Implementation of M-E PDG in Kansas

Transcription

1 Implementation of M-E PDG in Kansas Mustaque Hossain, Ph.D.,P.E. Kansas State University 1

2 Projects related to the M-E Guide Implementation and Calibration Kansas HMA Fatigue and Stiffness Study Pool Fund Study with KDOT and NYSDOT APT Pooled Fund Experiment for Flexible Pavement Response Verification 2

3 KS HMA Fatigue and Stiffness Study K-TRAN Project Four HMA base mixes tested for: Dynamic modulus (5 temperature, 5 frequencies and 4 replicates) Third-point beam fatigue (4 strain levels, 4 temperatures, 4 replicates) Final report submitted and published 3

4 KS HMA Characteristics Mix A Mix B Mix C* Mix D Binder Grade Asphalt Content (%) PG64-22 PG64-22 PG70-28* PG * contains SBS polymer B & C had the same aggregate structure 4

5 Fatigue life Mix B 1.E+08 Nf 1.E+07 1.E+06 1.E+05 4 C 10 C 20 C 30 C 1.E Strain (microstrain) 5

6 Measured vs. Predicted Fatigue Life Nf 1.E+10 1.E+09 1.E+08 1.E+07 A-NCHRP A B-NCHRP B C-NCHRP C D-NCHRP D 1.E+06 1.E Strain (microstrain) 6

7 Dynamic Modulus vs. 10 Hz 30,000 Dynamic Modulus (MPa) 25,000 20,000 15,000 10,000 5,000 Mix A Mix B Mix C Mix D 2:1 3: ,500 5,000 7,500 10,000 12,500 15,000 Flexural Stiffness (MPa) 7

8 FHWA Pool Fund Study TPF-5(079) Implementation Of The 2002 AASHTO Design Guide For Pavement Structures Two states are participating: NY and KS Work for Kansas is done; Final report reviewed Work for New York: On-going Modified work plan 8

9 Major Tasks in the NYSDOT Work Plan 1. Collect and Review Literature 2. Establish Pavement Performance and Material Database 3. Analytical Model Calibration DELETED 3. Traffic Data Processing 4. Reporting and Training 9

10 Dynamic Modulus NY Region 3 Mix1 12,000 Dynamic Modulus (MPa) 10,000 8,000 6,000 4,000 2, Hz 5 Hz 1 Hz 0.5 Hz 0.1 Hz Temperature ("C) 10

11 Witczak Equation log E * [ P V a ( log f log ) E* = Asphalt Mix Dynamic Modulus, in 10 5 psi η = Bitumen viscosity in 10 6 poise f = Load frequency in Hz V a = % air voids in the mix, by volume V beff = % effective bitumen content, by volume P34 = % retained on the ¾ inch sieve, P38 = % retained on the 3/8 inch sieve, P4 = % retained on the # 4 sieve, P200 = % passing the # 200 sieve P P P 1 e ( P ( V V beff beff V ) ( P a ) 2 ) P ]

12 Measured vs. Predicted Dynamic Modulus NY Region 5 Mix 1 15,000 Predicted Modulus (MPa) 12,500 10,000 7,500 5,000 2,500 1:1 4C 10C 20C 30C 35C 0 0 2,500 5,000 7,500 10,000 12,500 15,000 Measured Modulus (MPa) 12

13 Dynamic Modulus vs. Temperature KS Mix 9a (VTM=4%) 10Hz 20,000 Modulus (MPa) 15,000 10,000 5,000 DSR Predicted NCHRP Predicted Measured Hirsch Predicted Temperature (C) 13

14 20,000 DSR Predicted vs. Measured Dynamic Modulus - KS Mix 9a Predicted Modulus (MPa) 15,000 10,000 5,000 1:1 1:2 4C 10C 20C 30C 35C 0 0 5,000 10,000 15,000 20,000 Measured Modulus (MPa) 14

15 NCHRP Predicted vs. Measured KS Mix 9b (VTM = 7%) 20,000 Predicted Modulus (MPa) 15,000 10,000 5,000 1:1 1:2 4C 10C 20C 30C 35C 0 0 5,000 10,000 15,000 20,000 Measured Modulus (MPa) 15

16 Findings The ratio between measured and predicted moduli was close to 2 in most cases for Kansas mixes and between 0.8 and 1.2 for New York mixes The mix C (polymerized) had a longer fatigue life than mix B with unmodified binder Higher dynamic modulus at low temperatures and at higher frequencies for all mixes tested For Kansas mixes, the Witczak and Hirsch models gave better predictions for mixes with 7% air voids than for mixes with 4% air voids The moduli predicted by the Hirsch model were the lowest for most cases 16

17 KDOT PCCP Sensitivity Analysis and Comparison Eight in-service PCC projects (includes 3 LTPP SPS-2 sections) All material inputs were derived from actual construction test results Initial analysis was done for AASHTO MEPDG default traffic data Kansas traffic inputs and axle load spectra were derived from KDOT AVC and WIM data analysis 17

18 KDOT PCCP Sensitivity Analysis and Comparison (Contd..) COTE tests were done at FHWA Turner- Fairbanks Research Center Performance prediction results (for IRI and faulting) were compared with Kansas PMS or LTPP data Sensitivity analysis was conducted for each project 18

19 KS PCCP Comparison Project Route County Year Built Slab Thickness (in) AADT % Trucks Initial IRI (in/mile) 2003 (in/mile) K I-70 Geary , K I-70 Shawnee , SPS-2 (S5)* I-70 Dickinson , SPS-2 (S6)** I-70 Dickinson , SPS-2 (Control) I-70 Dickinson , K US-50 Chase , K US-50 Chase , K K-7 Johnson , * 4 Portland Cement -Treated Base and 6 Lime-Treated Subgrade ** 6 Lean Concrete Base and Fly Ash Subgrade 19

20 Findings IRI is the most sensitive to traffic volume With project-specific Kansas traffic input, IRI, Faulting and % Slabs cracked values are lower than those with default MEPDG traffic inputs Percent slabs cracked increases significantly with increasing truck traffic Faulting is the least sensitive output 20

21 APT Pooled Fund Study TPF-5(048) : VERIFICATION OF MECHANISTIC- EMPIRICAL DESIGN MODELS FOR FLEXIBLE PAVEMENTS THROUGH APT TESTING Four states: MO, IA, NE, KS Two HMA mixes tested for each state Two pavements tested for each mix Instrumented pavements Experiment with controlled temperature and loading Three year experiment 21

22 APT Pooled Fund Study 22

23 Wheel Load Assembly 23

24 Pavement Structures 24

25 Fatigue Cracking 10ºC (50ºF) Rutting 40ºC (104ºF) 3.5 in. HMA1 3.5 in. HMA2 7 in. HMA1 7 in. HMA2 6 in. AB-3 stone 6 in. AB-3 stone SUBGRADE SUBGRADE SS South Pit SN MS Middle Pit MN 25

26 Sensor Location for CISL Experiment #14 EAST WEST HAC AC 6 AB-3 Base Subgrade Sensors Thermocouple Longitudinal / Transverse Strain Gage Stress Cell Single Layer Deflectometer 26

27 Current Status Kansas experiment is complete Missouri rutting experiment is also complete Missouri fatigue test is currently going on Iowa sections will be constructed this Spring Project is expected to be complete this year 27

28 Acknowledgement Dr. Stefan Romanoschi, UT-Arlington Kansas Department of Transportation TAC for the MSAPT Pooled Fund Study 28

29 Thank you! 29