Influence of Recycled Asphalt Shingles (RAS) on Fatigue Properties of Asphalt Pavements Andrew Cascione & Dr. Christopher Williams
Introduction 8-11 million tons of asphalt shingles are placed in landfills each year More state agencies are allowing RAS as liquid asphalt prices continue a long-term climbing trend State agencies have little information on the long term fatigue performance of pavements with RAS
Pooled Fund Study TPF-5(213) Started in 2009 Collaboration of 8 state agencies Overall goals include: Developing RAS processing specifications Conducting field demonstration projects Evaluating field and laboratory performance of pavements with RAS Lab performance testing to include: Four Point Beam Fatigue Testing
TPF-5(213) Partner States Missouri (Lead State) Iowa Minnesota Indiana Illinois Colorado Wisconsin California
ResearchObjectives Agency Research Interest Mix Design Experimental Plan Iowa» Percentage of RAS 0% RAS 4% RAS 5% RAS 6% RAS Minnesota» Post-Manufacturer RAS vs. Post-Consumer RAS 0%RAS 30% RAP 5% Mfr RAS 5% Post- Cons. RAS Missouri» Coarse vs. Fine Grind RAS 0% RAS 15% RAP 5%FineRAS 10% RAP 5%CrseRAS 10% RAP Indiana» RAS with Foaming WMA 15%RAP HMA 3%RAS HMA 3%RAS WMA
RAS Properties (Gradations are prior to extractions) Iowa Minnesota Minnesota Missouri Missouri Indiana RAS Source Post-Cons. Manuf. Post-Cons. Post-Cons. Post-Cons. Post-Cons. 3/4" 100 100 100 100 100 100 1/2" 97 100 100 100 98 100 3/8" 95 95 99 99 94 97 #4 84 70 85 82 75 74 #8 67 56 73 67 62 62 #16 44 32 49 43 42 38 #30 22 12 24 21 22 18 #50 10 4 10 12 12 9 #100 3 1 3 5 5 4 #200 0.6 0.4 0.5 0.9 1.2 0.7 % AC 21.1% 14.6% 20.5% 25.0% 21.7% 21.1% High PG 124.1 109.1 122.5 137.3 146.1 NA
Testing Methodology 4 Point Beam Fatigue from IPC Global Beam samples contained 7% air voids 20 C testing temperature Followed AASHTO T-321 Strain Controlled 375, 450, 525, 650, 800, 1000 μstrain N f = N f = cycles to failure ε o = flexural strain K1 2 1 K εo K1 = regression constant K2 = regression constant
Fatigue Endurance Limit NCHRP Report 646 method Estimated endurance limit as strain level at which sample can withstand 50 million load cycles Calculated lower 95% prediction limit at 50 million load cycles from regression analysis Lower Prediction Limit = yˆ 1 ( x x ) 2 t s 1 o + + o α n S xx y o = the one-sided lower 95% prediction interval at the micro-strain level corresponding to 50,000,000 cycles t α = value of t distribution for n-2 degrees of freedom for a significance level of 0.05 s = standard error of the regression analysis n= number of samples S xx = sum of squares of the x values x o = log 50,000,000 = average of the fatigue life results x
Iowa DemonstrationProject July 2010 State Highway 10 2 surface coarse in 4 overlay 30k Tons 12.5 mm NMAS PG 58-28
Iowa DOT Mix Properties % RAS 0 4 5 6 % TotalAsphalt Content % Binder Replacement 5.5 5.4 5.5 5.4 0 15.1 17.5 19.8 Low PG -19.7-19.1-16.7-14.1 High PG 73 75.8 81.3 86.1
Iowa DOT Fatigue Results 1000 Strain (10-6 in./in.) 4% RAS 5% RAS 6% RAS 0% RAS Mix ID 200 1E+2 1E+3 1E+4 Nf 1E+5 1E+6 1E+7 Average Initial Stiffness (MPa) K1 K2 R 2 Endurance Limit (μstrain) 0% RAS 3497 1.43E-13-5.45 0.987 144 4% RAS 3090 6.75E-14-5.68 0.987 182 5% RAS 3106 1.97E-12-5.27 0.982 175 6% RAS 3156 7.07E-14-5.65 0.967 162
Minnesota 2008 Reconstruction of MnRoads on I-94 Cells 5, 6, 13-23 Placed in shoulders 3 lift 12.5mm NMAS PG 58-28
MinnesotaDOT Mix Properties 30% RAP 5% RAS Post-Manuf. 5% RAS Post-Consumer % TotalAsphalt Content % Binder Replacement 5.3 4.9 5.0 30 14.9 20.5 Low PG -22.4-21.7-21.6 High PG 68.8 71.3 71.1
Minnesota DOT Fatigue Results 1000 Strain (10-6 in./in.) 5% Mfr RAS 5% TORAS 30% RAP 200 1E+2 1E+3 1E+4 1E+5 Cycle to Failure (Nf) 1E+6 1E+7 Mix ID Average Initial Stiffness (MPa) K1 K2 R 2 Endurance Limit (μstrain) 30% RAP 3350 9.19E-12-4.90 0.994 131 5% Post-Manufacturer RAS 2272 2.22E-09-4.19 0.996 123 5% Post-Consumer RAS 2792 6.66E-11-4.51 0.982 89
Missouri US Route 65 south of Springfield June 2010 1.75 surface coarse in 4 overlay 26k Tons 12.5 mm NMAS PG 64-28 w/ 10%GTR
MissouriDOT Mix Properties 15% RAP 5% RAS Fine Grind 5% RAS Coarse Grind % TotalAsphalt Content % Binder Replacement 5.7 6.2 6.2 19.1 30.2 29.2 Low PG -16.8-8.4-3.2 High PG 75.0 90.1 88.3
Missouri DOT Fatigue Results 1000 Strain (10-6 in./in.) 5% Fine RAS, 10% RAP 5% Coarse RAS, 10% RAP 15% RAP 200 1E+2 1E+3 1E+4 Nf 1E+5 1E+6 1E+7 Mix ID Average Initial Stiffness (MPa) K1 K2 R 2 Endurance Limit (μstrain) 0% RAS 15% RAP 6341 5.15E-17-6.40 0.968 139 5% Fine RAS 10% RAP 5096 7.25E-19-6.91 0.992 145 5% Coarse RAS 10% RAP 4902 2.07E-20-7.37 0.968 159
Indiana US 6 in north central Indiana July 2009 1.5 lift over 8 of HMA and 8 of concrete 9.5mm PG 70-22
IndianaDOT Mix Properties 15% RAP HMA 3% RAS HMA 3% RAS WMA % TotalAsphalt Content % Binder Replacement 5.7 6.2 6.2 18.0 12.6 12.6 Low PG -19.8-15.3-15.7 High PG 75.6 77.6 78.8
Indiana DOT Fatigue Results 1000 Strain (10-6 in./in.) 3% RAS - HMA 3% RAS - WMA 15% RAP 200 1E+2 1E+3 1E+4 Nf 1E+5 1E+6 1E+7 Mix ID Average Initial Stiffness (MPa) K1 K2 R 2 Endurance Limit (μstrain) 15% RAP - HMA 5302 7.04E-12-4.87 0.993 114 3% RAS - HMA 5447 1.41E-11-4.77 0.970 118 3% RAS - WMA 5431 1.17E-11-4.81 0.985 110
Conclusions RAS can increase the ability of HMA to resist fatigue damage in strain-controlled loading RAS mixes can exhibit similar fatigue properties as commonly used DOT mixes that use RAP Post-manufacturer RAS mixes performed better than post-consumer RAS mixes in a strain-controlled testing Replacing RAS with RAP (up to 11% increase in % binder replacement) did not reduce the FEL Mixes with fine RAS had better fatigue performance than mixes with coarse RAS at higher strain levels Using foaming WMA technology with RAS did not change fatigue properties of the asphalt mixture
Thank You! Any Comments or Questions?