- mance. Evaluated the effect of the base ero- sion index on performance of CRCP and JPCP and determined that it needed im- provement. New erosion models were de- rived for each type of base course and sub- grade type so that the loss of support along the edge is computed automatically, not dependent on user input of erosion factor. 18. (#238) Add CRCP design criteria (Crack Width, Crack LTE, Crack Spacing) to Output Reliability Screen. CRCP design criteria include crack width, crack spacing, and crack LTE in addition to punchouts. These were added to the Reliability output screen to emphasize that they are just as im- portant as punchouts and IRI (even though there is no reliability level associated with them). • Crack Width < 0.02 in • Crack LTE > 95 percent • Crack spacing 3 to 6 ft The following modifications were completed, but were not included in the Bug Tracker database: 1. The definition of a CRCP punchout was re- vised, based on review comments, to include only medium- and high-severity punchouts and y-cracks. Distress maps for all CRCP sections were reviewed and the correct num- ber of punchouts (including y-cracks) was in- cluded for each time frame. This change is re- flected in the design input recommendations for critical levels for design. 2. The AC/JPCP or AC/CRCP overlay design procedure was found to contain various tech- nical deficiencies. One problem was that the 2004 version did not fully consider the width of transverse cracking (after many years of aging), the load transfer efficiency (which may have deteriorated), and the extent of erosion along the slab edge that exists in the field at the time of placement of a new overlay. In ad- dition, procedures to calculate the equivalent slab thickness (where the overlay is com- bined with the CRCP slab) with proper full- friction included errors. These deficiencies are being fixed in the software and this type of overlay is being tested to ensure reason- ableness. This fix is not in Version 0.900 but will be in the next version. 3. A major effort was expended to recalibrate all of the JPCP, CRCP, and rehabilitation distress models, including joint faulting (new and re- hab), slab top-down and bottom-up cracking (new and rehab), punchout (new and rehab), and crack spacing. The final newly calibrated models had a lower model error even with an expanded data set, reasonable sensitivity to changes in inputs, greater robustness because of an expanded data set, and better reliability than those originally developed under NCHRP 17

1-37A. Major steps in this effort included the following: • Improving the database. – Updating all existing sections with 4 to 5 additional years of performance data, traffic data, materials data, climatic data, and rehabilitation data. – Identifying additional LTPP and other sections for inclusion in the calibration. – Incorporating the weather stations that included 9 years of historic hourly data. – Establishing the proper input subgrade resilient modulus through an iterative process that included back-calculation of in situ moduli for the sections. • Re-establishing the model coefficients for all the models using the expanded database. • Re-establishing the reliability model co- efficients for all the models using the ex- panded database. • Conducting limited sensitivity analyses to validate the software and changes to the software. • Documenting work accomplished (not yet completed). Figures 18 through 20 show measured and pre- dicted JPCP cracking distress for new JPCP, unbonded JPCP overlays, and JPCP subjected to restoration. Figures 21 through 23 show the measured and pre- dicted JPCP faulting distress for new JPCP, unbonded JPCP overlays, and JPCP subjected to restoration. Figure 24 shows the measured and predicted CRCP punchout distress for new CRCP. The following improvements were made to the input guidelines: 1. Input resilient modulus, Mr, for sub- grade. Calibration required a complete re- establishment of the proper input subgrade Mr at optimum moisture content that would ultimately provide an in situ modulus that matched the FWD back-calculated value at each test section site. The modification of the ICM and subdrainage models resulted in a change of in situ Mr resulting in this required work. As a result of this, the JPCP and CRCP 18 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Measured percent slabs cracked Pr ed ict ed p er ce nt s la bs c ra ck ed R2 = 0.86 SEE = 5.02 percent N = 1585 Figure 18 Transverse cracking for new JPCP calibration. 0 10 20 30 40 50 60 70 80 90 100 0 10 20 30 40 50 60 70 80 90 100 Measured percent slabs cracked Pr ed ic te d pe rc en t s la bs c ra ck ed R 2 = 0.72 SEE = 3.95 percent N = 60 Figure 19 Transverse cracking for unbonded JPCP overlay calibration.

sections provide a large database from which to obtain typical values for Mr for each AASHTO soil class. It was discovered that the in situ Mr varied greatly for any given LTPP classification of subgrade soils and, thus, the selection of typical values for use as defaults or recommended values for Level 3 was diffi- cult. The values obtained were averages from all of the data and represent the best Level 3 source of information available. Separate re- commendations for rigid pavements made based on these results are presented in Table 2. 2. Input resilient modulus, Mr, for base/ subbase. The values recommended in the 2004 version of the Design Guide are recom- mended for the base and subbase. 3. Grain size distribution, Atterberg limits, and other soil properties were updated based on extensive data from LTPP test sites across the United States. 4. Contact friction between slab and base (for- merly called bonding). This value was found to be far more significant than before because of an error in the software. Guidelines on how many months of full friction to select for dif- ferent base courses were based on best fit to match cracking in the field. Results are as fol- lows which are recommended for input: • Asphalt stabilized base: 60 to 360 months with an average of 229 months. • Cement stabilized or lean concrete base: 0 to 360 months with an average of 136 months. • Unbound material base: 0 to 360 months with an average of 245 months. • Lime stabilized base: 0 to 360 months with an average of 176 months. • Unbonded overlay (with HMA separation layer): 0 months of full-friction (bond) to match cracking. • These values were used in the calibration and are recommended for design. 5. Recommendations for the modulus of exist- ing PCC slabs for use in unbonded overlay design were updated. 19 0 10 20 30 40 50 60 70 80 90 100 0 20 40 60 80 100 Measured percent slabs cracked Pr ed ic te d pe rc en t s la bs c ra ck ed R 2 = 0.90 N = 94 SEE = 6.5 percent Figure 20 Transverse cracking for restored JPCP (CPR) calibration. 0 0.05 0.1 0.15 0.2 0.25 0.3 0 0.05 0.1 0.15 0.2 0.25 0.3 Measured mean transverse joint faulting, in Pr ed ict ed m ea n tr a n sv e rs e joi nt fau lti ng , i n R2 = 0.62 SEE = 0.0276 in N = 1260 Figure 21 Transverse joint faulting for new JPCP calibration.

6. Permanent curl/warp gradient through slab. A value of −10°F was found in the original calibration in 2002–2004 to provide the lowest error in slab cracking prediction. This value was used for nearly all projects of JPCP and CRCP in the 2006 calibration. However, those projects cured with water or constructed at night required a lower value (−3°F) and those built under harsh curing conditions (e.g., morning paving, sunshine, wind) often required greater than −10°F, even up to −25°F. Additional research is needed to quantify this important input. 7. CTE recommendations were upgraded to the latest LTPP data analysis of this parameter as shown in Table 3. 8. Improved estimates of wheel base percent- ages were obtained from two states. They were somewhat different than the 33, 33, and 34 percent assumed for short, medium, and 20 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.00 0.05 0.10 0.15 0.20 0.25 0.30 Measured mean transverse joint faulting, in Pr ed ict ed m ea n tr a n sv e rs e joi nt fau lti ng , i n R2 = 0.74 SEE = 0.025 in N = 43 Figure 22 Transverse joint faulting for unbonded JPCP overlays calibration. 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 Measured transverse joint faulting, in Pr ed ict ed tr a n sv e rs e jo int fa u lti ng , i n R2 = 0.61 N = 40 SEE = 0.02 in Figure 23 Transverse joint faulting for restored JPCP (CPR) calibration.

21 y = 0.9993x R2 = 0.7395 0 10 20 30 40 50 60 0 10 20 30 40 50 60 Predicted punchouts, #/mile M ea su re d pu nc ho ut s, # /m ile Figure 24 Plot of predicted versus measured punchouts for new CRCP calibration. Subgrade AASHTO Soil Class Optimum Dry Density (mean, std. dev.)* Optimum Moisture Content (mean)* % Design Guide Input Resilient Modulus at Optimum Density/Moist. (mean, std. dev.)** Design Guide Back-calculated Output Dynamic k-value (mean, std. dev.)** Recommended Input Subgrade Resilient Modulus (Opt. Density/Moisture Content) A-1-a 128 pcf, 17 pcf 11 13,228 psi, 3,083 psi 322 psi/in, 68 psi/in 18,000 psi A-1-b 122, 9 11 14,760, 8,817 335, 92 18,000 A-3 NA NA NA NA 16,500 A-2-4 119, 7 11 14,002, 5,730 256, 79 16,000 A-2-5 NA NA NA NA 16,000 A-2-6 120, 6 12 16,610, 6,620 289, 51 16,000 A-2-7 NA NA NA NA 16,000 A-4 119, 7 12 17,763, 8,889 270, 88 15,000 A-5 NA NA NA NA 8,000 A-6 114, 5 14 14,109, 5,935 211, 54 14,000 A-7-5 103, 19 19 7,984, 3,132 148, 32 10,000 A-7-6 102, 8 20 13,218, 322 203, 53 13,000 *Information provided in these columns was obtained from the LTPP database (optimum density and moisture). **Information was obtained from Design Guide back-calculation and from use of the Design Guide (input subgrade resilient modulus, Mr, at optimum density and moisture). 1These results are based on about 250 JPCP and CRCP pavements located across the U.S. and used in the calibration of the Design Guide rigid pavements. 2Use of resilient modulus input (at optimum density and moisture) for a project that is significantly different than these test results and recommendations may result in erroneous model prediction. Specifically, input of higher resilient modulus that results in significantly higher output dynamic k-values may result in erroneous model prediction as very few LTPP sections across the country showed higher values. Note that bedrock close to the surface would be an exception to this guideline. 3Do not use these resilient modulus values for compacted base or subbase course. Use appropriate table for base/subbase course resilient modulus. TABLE 2 Recommended subgrade/embankment resilient modulus input (at optimum density and moisture) for rigid pavements and rehabilitation of rigid pavements1,2,3

long wheel bases used in the original design guide calibration. Further study is needed to more firmly establish the wheel base per- centages and then they will be used in a JPCP calibration. Major technical improvements in Version 0.900 for rigid pavements are summarized as follows: 1. Addition of algorithm to compute equivalent temperature gradient for “bonded” slab/base conditions (unbonded only was available). This was used in calibration whenever full- contact friction was specified. 2. Modification of JPCP computational algo- rithm so that it now includes month-by-month damage accumulation, rather than accumulat- ing over a multiyear period. 3. Fixed a bug in the algorithm for contact fric- tion between concrete slab and base (“time to de-bond”). The term “bond or bonding” was changed to “contact friction” to better describe the amount of slippage between the PCC slab and the base course. Recommen- dations were provided. 4. Erosion prediction for CRCP was improved. • Program calculates loss of support along edge over time as a function of – Base type and quality – HMA: asphalt content – CTB: Ec – Granular: fines content – Annual precipitation – Type and quality of subbase/subgrade (strength, fines) • Erosion calculated for 10 years, but uni- formly accumulated year by year with practical cap. 5. Re-calibrations of all JPCP and CRCP mod- els were successful. Partial sensitivity indi- cates the calibrated models are reasonable. Further sensitivity analyses are needed. 6. Re-calibration of all rehabilitation models for JPCP and CRCP were successful. 7. Results confirm that original 2004 models were valid over a much wider range of design parameters and a larger number of additional sections throughout the United States. 8. Joint spacing for skewed joints was increased by the amount of skew in a 12-ft-wide slab (normally 2-ft) to account for increased curl/ warp stresses over that of a perpendicular joint. 9. Improved recommendations for subgrade resilient modulus inputs are provided. 10. Slab widening algorithm for JPCP was im- proved to allow smaller widening values (e.g., 3 to 24 inches) with proper calculation of fatigue damage at each edge of slab. 11. Concrete pavement restoration algorithm for cracking was upgraded to predict top-down and bottom-up damage and cracking more accurately. 22 Primary Origin Primary Aggregate Class Average CTE2 Standard Deviation (s) Sample Count (n) Igneous (Extrusive) Andesite 5.3 0.5 23 Igneous (Extrusive) Basalt 5.2 0.7 47 Igneous (Extrusive)1 Diabase 4.6 0.5 4 Igneous (Plutonic) Diabase 5.2 0.5 17 Igneous (Plutonic)1 Gabbro 5.3 0.6 4 Igneous (Plutonic) Granite 5.8 0.6 83 Metamorphic Schist 5.6 0.5 17 Sedimentary Chert 6.6 0.8 28 Sedimentary Dolomite 5.8 0.8 124 Sedimentary Limestone 5.4 0.7 236 Sedimentary Quartzite 6.2 0.7 69 Sedimentary Sandstone 6.1 0.8 18 Lightweight1 Expanded shale 5.7 0.5 3 LTPP test section results. Testing conducted by FHWA at the Turner-Fairbank Highway Research Center. Mallela, J., et al. (2005) ìMeas urement and Significance of the Coefficient of Thermal Expansion of Concrete in Rigid Pavement Design” Transportation Research Record 1919, Transportation Research Board, Washington, DC. 1 Results based on very limited testing. 2 units are in in/in per °F x10-6. TABLE 3 PCC CTE results sorted by aggregate origin and classification

Transportation Research Board 500 Fifth Street, NW Washington, DC 20001 These digests are issued in order to increase awareness of research results emanating from projects in the Cooperative Research Programs (CRP). Persons wanting to pursue the project subject matter in greater depth should contact the CRP Staff, Transportation Research Board of the National Academies, 500 Fifth Street, NW, Washington, DC 20001. COPYRIGHT PERMISSION Authors herein are responsible for the authenticity of their materials and for obtaining written permissions from publishers or persons who own the copyright to any previously published or copyrighted material used herein. Cooperative Research Programs (CRP) grants permission to reproduce material in this publication for classroom and not-for-profit purposes. Permission is given with the understanding that none of the material will be used to imply TRB, AASHTO, FAA, FHWA, FMCSA, FTA, or Transit Development Corporation endorsement of a particular product, method, or practice. It is expected that those reproducing the material in this document for educational and not-for-profit uses will give appropriate acknowledgment of the source of any reprinted or reproduced material. For other uses of the material, request permission from CRP.

Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006 Get This Book
×
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

TRB's National Cooperative Highway Research Program (NCHRP) Research Results Digest 308: Changes to the Mechanistic Empirical Pavement Design Guide Software Through Version 0.900, July 2006, summarizes numerous changes made to the original Version 0.7 of the Mechanistic-Empirical Pavement Design Guide (MEPDG) software. Specifically, the digest describes the corrections incorporated in Version 0.8 (released in November 2005) and the mainly technical improvements and enhancements included in Version 0.900 (released in July 2006).

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!