Asphalt Pavement Response and Fatigue Performance Prediction Using. Warm Mix Asphalt

Transcription

1 Asphalt Pavement Response and Fatigue Performance Prediction Using the VECD Approach ApplicationApplication to Warm Mix Asphalt Y. Richard Kim, Ph.D, P.E. North Carolina State University Presented at the International Workshop on Cold and Warm Asphalt Mixture Design/Characterization and Pavement Design Identification of Worldwide Best Practices Fortaleza, Brazil October 5, 29

2 Outline FHWA PRS Project VEPCD Model Characterization Verification Application NCHRP 1-42A Integrated t VECD-FEP++

3 FHWA HMA PRS Project Four year long project started in Feb. 27 Objectives To develop various tools for testing ti and analysis of HMA mixture To develop a hierarchical system for performance-related specification The original research plan includes a wide range of HMA mixtures from various pavement sections. Recently incorporated RAP and WMA mixtures from NCAT Test Track and Manitoba projects

4 Summary of PRS Pavements FHWA ALF Pavements (control and modified) NY I-86 Perpetual Pavements NCAT RAP and WMA Pavements Control, OGFC w/15% RAP, High RAP (5%), High RAP plus WMA (Evotherm and Advera) Manitoba RAP and WMA Pavements WMA (Sasobit, Advera, Evotherm) RAP (, 15, 5%) Chinese Perpetual Pavements in Binzhou, Shandong KEC Test Road Pavements

5 Proposed Hierarchical PRS Model Description Level 1 Level 2 Level 3 Unconfined and IR E* and 55 C E* AMPT E* Confined E* Predictive Equation Predictive Equation for HMA Model Cracking (Tension) Rutting (Compression) Uniaxial VEPCD MVEPCD Uniaxial VEPCD VP at a Representative Confining Pressure VEPCD Coefficients from Mix Characteristics Predictive Equation for VP Coefficients from Mix Characteristics Pavement Model MVEPCD-FEP++ Layered Viscoelastic Layered Viscoelastic Model Model Testing Time 17 days 5 days Less than 1 day Analysis Time 3 days 2 days Less than 1 day Total Time 2 days 7 days 1 day

6 VEPCD Model Viscoelastoplastic Continuum Damage (VEPCD) Model Time- Temperature Superposition with Growing Damage Elastic- Viscoelastic Correspondence Principle Work Potential Theory Viscoplastic Model Linear Viscoelastic Effects Microcracking Related Degradation Permanent Deformation Growth Time- Temperature Effects

7 VEPCD Modeling Approach Monotonic Tests at 5 C or Cyclic Tests at 19 C Elastic LVE Damage R l ti hi (Microcracking) Damage Characteristic Relationship Strain Hardening Plastic VP Model VP ε = ε + total ve ε vp Linear Viscoelastic Creep Compliance VECD Model Dynamic Modulus Test and Interconversion Viscoplastic Coefficients Monotonic Tests at 4 C

8 VECD Experimental Program Dynamic modulus (LVE Characterization) -1, 5, 2, 4 and 54 C 25, 1, 5, 1,.5 and.1 Hz 5 75 microstrain peak-to-peak strain amplitude Tension-compression protocol Monotonic at 19 C or controlled crosshead cyclic Monotonic at 19 C or controlled crosshead cyclic at 19 C and 1 Hz (Damage Characterization)

9 Linear Viscoelastic i Behavior 3.E E* (MPa) 2E+4 2.E+4 1.E+4 Phase Angle (d deg) E+ 1.E-5 1.E-2 1.E+1 1.E+4 1.E-5 1.E-2 1.E+1 1.E+4 Reduced Frequency (Hz) Reduced Frequency (Hz) 1.E From Measure Fitted E* (MPa) 1.E+4 1.E+3 log at E+2 1.E-5 1.E-2 1.E+1 1.E+4 Reduced Frequency (Hz) Temperature (C)

10 C* Damage Characteristic Curve Cyclic and Monotonic 5C me 5C - 15 me 19C - 19 me 19C - 24 me 27C - 52 me 27C -668 me Monotonic.2..E+ 2.E+5 4.E+5 6.E+5 8.E+5 S

11 Study Mixtures FHWA ALF pooled fund study TPF-5(19) Four mixtures each the same coarse 12.5 mm gradation with the same asphalt content (5.3%) Unmodified PG 7-22 (Control) Crumb Rubber Terminal Blend (CRTB, PG 76-28) Styrene Butadiene Styrene (SBS, PG 7-28) Ethylene Terpolymer (Terpolymer, PG 7-28)

12 ALF Mixtures Comparison LVE Characteristics E* * (MPa) 4 Control 1 3 CRTB SBS Terpolymer E-8 1.E-5 1.E-2 1.E+1 1.E+4 Reduced Frequency (Hz) 1 1.E-8 1.E-5 1.E-2 1.E+1 1.E+4 Reduced Frequency (Hz)

13 Simplified Formulation Verification i CX-L 19-CX-L (2) 19-CX-H 5-CX-L 5-CX-H(2) 5CXH 5-CX-H 5CXH 19-CS-L 19-CS-H 5-CS-H Monotonic CX-L 19-CX-H 19-CS-L 19-CS-H 5-CS-H Monotonic C* C* Control..E+ 2.E+5 4.E+5 6.E+5 8.E+5 1.E+6 S.2 CRTB..E+ 1.E+5 2.E+5 3.E+5 4.E+5 5.E+5 S C* CX-L 19-CX-H 19-CS-L 19-CS-H 5-CS-H Monotonic C* CX-L 19-CX-H 19-CS-L 19-CS-H 5-CS-H Monotonic SBS.E+ 1.E+5 2.E+5 3.E+5 4.E+5 5.E+5 6.E+5 7.E+5 S 2.2. Terpolymer.E+ 1.E+5 2.E+5 3.E+5 4.E+5 5.E+5 S

14 VECD Comparison of ALF Mixtures Damage Characteristics Control SBS Terpolymer CRTB C* E+ 2.E+5 4.E+5 6.E+5 8.E+5 1.E+6 S

15 VEPCD Model Verification

16 Random Loading Validation Random Stress level el and frequency, kpa, 1 2 Hz, 25 C C 7 6 Stress s (kpa) tress (kpa) St Time (s) Loading Block Num mber of Cycles Loading Block Fre equency (Hz) Loading Block

17 Random Loading Verification eiiaio 1.5E-2 8.E-3 Control CRTB Measured Strain 1.E-2 Measured Predicted Total Strain 6E3 6.E-3 4.E-3 Predicted Total 5.E-3 3 Predicted VP Predicted VP 2.E-3.E+ Predicted VE.E+ Predicted VE E-2 SBS Time (s) Measured Predicted Total 6.3E-3 Terpolymer Time (s) Measured Predicted Total 8E3 8.E-3 Strain Strain 4.2E-3 4.E-3 Predicted VP 2.1E-3 Predicted VP.E+ Predicted VE.E+ Predicted VE Time (s) Time (s)

18 MEPDG Fatigue Model 1 5C Measured Strain Amplitu ude 19C Measured 27C Measured 1 1.E+1 1.E+3 1.E+5 1.E+7 Nf (Cycle)

19 Fatigue Life Prediction Using Monotonic Data C* C* CX-VL-Measured.2 19-CX-VL-Measured 19-CX-H-Predicted 19-CX-H-Measured.E+ 1.E+6 2.E+6 3.E+6 4.E+6 Reduced Time (s) 4.4 5CXLP 5-CX-L-Predicted d 5-CX-L-Measured.2 5-CX-H-Predicted 5-CX-H-Measured.E+ 5.E+3 1.E+4 1.5E+4 2.E+4 2.5E+4 Reduced Time (s)

20 Failure Criteria 5 4 E* E* Phase Angle 4 3 N f (a) 2 E+.E+ 1E+3 1.E+3 2E+3 2.E+3 3E+3 3.E+3 4E+3 4.E+3 5E+3 5.E+3 Number of Cycles e Agnle Phas

21 VECD Failure Criteria Failur re ALF Control ALF CRTB ALF Terp. Mix A Failure Envelope Temperature (deg. C)

22 Prediction of N f vs. ε t Fatigue Relationship mplitude 1 5C Measured 19C Measured 27C Measured Predicted Strain A 1 1.E+1 1.E+3 1.E+5 1.E+7 Nf (Cycle)

23 Prediction of Fatigue Life Characterized with ihcyclic Predicted Nf 1.E+6 1.E+5 1.E+4 1.E+3 R 2 =.91 Se/Sy =.37 Ei Eric-CX Control-CX CX CRTB-CX Terpolymer-CX Eric-CS Control-CS CRTB-CS SBS-CS Terpolymer-CS LOE 1.E+2 1.E+2 1.E+3 1.E+4 1.E+5 1.E+6 Nf 4.E+5 3.E+5 R 2 =.86 Se/Sy =.37 Measured, N f Predicted 2.E+5 1.E+5 Eric-CX CRTB-CX Eric-CS CRTB-CS Terpolymer-CS Control-CX Terpolymer-CX Control-CS CS SBS-CS LOE.E+.E+ 1.E+5 2.E+5 3.E+5 4.E+5 Measured, N f

24 f Pred dicted N Fatigue Life Verification Multiple mixtures, 1 Hz, 1 7 με, 5, 19, and 27 C C 1.E+6 1.E+5 1.E+4 1.E+3 S9.5C I19C I19B RI19B RB25B S9.5B B25B RS12.5C RI19C LOE 5 ~ 3% Error (in arithmetic space) 1E+2 1.E 1.E+2 1.E+3 1.E+4 1.E+5 1.E+6 Measured Nf

25 Fatigue Endurance Limit Using VECD Model 1E+3 Endurance Limit n Level Strain 1E+2 5C 19C 1E+1 1.E+1 1.E+5 5 million 1.E+9 Nf

26 Effect of Mixture Variables Solid Symbols = Modified Mixes Endura ance Limit (με ) ) Aggregate Size Endura ance Limit (με Asphalt Content NMAS % AC durance Limit (με) En Asphalt Grade durance Limit (με) En RAP Effect No-RAP 2 RAP 3 4 High Temperature PG Grade Use of RAP Materials

27 Thermal Cracking Verification C/hr -8.6C/hr -4.4C/hr 3 Stres ss (kpa) 2 1 LVE VECD VEPCD Actual Time (min)

28 TSRST Prediction Failure Stress (kpa) Filled Symbols: Predicted Clear Symbols: Measured Cooling Rate (C/hr) -2 Failure Time (min) Filled Symbol: Predicted Clear Symbols: Measured Cooling Rate (C/hr) ure at (C) Temperat Failure( Filled Symbols: Predicted Clear Symbols: Measured Cooling Rate (C/hr)

29 Predictions from the VEPCD Model Stress-strain behavior of asphalt mix in: monotonic tests at varying rates of loading and temperature; and random load cyclic tests under varying stress/strain magnitudes, temperatures, and loading frequencies. N f vs. ε t relationship at various temperatures Endurance limit TSRST results under different cooling rates including: thermal stress development history; fracture time, fracture stress, and fracture temperature.

30 VECD Model Application

31 Fatigue Performance Prediction Two-Step Method of HMA Pavement Pavement response model (3-D FEP++) plus VECD model Integrated Method Finite element simulation of damage under continuous loading cycles

32 Two Step Method 3-D FEM Linear VE HMA Elastic Base/SG Pavement Response Model ε kernel VECD Model Transfer Functions Mt Material il Level Performance Model Predicted Distresses

33 Effect of Material Type Vertical Strain Control SBS-Modified

34 ALF Pavement Transverse Strain Response VECD Input Kernel ε f(ξ) 3E-4 2E-4 1E-4 Control CRTB SBS Terpolymer E Time (s)

35 ALF Mixtures Comparison Damage Characteristics Control SBS Terpolymer CRTB C* E+ 2.E+5 4.E+5 6.E+5 8.E+5 1.E+6 S

36 ALF Fatigue Life Prediction log Pre edicted Nf SBS Control CRTB EVY No Terpolymer With Terpolymer y =.7119x R 2 =.8473 y =.6189x R 2 = log Field N f

37 Integrated Method M-E PDG Layered Elastic Model σ, ε Performance Prediction Model Transfer Functions Pavement Response Model Material Level Performance Model Fully Mechanistic 3-D FEM with VEPCD Interface Fracture Healing Aging Nonlinear Base/SG Transfer Functions Predicted Distresses Predicted Distresses

38 Simulation Results (Damage) Only Mechanical Loading Thin Pavement Thick Pavement C:

39 Pavement Simulation for Lime Modified Mix Evaluation cm Pressure kPa Viscoelastic AC Layer 1cm Linear Base 75ksi (51717kPa) 2.3cm Subgrade 1.9ksi (7556kPa)

40 Effect of Moisture Conditioning Cycles Control Control Moisture Lsub Lsub Moisture 5, , , ,, ,5, ,, ,5, ,, C:

41 NCHRP 1 42A VECD FEP++ Aging g Stiffness Damage Viscoplastic Healing E * Monotonic, CS, CX Monotonic Cyclic Pavement Structure Traffic Loading (Non-uniform, Single, Dual) Mesh Size Sensitivity Check Material Characteristics (VECD + Aging) VECD-FEP++ Response Strain, Damage Base Layer (Non-linear Elastic) Subgrade (Linear Elastic) Moisture (EICM) Temperature (EICM) Interpretation Condition Index

42 Effect of Aging on E* 1 AL-STA AL-LTA1 * (MPa) 1 AL-LTA2 LTA2 AL-LTA3 E 1 (b) 1 1.E-8 1.E-5 1.E-2 1.E+1 1.E+4 Reduced Frequency (Hz)

43 Effect of Aging on Phase Angle 45 g) Phase Angle (de AL-STA AL-LTA1 AL-LTA2 AL-LTA3 1.E-8 1.E-5 1.E-2 1.E+1 1.E+4 Reduced Frequency (Hz)

44 Effect of Aging on VECD Model AL-STA AL-LTA1 LTA1 AL-LTA2 AL-LTA3 C E+ 1.E+5 2.E+5 3.E+5 4.E+5 S

45 Healing Model Kim, Lee, and Little (AAPT 1997) Region I Region II iffness (kp Pa) Pseudo St R S B R S C B B C D ΔN f D No. of Cycles

46 2 Modified Healing Model NCHRP 1 42A ΔC C C S 1, ΔS i S 2, ΔS i S 2, ΔS j ξ rest Pseudo Stiffness (k kpa) Lo oading Time Damage, S

47 Simulation io Results eu (Cracking Index) Mechanical and Thermal Loading, Aging, Healing, and Viscoplasticity Thin Pavement Thick Pavement CI:

48 Summary VEPCD model s ability to predict material s behavior at a wide range of conditions Cracking simulation of VECD-FEP++ does not need to know the crack location a priori. Thermal stress, aging, healing, viscoplasticity models implemented into VECD-FEP++ VECD-FEP++ as a tool to investigate and model WMA materials and pavements

49 Thank you!

50 Control, OGFC, Foam WMA, Evotherm WMA NCAT Test Track Cross Section Surface (9.5 mm) Intermediate (19 mm) 1.25 (32 mm) 2.75 (7 mm) NCAT Test Track Base (19 mm) 3. (76 mm) 5% RAP, 5% RAP with Foam WMA Dense Graded Aggregate Base 6. (152 mm) Stiff Subgrade

51 Summary of NCAT Mixtures HMA Mixtures Control, OGFC (15% RAP) High RAP (5%), High RAP + WMA WMA Evotherm (additive), Advera (foam) RAP, 15, 5% RAP Binder Grades Surface/Intermediate layers No RAP PG With High RAP PG Base layer PG 67-22

52 Summary of MIT Mixtures WMA Project WMA Additives Advera, Sasobit, Evotherm Layer Properties Surface layer % RAP, 15/2 pen Intermediate layer 3% RAP, 2/3 pen RAP Project RAP, 15, 5% Binder 15/2 pen for all % RAP, also 2/3 pen for 5% RAP Base Materials (not sampled) 7% RAP

53 km MIT RAP Sections % RAP 15% RAP 5% RAP 5% RAP 15/2 pen 15/2 pen 15/2 pen 2/3 pen 1 mm 7% RAP Base Layer 1 mm m MIT WMA Sections.5 km 3 km 1 km 3 km 1 km 3 km.5 km HMA Advera HMA Sasobit HMA Evotherm HMA +3% RAP +3% RAP +3% RAP +3% RAP +3% RAP +3% RAP +3% RAP 5 mm 5 mm