Stress Sweep Rutting (SSR) Test: AMPT

Transcription

1 Stress Sweep Rutting (SSR) Test: AMPT Amir Ghanbari Graduate Research Assistant Y. Richard Kim Jimmy D. Clark Distinguished University Professor Alumni Distinguished Graduate Professor B. Shane Underwood Associate Professor Department of Civil, Construction, and Environmental Engineering North Carolina State University Asphalt Mixture and Construction ETG Meeting, Fall River, MA May 8, 2018

2 Topics Overview: the basics of the SSR experiment Research: development and verification Practical: draft standard specification 2

3 What is the Stress Sweep Rutting (SSR) Experiment? AMPT compatible test for rutting performance RLPD and TRLPD (Flow number), Creep test (Flow time), irlpd Test parameters Specimen: 100 mm dia. x 150 mm tall specimen cut and cored from 150 mm dia. x 180 mm tall gyratory sample Test temperature: T H and T L Loading: axial compression cyclic test under 10 psi confining pressure (UTS044 in Controls AMPT) Loading time = 0.4 seconds Rest period = 3.6 seconds for T H and 1.6 seconds for T L 3 Deviator stress = 100, 70, and 130 psi for T H and 70, 100, and 130 psi for T L Measurements taken = applied load and actuator displacement Two samples for each temperature for a total of four samples in one day

4 4 SSR Experiment Setup

5 SSR Experiment Stress Pulses T H or T L Confining Pressure 10 psi 5 Controls AMPT: UTS044 Temperature Control

6 6 SSR Experiment Outcome

7 Why Stress Sweep Rutting (SSR) Test? SSR Test in AMPT Structure Traffic Climate Rutting Prediction 7

8 Topics Overview: the basics of the SSR experiment Research: development and verification Practical: draft standard specification 8

9 9 Where did this test come from? Origins in NCHRP 9-19 and DTFH61-03-H where the effect of repeated pulses of load, wave shape, temperature, rest period, etc. were examined More directly 1. Multiple TRLPD tests at different load durations, load levels, and temperatures Established a model form 2. Triaxial stress sweep test (TSS) using repeated stress pulses of varying magnitude at multiple temperatures and one TRLPD test at a reference condition Simplified testing 3. Stress sweep rutting (SSR) using repeated stress pulses of varying magnitude, pulse time, and temperatures Further simplified testing

10 Model Foundation and Experimental Model basis: Evolution The equivalence of loading frequency and temperature in permanent strain accumulation. The equivalence of stress level and number of loading repetitions in permanent strain accumulation. 10

11 Equivalence of Loading Frequency and Temperature Linear Viscoelasticity Time (loading frequency) and temperature are interchangeable. Create a continuous mastercurve by horizontally shifting data from different temperatures E* (MPa) E-08 1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 1.E+08 Reduced Frequency (Hz) C 10.6 C 35.0 C 53.5 C Shift Factor, a Mastercurve T Shift Factor y = x x R 2 = Temperature( C) 35.0 Shift function 54.0

12 Equivalence of Loading Frequency and Temperature with Damage Time-temp. shift factor from E* test y = x x R 2 = σ t/a T Repeated creep and recovery test at 40º and 55ºC Shift Factor Shift function % -6 Temperature( C) Viscoplastic Strain 1.5% 1.0% 0.5% 140kPa-827kPa 55 C VT (1) 140kPa-827kPa 55 C VT (2) 140kPa-827kPa 40 C VT (1) 140kPa-827kPa 40 C VT (2) 0.0% Cumulative Loading Time (sec)

13 Equivalence of Loading Frequency and Temperature with Plastic Strains Experiments = Multiple TRLPD at 90 psi deviatoric but with varying loading times and temperatures. FHWA ALF Control 90 psi After Shifting Reference: 0.1s-54 C Permanent Strain (%): FHWA 2.0% 1.6% 1.2% 0.8% 0.4% 1.6s-54C 0.4s-54C 0.1s-54C 1.6s-40C 0.1s-40C Permanent Strain (%): FHWA 2.0% 1.6% 1.2% 0.8% 0.4% Horizontal shift distance is termed the loading time reduced load time shift factor, a ξp 0.0% 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 Physical Cycles (N) 0.0% 1.E-02 1.E-01 1.E+00 1.E+01 1.E+02 1.E+03 1.E+04 Effective Cycles (N) 13

14 Equivalence of Stress Level and Number of Loading Repetitions Experiments = Multiple TRLPD at varying deviatoric stresses and with varying loading times and temperatures. After Load Time Shifting Mastercurve at 90 psi Permanent Strain (%): FHWA 2.0% 1.6% 1.2% 0.8% 0.4% 90 psi 150 psi 120 psi Permanent Strain (%): FHWA 2.0% 1.6% 1.2% 0.8% 0.4% Vertical shift distance is termed the vertical stress shift factor, a σv 0.0% 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 Effective Cycles (N) 0.0% 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06 Effective 14 Cycles (N)

15 Shift Model for Rutting ε vp = ε N 0 red ( N + N ) I red β Describes permanent strain evolution at the shifted, or reference condition Nred ξ p = N a a σv ( ξ p ) a = p log + p ξ p 1 2 ( log ( σ ) 0.877) a = D P σ v v a D = d1 T + d2 Explains how cycle wise repetitions at a given loading time and temperature are equivalent to those repetitions at the reference condition N red p ( ξ p D p σ v = N ) 1 Pa D 15

16 Shift Model for Rutting 16 N red ( ξ p ) ε vp a = p log + p ξ p Nred 1 2 = ε N 0 red ( N + N ) I red ξ p = N p ( ξ p D p σ v = N ) 1 Pa a β a σv ( log ( σ ) 0.877) a = D P σ v v a D = d1 T + d2 D With Multiple TRLPD tests shifting and fitting of the results yields the variables circled here. Problem: Testing is extensive as TRLPD must be run in its entirety for multiple temperatures, pulse times, and stress levels.

17 Triaxial Stress Sweep (TSS) Test Reference Test (T H ) 0.4s x 600 cycles or more 2.5% Permanent Strain (%) 2.0% 1.5% 1.0% 0.5% 0.0% Cycles (N) 100 psi sec (1) Stress Sweep Test at T H, T I, T L 0.4s x 200 cycles for each loading block Permanent Strain 1.60% 1.20% 0.80% 0.40% NY9.5B-47C NY9.5B-37C NY9.5B-17C 0.00% Cycles (N)

18 Shift Model for Rutting ε vp = ε N 0 red ( N + N ) I red β Characterized using reference test at T H ( ξ p ) a = p log + p ξ p Nred 1 2 ξ p = N a a σv ( log ( σ ) 0.877) a = D P σ v v a D = d1 T + d2 Shift Factors N red p ( ξ p D p σ v = N ) 1 Pa 1 D Load Time SF (aξp) (d) E-04 1.E-03 1.E-02 1.E-01 1.E+00 Reduced Load Time (sec) Stress SF (aσd) (e) Separated 18 by Stress Level and Temperature Vertical Stress (psi)

19 Model Verification Loading history Random Loading Test Permanent Strain: Random 2.5% 2.0% 1.5% 1.0% 0.5% 0.0% Predicted Measured Deviator (psi) Cycles (N) Testing Time (sec) Measured Predicted 19

20 Shift Model in FlexPAVE TM T 1, t p,1, σ v,1 T i, t p,i, σ v,i T N, t p,n, σ v,n ε vp = ( N + N ( T, ξ, σ )) RD ε0 Nred ( T, ξp, σv ) I red p v N = ε vp, i i= 1 h i β 20

21 21 Rutting Prediction by TSS/FlexPAVE TM

22 Simplification of TSS to SSR Test Method TSS Reference 1 (T H ) TSS Temp. 3 (T H, T I, and T L ) Pulse Time (s) Rest Period (s) Deviator Stress (psi) Number of Samples Displacement Measurement Testing Time incl. Pre-conditioning 10 (T H ), 1.6 (T I ) 1.6 (T L ) 70, 100, 130 (T H and T L ) 8 (2 replicates of each experiment) On-specimen LVDTs 16 hrs for 8 TSS experiments

23 Simplification of TSS to SSR Test Method TSS SSR Reference 1 (T H ) - TSS Temp. 3 (T H, T I, and T L ) 2 (T H and T L ) Pulse Time (s) Rest Period (s) 10 (T H ), 1.6 (T I ) 1.6 (T L ) 3.6 (T H ) 1.6 (T L ) Deviator Stress (psi) 70, 100, 130 (T H and T L ) 100, 70, 130 (T H ) 70, 100, 130 (T L ) SSR Number of Samples 8 (2 replicates of each experiment) 4 (2 replicates of each experiment) Displacement Measurement On-specimen LVDTs Actuator 23 Testing Time incl. Pre-conditioning 16 hrs for 8 TSS experiments 6 hrs for 4 SSR experiments

24 Shift Factors between TSS and SSR Total SF-TSS Total SF-SSR Reduced Load Time SF Vertical Stress SF 24

25 Effects of Simplifications on Rut Depths Predicted by FlexPAVE TM 4 in. RS9.5B 10 in. Aggregate Base 700 AADTT 40 kn wheel load Raleigh, NC 6.0 Actuator Displacement vs. LVDT 10 Temperature, Reversed Loading Block, and Rest Periods Rut Depth (cm) Rut Depth Comparison (RS9.5B) (1) LVDTs (2) X head (3) Corrected X head Total Rut Depth (mm) TSS S-TSS S-TSS S-TSS Time (Month) Time (Months)

26 Draft AASHTO SSR Specification 26

27 Test Parameter Overview of SSR Quantity Equipment AMPT Temperature 2 (T H and T L ) Pulse Time (s) 0.4 Rest Period (s) 3.6 (T H ) or 1.6 (T L ) Deviator Stress (psi) 100, 70, 130 (T H ) or 70, 100, 130 (T L ) Displacement Measurement Actuator Test Outputs ε 0, N I, β, d 1, d 2, p 1, and p 2 Replicates 2 27 Testing Time incl. Pre-conditioning 6 hrs for 4 SSR tests

28 Test Temperature Update TT HH = 0.87( DDDD 15 log(hh + 45)) T H = high test temperature, C, DD = Degree-Days >10 C ( 1000) from LTPPBind v 3.1, and H = depth of layer, mm (0 for surface layer). T L T + T HPG,98% LPG,50% = T L = low test temperature, C, and T HPG,98% = continuous high temperature grade for 98% reliability T LPG,50% = continuous low temperature grade for 50% reliability 28

29 Draft AASHTO SSR Specification Updates from first ETG review Wording modified to be consistent with AASHTO T378 Maximum aggregate size, High vacuum grease, Balances requirements, Dummy specimen preparation, Latex membrane preparation moved to Annex A, Calibration section, Air void content Samples must be compacted to 150 mm diameter x 180 mm tall height with the gyratory compactor, and then test specimens cored and cut from these. Calculation section completely rewritten so that it is easier to follow and conduct the analysis. 29

30 Section 10 Updates 30 N red ( ξ p ) ε vp a = p log + p ξ p Nred 1 2 = ε N 0 red ( N + N ) I red ξ p = N p ( ξ p D p σ v = N ) 1 Pa a β a σv ( log ( σ ) 0.877) a = D P σ v v a D = d1 T + d2 D Section 10 describes the calculation process to obtain coefficients, ε 0, β, N I, d 1, d 2, p 1, and p 2. FlexMat TM Rutting mentioned With these quantities a user can run FlexPave TM to predict the rutting of a pavement. Annex B also includes a method to use the model to predict permanent strain as possible index parameter.

31 Section 10: Calculations ε vp = ε N 0 ref ( NI + Nref ) Finding ԑ 0, β, and N I from Reference Curve β 1. Construct the reference curve (Sec. 10.3) Using optimization based on the 100 psi loading block at T H 2. Compute the total shift factor (Sec. 10.4) Nref = N a 10 total Reference from step 1 SSR at T H or T L Shift Factors ref atotal log N =

32 Section 10: Calculations 32 ε vp Nref total = ε N 0 ref ( NI + Nref ) Finding ԑ 0, β, and N I from Reference Curve = N p β a 10 total a = a + a N ref ξ p1 σ ξ p σ v = A N 1 Pa p D A = D= dt 1 + d2 v D 3. Compute the reduced time shift factor and vertical stress shift factor (Sec & 10.6) Load Time SF (a ξp ) Total SF (a total ) (d) Reduced load time SF (c) Total SF (100 psi) E E E E E E E+01 Reduced Load Time (Sec) -2 T L 70 psi 100 psi 130 psi E E E E E E E+01 Reduced Load Time (Sec) a = p log( ξ ) + p ξ p 1 p 2 Stress SF (a σd ) T H (e) Vertical Stress SF a = DT ( )*(log ( σ / P) 0.88) σ v v a D= dt+ d Vertical Stress (=dev+con)

33 FlexMAT TM -Rutting Determines shift model coefficients from AMPT data files and generates input files for FlexPAVE TM 33 Note: Old screenshot: Updated FlexMat TM Rutting sheet that is 508 compliant is currently being developed

34 Export Data to FlexPAVE TM 34

35 Material Properties Input in FlexPAVE TM 35

36 Stress Sweep Rutting (SSR) Test SSR Test in AMPT Structure Traffic Climate Rutting Prediction 36

37 37 Thank you