WELCOME TO THE WEBINAR This Webinar is being

WELCOME TO THE WEBINAR • This Webinar is being presented from a computer with a resolution of 1280 X 1024. • Using a similar setting on your machine may improve your viewing experience.

FY 2018 Enhancements Released with Version 2. 5 21 August 2018

FY 2018 Enhancements Released with Version 2. 5 Presenters: Moderator: John Donohue, Missouri Department of Transportation; Chair Harold Von Quintus, ARA Wouter Brink, ARA Presentation will be available for viewing on the MEDesign Resource website: http: //www. me-design. com

FY 2018 Enhancements Released with Version 2. 5 Phones are being muted. Please post your questions in the Q&A box. This can be accessed by clicking on the Webex Q&A button. The presenters will answer all questions at the end of the webinar/demonstration as time permits. Questions not answered because of time, will be responded to separately.

Process for Enhancing the Software: 1. Suggested revisions are received from users and AASHTO members. 2. Pavement ME Design Task Force monitors ongoing/completed NCHRP, FHWA, and pool fund projects applicable to the MEPDG. 3. Pavement ME Design Task Force members review and prioritize all potential suggested revisions and enhancements.

Pavement ME Task Force Members 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. John Donahue, P. E. ; Missouri DOT, Chairperson Vicki Schofield, AASHTO Project Manager Marta Juhasz, P. E. ; Alberta Transportation, Vice-Chair Clark Morrison, P. E. , North Carolina DOT Robert Shugart, P. E. ; Alabama DOT Karen Strauss, P. E. ; Oregon DOT David Holmgren, P. E. ; Utah DOT Patrick Bierl, P. E. ; Ohio DOT Felix Doucet; TAC Liaison Tom Yu, P. E. ; FHWA Liaison Shane Marshall, P. E. ; Utah DOT, SCOJD Liaison Travis Tackett, Florida DOT; T&AA Liaison

FY 2018 Enhancements Released with Version 2. 5 Recap of Past Webinars on FY 2018 Enhancements I. Webinar #1: August 1 a) b) c) d) Manual of Practice Integration Customization of output reports Comparator tool Use of MERRA climate II. Webinar #2: August 14 a) Reset Performance Predictions b) Input Level 1 for Indirect Tensile Strength

FY 2018 Enhancements Released with Version 2. 5 Todays Webinar on FY 2018 Enhancements, The Purpose: I. We will present results from the recalibration of flexible and semi-rigid pavements, and the impact of the new model coefficients. We are not describing or presenting details of the recalibration process itself.

FY 2018 Enhancements Released with Version 2. 5 Before we get started: Poll 1: Questions 1 and 2

Questions Asked from Recalibration Results General questions: 1. Why were the flexible pavement prediction models recalibrated? 2. Is there a significant difference between the NCHRP 1 -40 D global model coefficients and those derived from the recalibration using version 2. 5? 3. Why are the predictions between versions 2. 5 and 2. 3. 1 using the appropriate calibration coefficients different? 4. Which version of the software and associated calibration coefficients are the more accurate?

Recalibration: Flexible and Semi-Rigid Pavements I. III. IV. V. VI. Why Recalibrate? Rut Depth Transverse Cracking, new flexible Alligator Area Cracking, new flexible Reflection Cracking Smoothness

Why Recalibrate Flexible and Semi-Rigid Pavements? I. Flexible pavements have not been globally calibrated since FY 2002. II. Semi-rigid pavements were never globally calibrated, except for reflection cracking. III. Technical audit of software completed in FY 2016, and a few discrepancies were identified. IV. LTPP made recent decision to include MERRA climate database.

Recalibration Results for Flexible and Semi. Rigid Pavements AASHTO recognizes calibration is complex and is preparing a tool (The Calibrator) to assist agencies in the verification/calibration process to: 1. Reduce the level of effort. 2. Formalize or establish a structured procedure that is easier to follow. 3. Provide direction on making decisions during calibration process. 4. Etc.

Why Recalibrate Flexible and Semi-Rigid Pavements? Flexible pavements globally calibrated in FY 2002, so an additional 15+ years of data in LTPP. Question to be Answered: Do earlier global calibration coefficients accurately predict the measured growth of distress and smoothness?

Why Recalibrate Flexible and Semi-Rigid Pavements? Verify impact of added data used in recalibration.

Why Recalibrate Flexible and Semi-Rigid Pavements? Bottom-up fatigue cracks are much greater.

Recalibration Results for Flexible and Semi. Rigid Pavements Multiple addenda/memoranda prepared to present results from recalibration: 1. Revised Global Model Coefficients and Field Adjustment Parameters for version 2. 5. Addendum, FY 2018. 4, June 30, 2018. 2. Using Local Calibration Coefficients with Pavement ME Design Version 2. 5. Addendum, FY 2018. 5, June 30, 2018. 3. Why Recalibrate? Memorandum, FY 2019. 2; August 15, 2018

Recalibration Results for Flexible and Semi. Rigid Pavements Multiple addenda/memoranda prepared to present results from recalibration; Go to: ME-Design. com

Recalibration: Flexible and Semi-Rigid Pavements I. III. IV. V. VI. Why Recalibrate? Rut Depth Transverse Cracking, new flexible Alligator Area Cracking, new flexible Reflection Cracking Smoothness Process used for all distress/performance prediction models was to first verify previous global prediction model/transfer function coefficients.

Total Rut Depth Iterative Process; segregated sections into 3 groups: I. AC Layer Rut Depths for AC Overlays • Assumption: rutting in unbound layers is minimal. II. Unbound Layer Rut Depths for Thin Flexible Pavements • Assumption: AC layer rutting properly predicted from first step. III. Total Rut Depths for Intermediate to Thick AC Pavements to Remove any Bias

AC Rut Depth I. Version 2. 3. 1 and earlier versions • • Kr coefficients were derived from global calibration combining laboratory test results and field adjustments. NCHRP 1 -37 A reported laboratory and field values; page 3. 3. 49 in the report. Lab Values: Kr 1 = -3. 15552 Kr 2 = 1. 734 Kr 3 = 0. 39937 Field Values: Br 1 = 0. 509 Br 2 = 0. 90 Br 3 = 1. 2

AC Rut Depth I. II. Version 2. 3. 1 and earlier versions: Kr coefficients combined laboratory test results and field adjustments. Version 2. 5 – separates laboratory test results and field adjustments. Version 2. 5: Kr lab values Version 2. 3. 1: Kr values Version 2. 5: βr values

AC Rut Depth, Laboratory Derived Coefficients

AC Rut Depth, Laboratory Derived Coefficients Tertiary Region Primary Region IT R = Elastic Strain T = Temperature Steady State Region ST I = Intercept from steady state region S = Slope in the steady state region

AC Rut Depth Multiple data sources were used to determine the Kr, laboratory-derived, values. I. NCHRP 1 -37 A Predominantly II. NCHRP 9 -30 A, test procedure dense-graded neat III. FHWA, ETG test programs asphalt concrete IV. AAMAS project mixtures, with a few V. Colorado DOT PMA mixtures. VI. Pennsylvania DOT VII. Asphalt Institute (confined test results) VIII. Etc.

AC Rut Depth Final results from the global calibration, as compared to earlier version. Type of Mixture Dense. Graded Neat Asphalt Version 2. 3. 1 Version 2. 5; Lab-Field Coefficient Lab-Field, [1 -37 A Rpt. ] Combined, [1 -40 D] Kr 1 -3. 15552 -3. 35412 -2. 45 1. 734 1. 5606 3. 01 Kr 3 0. 39937 0. 4791 0. 22 0. 509 1. 0 0. 40 βr 2 0. 90 1. 0 0. 52 βr 3 1. 2 1. 0 1. 36 Kr 2 βr 1

AC Rut Depth Kr 3 Br 3 = 0. 4791 10 Kr 1 Br 1 = -3. 35791 s with n e t t a l f e Slop g K r 3 decreasin Kr 3 Br 3 = 0. 299 10 Kr 1 Br 1 = -2. 85

Unbound Layer Rut Depth Total rut depths for thin flexible pavements exhibit much higher variability between predicted and measured values. Thin AC: < 6 inches

Unbound Layer Rut Depth Final results from the global calibration, as compared to earlier versions. Type of Material Fine-Grained Subgrade Sand, A-3 Coarse-Grained Subgrade Aggregate Base Coefficient Ks(Fine-Grained) Ks(Coarse-Grained) Ks(Sand) βs 1(Subgrade) Ks 1(Base) βs 1(Base) Version 2. 3. 1 Version 2. 5 0. 675 1. 35 0. 635 0. 965 1. 0 2. 03 0. 965 1. 0

Unbound Layer Rut Depth Total rut depths for thin flexible pavements exhibit much higher residual errors (differences) between predicted and measured values.

Unbound Layer Rut Depth Total rut depths for thin flexible pavements exhibit much higher residual errors between predicted and measured values.

Total Rut Depth High residual errors (predicted minus measured values) are thin (< 6 inches) AC pavements.

Total Rut Depth Statistical Parameters Original Values Version 2. 5 Values n, Observations R 2 Se Se/Sy 334 0. 577 0. 107 0. 818 916 0. 51 0. 11 0. 78 N, Projects 126 Remember: Total rut depths for thin flexible pavements exhibit much higher residual errors between predicted and measured values.

Total Rut Depth In summary: use of version 2. 5 model coefficients results in lower predicted rut depths, especially for higher volume roadways.

Total Rut Depth, Summary Rut depth model and calibration coefficients Dense-Graded Neat Asphalt Concrete Mixes Field adjustment coefficients Lab derived model coefficients

Recalibration: Flexible and Semi-Rigid Pavements I. III. IV. V. VI. Why Recalibrate? Rut Depth Transverse Cracking, new flexible Alligator Area Cracking, new flexible Reflection Cracking Smoothness

Transverse Cracks Global model coefficients derived for only new flexible pavements. AC overlays and semi-rigid pavements excluded because of possible transverse reflection cracks.

Transverse Cracks Indirect tensile strength and creep compliance are the two critical properties measured in lab, and both are temperature dependent.

Transverse Cracks Prediction Model: transverse cracks are caused by cold temperature events. However, transverse cracks exhibited on many flexible pavements in mild to warm climates. Sites in warm to cold climates included in verification process.

Transverse Cracks

Transverse Cracks Residual errors (predicted minus measured values) found to be dependent on mean annual air temperature or freezing index.

Transverse Cracks MAAT used to estimate Kt values for predicting transverse cracks. MAAT at 57 ⁰F is the transition between cold temp. events and another mechanism. Transverse cracks caused by shrinkage or another mechanism? Transverse cracks caused by cold temp. events.

Transverse Cracks Final results from the global calibration for input level 3, as compared to earlier versions. Parameter Kt MAAT Version 2. 3. 1 > 57 ⁰F 1. 5 < 57 ⁰F 1. 5 Version 2. 5

Transverse Cracks

Transverse Cracks Statistical Parameters N, Projects n, Observations R 2 Se Se/Sy Original Values 0. 064 Version 2. 5 Values 95 594 0. 25 462 0. 85

Transverse Cracks

Transverse Cracks Comparison of predicted transverse cracks using versions 2. 3. 1 and 2. 5. In summary: use of version 2. 5 model coefficients results in higher predicted transverse cracks for mild to warm climates and little change for cold climates.

Transverse Cracks Transverse crack model and calibration coefficient; Ks 1

FY 2018 Enhancements Released with Version 2. 5 Poll 2: Questions 3 and 4

Recalibration: Flexible and Semi-Rigid Pavements I. III. IV. V. VI. Why Recalibrate? Rut Depth Transverse Cracking, new flexible Alligator Area Cracking, new flexible Reflection Cracking Smoothness

Questions Specific to Bottom-Up Alligator Cracking 1. How or does the higher magnitudes of measured alligator cracking affect the predicted growth of alligator cracking with damage or time? 2. Are the flexible LTPP test sections representative of alligator cracking over time? 3. Could some of the LTPP sections exhibiting high amounts of cracking be top-down cracking? 4. The impact of total AC thickness on alligator cracking seems low. Is that a correct observation?

Bottom-Up Alligator Area Cracking Iterative Process: I. Alligator cracking for new AC layers in the intermediate thickness range. • Assumption: AC fracture model coefficients are the same for all dense-graded mixtures. II. Alligator cracking for new AC layers for the thin (< 6 inches) and thick (> 14 inches) thickness range. III. Combine all new AC flexible pavements. IV. Verify AC overlays with new model coefficients with reflection cracking.

Bottom-Up Alligator Area Cracking Examples of AC thickness on fatigue cracking from SPS-1 Projects. Data exclude all SPS-1 sections with a PATB layer.

Bottom-Up Alligator Area Cracking Examples of AC thickness on fatigue cracking from SPS-1 Projects.

Bottom-Up Alligator Area Cracking Examples of AC thickness on fatigue cracking from SPS-1 Projects. Observation: The measurement error for alligator cracking is large. There is no smooth transition between alligator cracking and AC layer thickness for just about any SPS-1 site. Question: Do the ATB mixtures have the same fracture strength coefficients as for the other AC mixtures?

Bottom-Up Alligator Area Cracking Examples of AC thickness on fatigue cracking from SPS-1 Projects.

Bottom-Up Alligator Area Cracking Examples of AC thickness on fatigue cracking from SPS-1 Projects. Observation: Assuming one set of fracture coefficients for all AC mixtures, the SPS-1 data of cracking for different AC thicknesses along the same project, compresses the thickness impact on fatigue cracking. Emphases the importance of fracture strength testing!

Bottom-Up Alligator Area Cracking LTPP study found that C 2 and Bf 1 were dependent on AC layer thickness. Thick AC layers (> 12 inches) –may exhibit topdown cracking. Thin AC layers (< 5 inches) – impact of crushed stone bases becomes much more important.

Bottom-Up Alligator Cracking

Bottom-Up Alligator Cracking Intermediate AC thicknesses are close together. Does this make sense? Initially – No, but remember C 2 and βf 1 are thickness dependent.

Bottom-Up Alligator Area Cracking Survivability analysis from LTPP data: International Conference on Perpetual Pavements, 2006. 12 to 14 inches of AC Range of less change in survivability

Bottom-Up Alligator Area Cracking Applied Strain, µstrain Fatigue Strength: Flexural Fatigue Test 100 xz Endurance Limit 10 1. 0 E+02 1. 0 E+04 1. 0 E+06 1. 0 E+08 1. 0 E+10 Cycles to Failure

Bottom-Up Alligator Cracking Bending beam tests unlikely to be performed other than for research purposes. Need a simpler and quicker fracture test to estimate or measure the fatigue strength for significantly different mixtures.

Bottom-Up Alligator Cracks Final results from the global calibration of K values, as compared to earlier versions. Parameter AC Thick. Version 2. 3. 1 Version 2. 5; Lab. Derived Values Kf 1 Intercept --- 0. 007566 3. 75 --- 1. 281 1. 46 Kf 3 Strain Exponent --- 3. 95 2. 87 Kf 1 E Exponent

Bottom-Up Alligator Cracks Final results from the global calibration of β values, as compared to earlier versions. Parameter AC Thick. Version 2. 3. 1 Version 2. 5; Field-Derived Values βf 1 Intercept <5 1. 0 0. 02054 5 to 12 > 12 0. 001032 βf 1 E Exponent --- 1. 0 0. 88 βf 3 Strain Exponent --- 1. 0 1. 38

Bottom-Up Alligator Cracks Final results from the global calibration of C values, as compared to earlier versions. Parameter AC Thick. Version 2. 3. 1 Version 2. 5; Field-Derived Values C 2 <5 1. 0 2. 1585 5 to 12 > 12 C 1 --- 3. 9666 1. 0 1. 31

Bottom-Up Alligator Cracking What is the impact of the transfer function C 1 and C 2 parameters and the fatigue strength βf 1*Kf 1 product? Increasing C 2, steepens slope. Increasing Kf 1, βf 1 decreases cracking. Increasing C 1, extends curve out.

Bottom-Up Alligator Cracking

Bottom-Up Alligator Cracking Statistical Parameters Original Values Version 2. 5 Values n, Observations R 2 Se Se/Sy 461 0. 275 5. 01 0. 815 659 0. 41 6. 1 0. 76 N, Projects 101

Bottom-Up Alligator Cracking Comparison of predicted alligator cracks using versions 2. 3. 1 and 2. 5. In summary: use of version 2. 5 model coefficients results in higher predicted area of alligator cracking for larger design criteria on higher volume roadways.

Questions Specific to Bottom-Up Alligator Cracking 1. 2. 3. 4. How or does the higher magnitudes of measured alligator cracking affect the predicted growth of alligator cracking with damage or time? C 2 increases or is much larger. Are the flexible LTPP test sections representative of alligator cracking over time? We believe so; or no reason to suggest they are not representative. Could some of the LTPP sections exhibiting high amounts of cracking be top-down cracking? Definitely yes. The impact of total AC thickness on alligator cracking seems low or AC thickness has little impact on alligator cracking. Is that a correct observation? Increasing AC thickness in the intermediate range (6 to 10 inches) has a lower impact on predicted cracking in comparison to version 2. 3. 1.

Bottom-Up Alligator Cracks Transfer function coefficients Dense-graded neat AC mixtures. Field adjustment coefficients Lab derived model coefficients

Recalibration: Flexible and Semi-Rigid Pavements I. III. IV. V. VI. Why Recalibrate? Rut Depth Transverse Cracking, new flexible Alligator Area Cracking, new flexible Reflection Cracking Smoothness

Reflection Cracking LTE is important; none of the LTPP flexible and semi-rigid pavement sites included LTE measurements prior to overlay. Crack severity level used to estimate LTE. No Change in reflection cracking model coefficient for fatigue and transverse reflection cracks.

Reflection Cracking; AC over AC

Reflection Cracking; AC over PCC

Reflection Cracking; Semi-Rigid Pavements

CTB Fatigue Cracking Final results from the global calibration of CTB fatigue after reflection cracking was completed. Type of Mixture Coefficient Version 2. 5 K 1 0. 972 K 2 CTB Layers; defined by compressive strength and elastic modulus β 1 0. 0825 1. 0 β 3 1. 0 C 1 0 C 2 75 C 3 2. 0 C 4 2. 0

Recalibration: Flexible and Semi-Rigid Pavements I. III. IV. V. VI. Why Recalibrate? Rut Depth Transverse Cracking, new flexible Alligator Area Cracking, new flexible Reflection Cracking Smoothness

Smoothness, IRI No change in global model calibration coefficients of the regression equations.

Smoothness, IRI Statistical Parameters Original Values Version 2. 5 Values n, Observations R 2 Se Se/Sy 1926 0. 56 18. 9 in. /mi. 0. 775 916 0. 66 11. 1 0. 59 N, Projects 126

Smoothness, IRI

FY 2018 Enhancements Released with Version 2. 5 Poll 3: Questions 6 and 7

AASHTOWare: Pavement ME Design FY 2018 Enhancement QUESTION AND ANSWER SESSION We welcome comments & suggestions for future webinars; Send an email to pavementmedesign@ara. com.

FY 2018 Enhancements Released with Version 2. 5 Remember: Pavement ME Design Users Group Meeting scheduled for November 7 and 8, 2018 in Nashville, TN. An overview/demo of the Calibrator will be included as a topic.

FY 2018 Enhancements Released with Version 2. 5 Remember to visit https: //www. allanswered. com/community/s/pavem ent-engineering/ to ask questions and participate in the Pavement ME design community.

FY 2018 Enhancements Released with Version 2. 5 For FY 2019: Develop a tool to assist agencies in local calibration – The Calibrator. Continue to prepare a web version of the software – high priority of AASHTO.

Thank you for Attending the Webinar! AASHTOWare Pavement ME-Design Contacts: • Vicki Schofield, AASHTO vschofield@aashto. org Phone: (202) 624 -3640 • John Donahue, Mo. DOT john. donahue@modot. gov ME Design Resource Website http: //www. me-design. com Pavement ME Design Users Group Contact: • Christopher Wagner, FHWA Christopher. wagner@dot. gov Phone: (404) 562 -3693 Help Desk, Customer Support: PREFERRED • Pavement ME Design Help Desk pavementmedesign@ara. com Phone: (217) 356 -4500 • Other ARA Staff: • Chad Becker cbecker@ara. com • Wouter Brink, wbrink@ara. com • Harold Von Quintus, P. E. hvonquintus@ara. com Phone: (217) 356 -4500