Oxidative Aging of Asphalt Binders in Hot g g p Mix Asphalt - PowerPoint PPT Presentation

Oxidative Aging of Asphalt Binders in Hot g g p Mix Asphalt Mixtures Nathan Morian, Ph.D. Candidate, NDOT Elie Y. Hajj, Ph.D., UNR (Presenter) Charles J. Glover, Ph.D., Texas A&M Peter E. Sebaaly, Ph.D., UNR Transportation Research Board 90 th A 90 th Annual Meeting l M ti g Characteristics of Asphalt Materials (AFK20) January 24, 2011

Introduction •Binder aging affects nearly all critical performance Binder aging affects nearly all critical performance aspects of HMA pavements  important to quantify! •Binders aged outside of mixtures have been frequently studied. studied. •Will binders aged in HMA mixtures have same •Will binders aged in HMA mixtures have same engineering properties?

Research Objective • Quantifying Oxidation of Asphalt Binders Aged in Quantifying Oxidation of Asphalt Binders Aged in Compacted Mixtures Others have compared aging to binder viscosity or p g g y o o stiffness Lack sufficient aging measurements of the binder g g o lack of previous studies specifically relating mixture o properties to adequate aging measurements

Overview Experimental Design Experimental Design A. Aggregate sources: 2 (NV & CO) B. Binders, single source: 2 (PG64-22 & PG64-28) C. Mixture oven-aging levels: 4 (0, 3, 6 and 9 months at 140 ° F) months at 140 F)

Experimental Design A. Aggregate sources: NV & CO 100 Water Abs: 90 NV : 2.7% 80 CO: 0.9% 70 nt Passing 60 50 Perce 40 Nevada 30 Colorado 20 Max Density Line 10 0 No. 200 No. 100 No. 50 No. 30 3/4" No. 40 No. 10 No. 8 3/8" 1/2" No. 16 No. 4 1 " 1 " 1 1/2" 1 1/2" U S. Sieve Opening

Experimental Design B. Binders: single base stock and supplier (Paramount Petroleum Corp. ) Neat PG64-22  SBS Modified PG64-28 

Experimental Design C. Agg. sources and binders combine to 4 different Superpave designed mixtures (6 × 10 6 ESALS) Binder Binder Binder Binder App App Film App. Film App Film Film Sour Source ce Sour Source ce Agg. W gg. Water ter Binder Binder Mineral Mineralogy gy Content Content Thickness Thickness ID ID Location Location Abs. (%) s. (%) Gr Grade ade (  m) (% T (% TWM) M) m) PG64 22 PG64-22 5 4 5.4 9 9 Rhyolite, Nevada Sparks 2.7 Silica Sand PG64-28 5.2 9 Mica Gneiss, PG64-22 4.5 11 Colorado Morrison Mica Schist, 0.9 PG64-28 4.5 11 Quartz Sand

Experimental Design C. 4 Mixture oven-aging levels: 0, 3, 6, & 9 months at 140 ° F  All samples short-term aged loose 4 hrs at 275  F   SGC Compacted Specimens 7 ± 0.5% Air Voids 

Experimental Plan Virgin Long-term oven aging : 3, 6, 9 mo. at 140  F Aggregate Dynamic Modulus, |E*| Short-term oven aging : No Aging 4 hrs at (i.e. 0 mo.) 275  F 275 F Loose Mix Compacted Asphalt Specimen Binder FTIR, Carbonyl Area, CA Mix CA Original CA

Experimental Analysis • CA vs Aging • |E*| vs Aging • |E | vs Aging • |E*| vs CA

Results, Example Carbonyl Area, CA (measurements are being done by Glover at A&M) Carbonyl Area, CA (measurements are being done by Glover at A&M) 100 Wave numbers for 90 th CA the CA 80 measurements 70 bsorbance between 1,650 and 60 1,870 cm -1 50 % Ab 40 30 20 10 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 Wavelength (cm-1)

Results, Carbonyl Area

Statistics, Carbonyl Area CA = β 0 + β 1 (Age) + β 2 (Mix) + β 3 (Mix)(Age) β 0 β 1 ( g ) β 2 ( ) β 3 ( )( g ) [ q [Eqn 1] ] • CA – measured Carbonyl Area; CA measured Carbonyl Area; • β i – regression coefficients, i = 0,…,3; • Age – months of oven aging at 140 ° F; • Mix – categorical variable to differentiate the two mixtures being compared, value of 1 or 0 depending on which agg. and binder combination being considered combination being considered.

Statistics, CA CA = [ β 0 + β 2 (Mix)] + [ β 1 + β 3 (Mix)] (Age) [ β 0 β 2 ( )] [ β 1 β 3 ( )] ( g ) [Eqn 1] [ q ] Compariso Comparison I I Compariso omparison II n II Compariso Comparison III III Comparison IV Compariso IV CA vs. vs. Ag Age CO22 NV22 CO28 NV28 CO22 CO28 NV22 NV28 Mix 0 0 1 1 0 0 1 1 1 1 0 0 1 1 0 0 Variable

Statistics, CA Example CA = [ β 0 + β 2 (Mix)] + [ β 1 + β 3 (Mix)] (Age) [ β 0 β 2 ( )] [ β 1 β 3 ( )] ( g ) [Eqn 1] [ q ] Comparison I Comparison I • Mix = 0, CO22  CA CO22 = [ β 0 ] + [ β 1 ] (Age) • Mix = 1, NV22  CA NV22 = [ β 0 + β 2 ] + [ β 1 + β 3 ](Age)

Statistics, CA CA = [ β 0 + β 2 (Mix)] + [ β 1 + β 3 (Mix)] (Age) [ β 0 β 2 ( )] [ β 1 β 3 ( )] ( g ) [Eqn 1] [ q ] Mi Mixes xes  2  3 P-valu P-v lue Sig. ig. P-value P-value Sig. ig. Compared mpared CO22 CO22 -0.0137 0.600 NS -0.0118 0.013 SH NV22 NV22 CO28 CO28 0.0089 0.702 NS -0.0077 0.052 SH NV28 NV28 NV28 NV28 CO22 CO22 0.1343 <0.001 SH 0.0019 0.652 NS CO28 CO28 NV22 NV22 NV22 NV22 0.1122 <0.001 SH -0.0022 0.614 NS NV28 NV28

Findings, Carbonyl Area 1) CA i 1) CA increased linearly with Age; d li l ith Ag 2) Generally, CA was higher for PG64-22; 3) Within each binder, the intercepts were stat. the same; a) Aggregate source did not significantly affect short-term oxidation; gg g g y 4) Oxidation rates were different between agg. sources; a) a) Agg. source, as it influences mix properties affected binder aging Agg. source, as it influences mix properties affected binder aging

Findings, Carbonyl Area 5) Within each agg., the intercepts were stat. different; ) gg , p ; a) Short-term aging of binders were not the same (original CA was the same) b) b) Polymer modification influences the Non-Linear Fast Rate Oxidation Polymer modification influences the Non Linear Fast Rate Oxidation (short-term region) 6) Within each agg source after Fast Rate Oxidation the 6) Within each agg. source, after Fast Rate Oxidation, the binders aged at the same rate; a) ) Binders from same base stock (similar oxidation characteristics) ( ) b) Indicating Mix Characteristics Influence the rate of binder oxidation

Results, |E*| Nevada Mixes Nevada Mixes

Results, |E*| Colorado Mixes

Statistics, |E*| vs. CA |E*| = β 4 + β 5 (CA) + β 6 (Mix) + β 7 (Mix)(CA) [Eqn 2] | | β 4 β 5 ( ) β 6 ( ) β 7 ( )( ) [ q ] • |E*| – measured Dynamic Modulus, 0.1Hz; |E | measured Dynamic Modulus, 0.1Hz; • β j – regression coefficients, j = 4,…,7; • CA – measured Carbonyl Area; • Mix – categorical variable to differentiate the two mixtures being compared, value of 1 or 0 depending on which agg. and binder combination being considered combination being considered.

Statistics, |E*| vs. CA |E*| = [ β 4 + β 6 (Mix)] + [ β 5 + β 7 (Mix)] (CA) | | [ β 4 β 6 ( )] [ β 5 β 7 ( )] ( ) [Eqn 2] [ q ] Compariso Comparison I I Compariso omparison II n II Comparison III Compariso III Comparison IV Compariso IV 70 and 70 and 100 ° F 10 CO22 NV22 CO28 NV28 CO22 CO28 NV22 NV28 Mix 0 0 1 1 0 0 1 1 1 1 0 0 1 1 0 0 Variable • Analysis conducted for both 70 and 100 ° F

Statistics, |E*| Example |E*| = [ β 4 + β 6 (Mix)] + [ β 5 + β 7 (Mix)] (CA) | | [ β 4 β 6 ( )] [ β 5 β 7 ( )] ( ) [Eqn 2] [ q ] Comparison I at 70 ° F Comparison I at 70 F • Mix = 0, CO22  |E*| 70-CO22 = [ β 4 ] + [ β 5 (CA)] • Mix = 1, NV22  |E*| 70-NV22 = [ β 4 + β 6 ] + [ β 5 + β 7 ](CA) • Same model form for both 70 ° F and 100 ° F

Statistics, |E*|, 70 ° F | |E*| 70 = [ β 4 + β 6 (Mix)] + [ β 5 + β 7 (Mix)](CA) [Eqn 2.a] | 70 [ β 4 β 6 ( )] [ β 5 β 7 ( )]( ) [ q ] Mi Mixes xes  6  7 P-valu P-v lue Sig. ig. P-v P-valu lue Sig. ig. Compared mpared CO22 CO22 -67.58 0.255 NS 160.03 0.005 SL NV22 NV22 CO28 CO28 -36.43 0.530 NS 111.05 0.049 SL NV28 NV28 NV28 NV28 CO22 CO22 -73.79 0.173 NS 184.91 0.001 SH CO28 CO28 NV22 NV22 NV22 NV22 -104.94 0.119 NS 233.89 0.001 SH NV28 NV28

Findings, |E*|, 70 ° F 7) Within each binder, the intercepts were stat. the same; a) Agrees with CA vs Age analysis, Item 3.a 8) Within each binder, Rates of |E*| increase lower for CO; ) , | | ; a) Rate of |E*| increase dependent upon mixture characteristics; 9) Within each agg rate of |E*| increase lower for PG64-28; 9) Within each agg., rate of |E | increase lower for PG64 28; a) Supports that different binders influence the binder aging, particularly polymer modification

Statistics, |E*|, 100 ° F | |E*| 100 = [ β 4 + β 6 (Mix)] + [ β 5 + β 7 (Mix)](CA)[Eqn 2.b] | 100 [ β 4 β 6 ( )] [ β 5 β 7 ( )]( )[ q ] Mi Mixes xes  6  7 P-value 1 P-value Sig. Sig. P-value P-value 1 Sig. Sig. 1 1 Compared mpared CO22 CO22 -54.21 0.027 SL 1 74.38 0.001 SL NV22 NV22 CO28 CO28 -19.51 0.110 NS 35.49 0.004 SL NV28 NV28 NV28 NV28 CO22 CO22 -61.35 0.004 SL 1 77.65 <0.001 SH CO28 CO28 NV22 NV22 NV22 NV22 -96.05 <0.001 SL 1 116.54 <0.001 SH NV28 NV28 1 – Change in result as compared to 70 ° F analysis