Figure 104 – uploaded by zhaoxing xie

See full PDF downloadDownload figure

using the centrifuge extractor and tetrahydrofuran (THF) as the solvent. — Figure 104 using the centrifuge extractor and tetrahydrofuran (THF) as the solvent.

Related Figures (179)

purposes is one practical solution (Mashaan et al. 2014). Crumb rubber modifier (CRM), sometimes referred to as ground tire rubber (GTR), landfills is not. Recycling tire-derived aggregates and crumb rubbers for construction vith them, retreading and recycling are environmentally friendly, while putting them into

Figure 2-2. Ambient ground rubber (left); cryogenically fractured rubber (right) (Shen & Amikhanian 2005)

modification of asphalt in a continuous blend system at an asphalt mix plant. Figure 2-3. CR modification of asphalt in a continuous blend system (http://maxlinktyrerecycling.com)

Figure 2-4. Feed system for the dry process (Hines 2007) In 1950, the use of CRM in asphalt as a stress-absorbing membrane interlayer and open- graded mixtures (Heitzman 1992). Figure 2-4 shows the feed system used for

Table 2-1. Some applications of dry-processed CRM asphalt mixtures by state

Table 2-2. Some applications of wet-processed CRM asphalt mixtures by state

Figure 2-5. Asphalt rubber (left); terminal blends (right) 9-10% CRM weight of total binder. Terminal blends look different from traditionally particulate matter remaining. Sometimes, polymers are added with fine CRM, typically at

Figure 3-1. Flow chart for testing rubberized asphalt binder properties asphalt binder to evaluate the effects (Fig. 3-1). DSR, BBR, and separation tests were then conducted on unaged and aged rubberized

Aging was accelerated in the standard rolling thin-film oven (RTFO) at 163 °C for 85

Figure 3-2. CRM (left) and TOR (right) binder was mixed at 170 °C and 900 RPM for 45 minutes in the laboratory. Table 3-1

Figure 3-5. Dynamic shear rheometer

The separation tube test was performed to determine CRM tendency to separate

specimens. Tendency of Polymer from Polymer Modified Asphalt. Figure 3-7 shows th

Figure 3-8 shows the values of G*/sin(5) and phase angle at 76 °C for the binders temperatures, a complex shear modulus and phase angle were measured for both

Note: Y: P-value < a = .05 (significant difference); N: P-value > o = .05 (no significant difference). The phase angle of the unaged rubberized asphalt binders decreased as the dose of asphalt binder. It also increased as CRM dose increased from 8% to 12%, regardless

Note: Y: P-value < a = .05 (significant difference); N: P-value > a = .05 (no significant difference).

Table 3-3. Statistical analysis of the phase angle of unaged rubberized asphalt binders

PG grades (12.0 °C) at high temperature, regardless of the type of base asphalt.

“or an RTFO-aged rubberized asphalt binder consisting of a PG 67-22 asphalt and 8%

For RTFO-aged, rubberized asphalt binder consisting of a PG 67-22 base binder increased as CRM dose increased from 8% to 12%, regardless of TOR dose and base

Figure 3-14. m-values of PAV -aged rubberized asphalt binder at -12 °C 3.3.3 Effect of TOR on separation resistance

to some extent, TOR reduces separation severity in rubberized asphalt binder. 3.4 Summary and Conclusions for Rubberized Binder Properties

properties with both dry- and wet-processed CRM binders (Fig. 4-1). 4,2 Materials and Test Procedures

Table 4-1. Properties of materials used Table 4-2. Properties of cellulose fibers* the materials tested.

* The data are from the manufacturer, Fiberand, Inc. (Miami, FL).

Table 4-3. Properties of mineral fibers* Figure 4-2. Mineral fibers

These materials were combined to produce gradations similar to those for

Figure 4-4. A ggregate gradations of PEM and SMA used for this study Specifications, Section 828. gradations meet the control tolerances and design criteria of the Standard

Table 4-5. Surface-capacity test results

Figure 4-5. Surface-capacity test

Modified Marshall Method Dry-processed PEM is mixed in the following steps: (1) aggregates are mixed with

Table 4-6. Asphalt content as determined by the modified Marshall method determined as the value at which V MA reached the lowest point on the curve plot.

Drain-down test

Table 4-7. Results of drain-down testing Table 4-8. Volumetric properties of PEMs 4-8 presents the OAC and other volumetric properties of the four PEM.

Note: OAC = optimal asphalt content; Gm = maximal specific gravity; Gmb = bulk specific gravity; AV =air voids; VMA = voids in mineral aggregate; VFA = voids filled with asphalt. Based on these tests, the final OAC was 6.5% for wet-processed rubberized

Note: OAC = optimal asphalt content; Gmm = maximal specific gravity; AV = air voids; VMA =voids in mineral aggregate; VFA = voids filled with asphalt.

Table 4-9. Volumetric properties of SMA mixtures mixtures at OAC. Marshall hammer (Fig. 4-8). Table 4-9 presents the volumetric properties for all SMA

steel wheel on a pneumatic hose at 100 psi of pressure for 8,000 cycles. Figure 4-9. APA Test

respectively. Figure 4-11. SMA average rut depths

Figure 4-10. PEM average rut depths

Figures 4-13 and 4-14 show the results for PEM and SMA samples. Rutting depth

Table 4-10. Results of drain-down testing drain-down while the others had none.

and placed in a foil pan for transfer into a 177 °C +2 °C oven (Fig. 4-15).

4.5 Summary and Conclusions for Mix Design Verification The volumetric properties of both PEM and SMA mixed with wet- and dry-processed

Table 5-1. Dynamic modulus test matrix collected. Figure 5-1 presents a flow chart of testing. E* were measured at different and a typical aggregate granite used by GDOT. Table 5-1 shows the types of date

Figure 5-1. Flow chart for determining the dynamic moduli of rubberized PEM and 5,2 Materials and Test Procedures mixture specimens used in this study had target air voids of 170.5% and 5+0.5%,

Figure 5-2. Gyratory compaction of PEM and SMA specimens 0.1 Hz, 1 Hz, 10 Hz) for 18 hours at 4 °C, 3 hours at 20 °C, and 3 hours at 45 °C.

Figure 5-3. Specimen coring and cutting Dynamic modulus testing was conducted for each PEM and SMA mixture with an

Figure 5-4. Dynamic modulus samples and conditioning

Figure 5-5. AMPT device and instrumentation 5.3 Results and Discussions

asphalt binder types at 45 °C was 90-92% lower than at 4 °C.

Figure 5-6. PEM dynamic modulus vs. temperature at 10 Hz

Figure 5-8. PEM dynamic modulus vs. temperature at 0.1 Hz Dynamic modulus dependence on load frequency frequency. Under a constant test temperature, the dynamic moduli of all PEM

Figure 5-9. PEM dynamic modulus vs. frequency at 45 °C

20 °C; and slightly lower than hybrid and SBS-modified PEM at 4 °C.

Figure 5-14. PEM phase angle vs. temperature at 0.1 Hz Figure 5-13. PEM phase angle vs. temperature at 1 Hz

Master curve of dynamic moduli Master curves of asphalt mixtures allow performance comparisons over extended

To determine whether dry-processed rubberized PEM performs as well as the dynamic modulus; 6 +a = maximum value of dynamic modulus; and B, y = parameter

Note: Y: P-value < o = .05 (significant difference); N: P-value > a = .05 (no significant difference). Table 5-2. Statistical analysis of PEM dynamic modulus Master curve of phase angles [he master curve of the phase angles of four PEM mixtures (Fig. 5-19) shows that the

5.3.2 SMA dynamic modulus and phase angle

modulus at low frequency and high temperature (0.01 or 0.1Hz at 45 °C).

Figure 5-22. SMA dynamic modulus vs. temperature at 0.1 Hz

Figure 5-23. SMA dynamic modulus vs. frequency at 45 °C

Master curve of dynamic moduli

Note: Y: P-value < a = 0.05 (significant difference); N: P-value > a =0 .05 (no significant difference).

Table 5-3. Statistical analysis of SMA dynamic modulus

samples decreased with increased frequency, while at 45 °C, they increased.

Note: Y: P-value < o = .05 (significant difference); N: P-value > a = .05 (no significant difference). Table 5-4. Statistical analysis of SMA phase angle 5.4 Summary and Conclusions on Dynamic Modulus The effect on the dynamic modulus of modifying PEM and SMA mixtures four

Figure 6-1. Flow chart for determining the fatigue life of rubberized PEM and SMA 6.2 Specimen Fabrication and Test Procedures

Figure 6-2. Specimen fabrication S-VECD Direct Tension F atigue Test

7.2 Materials and Test Procedures

Figure 7-2. Georgia asphalt weathering device

Figure 7-3. |E*| master curves: (a) unaged and 1000-hour aging; (b) unaged and 3000-hour aging Furthermore, a t-test (a = 0.05) was performed to compare |E*| between aging

factors (Figure 7-4). The ratio of |E*| between the long-term aged and unaged specimens can elucidate the PEM samples, indicating that it foamed due to the interaction of the environment

Figure 7-6. Modular aging ratio for wet- processed rubberized PEM

Figures 7-5 to 7-8 indicate that both AR1 and AR2 for all PEM mixtures increased

on |E*| at lower frequency (0.1 Hz) or higher temperature (45 °C). Figures 7-10 to 7-13 show aging ratio (AR) results for all SMA mixtures. Both etween unaged and 1,000-hour aging, but 3,000-hour aging had a significant effect

Figure 7-10. Modular aging ratio for dry-processed rubberized SMAs mixes after 1,000- or 3,000-hour weathering.

Figure 7-11. Modular aging ratio for wet- processed rubberized SMAs

7.3.3 Influence of weathering on fatigue life jor damage throughout the test. End-failure tests are not as good since

Figure 7-14. Failure locations: (a) mid-failure; (b) end-failure (Hou 2009) more end-failures than unaged samples; few unaged samples had end-failures. had mid-failures; the CX test was successful. It also found that aged samples had

Figure 7-15. SMA fatigue life under strain control after 3,000-hour aging

depth, regardless of aging. Figure 7-18 shows the rutting depth results for unaged and aged SMA mixtures. As SBS-modified PEM mixtures. Furthermore, hybrid PEM mix had the lowest rutting samples, dry-processed CRM SMA had the deepest rutting followed by SBS-modified

7.3.9 Influence of weathering on C antabro loss SMA, like hybrid PEM, had the least rutting, regardless of aging.

7.3.6 Influence of weathering on indirect tensile strength

Figures 7-22 and 7-23 illustrate the IDT results for unaged and aged PEM and

nano-structure characteristics. Figure 8-1 shows the experimental design.

Figure 8-3. Separated asphalt binder and THF Gel Permeation Chromatography

Figure 8-4. GPC system used in this study manufacturer. Each binder sample, dissolved in THF, went through a 0.45-um syringe filter prior concentration rate was 0.5% by weight of binder as recommended by the equipment

Figure 8-5. A typical chromatogram of an asphalt binder proximately 21 minutes (Fig. 8-5). Each sample was tested three times, and the In Figure 8-5, the area under the curve represents 100% of the binder molecules to the injection module. A sample volume of 50 ul was injected for each test, which several groups. In this study, the chromatogram profile was partitioned into 13 slices

Table 8-1. Assignments of the main bands of asphalt binder in FTIR spectra

Figure 8-6. FTIR system used in this study

Figure 8-7. G* of asphalt binder extracted from rubberized PEM highest value. Binders extracted from mixtures stored for 30 minutes had the same G* increased their G*, and binder extracted from 90-minute aged mixtures had the

Table 8-2. Statistical analysis of G* of asphalt binder from rubberized PEM All binders were tested at 76 ‘C and thus exhibit viscoelasitc properties. As

Figure 8-8. Phase angle of asphalt binders extracted from rubberized PEM recovered binders stored for 30 minutes, indicating that their elasticity is equivalent.

Table 8-3. Statistical analysis of phase angles of binders from rubberized PEM Figure 8-9 shows that the G*/sinéd value of binders extracted from dry-processed

Figure 8-9. Rutting resistance of asphalt binders extracted from rubberized PEM thus, their rutting resistance is equivalent. rubberized PEM stored for 30 minutes had the same value as the wet-processed and

Table 8-4. Statistical analysis of asphalt binder rutting resistance and binder results in a stiffer asphalt binder due to dry-process mixing and aging.

Figure 8-10. LMS of asphalt binder extracted from rubberized PEM FTIR results Figure 8-11 shows infrared (IR) spectrum analysis images of four asphalt binders

Figure 8-11. FTIR spectra of CRM asphalt extracted from rubberized PEM

Table 8-5. Statistical analysis of the G* of asphalt binder from rubberized SMA after 90 minutes’ storage and wet-processed SMA was found.

Figure 8-13. G* of asphalt binder extracted from rubberized SMA same level. that the G* of asphalt binder in dry- and wet-processed rubberized SMA can reach the Table 8-5 shows the results of a one-way analysis of variance of the significance

Table 8-6. Statistical analysis of phase angle of binder from rubberized SMA

Figure 8-14. Phase angle of asphalt binder extracted from SMA Table 8-6 indicates no significant difference between asphalt binders extracted after 90 minutes’ storage was lowest. For the binders tested, longer storage increases oinder. The phase angle of binders extracted from dry-processed rubberized SMA

Note: S=significant difference; N=no significant difference; significance level=0.05 Figure 8-15 shows that the G*/siné value of binders extracted from dry-processed

Figure 8-15. Rutting resistance of binder extracted from rubberized SMA in dry-processed rubberized SMA reach a level similar to that of wet-processed.

Table 8-7. Statistical analysis of asphalt binder rutting resistance

Figure 8-16. LMS of asphalt binder extracted from rubberized SMA dry-processed rubberized SMA based on the GPC test as a function of aging time. Figure 8-16 shows the LMS values of control and binders extracted from a loose FTIR results

Figure 8-17. FTIR spectra of CRM asphalt extracted from rubberized SMA

Figure 8-18. Bonding ratio (C=O) in CRM asphalt extracted from dry- and to that of wet-processed SMA. wet-processed rubberized SMA dry-processed CRM SMA mixture increases with storage. After 30 minutes, it is close

were used in the AFM tapping mode to avoid damaging the asphalt binder surface Wet- processed rubberized asphalt binder was produced by mixing -30 mesh CRM

Figure 8-20. Topographic and phase images of PG 67-2:

Figure 8-21. Phase images of unaged rubberized asphalt binders

Figure 8-22. Phase images of short-term aged rubberized asphalt binders

8.4 Summary and Conclusions on CRM-Asphalt Interaction A series of rheological property tests and chemical analyses were conducted on

dry- processed rubberized CRM asphalt binder and least in SBS asphalt binder. GPC results

Figure 9-2. LMS of asphalt binder extracted from PEM and lightweight elements, reducing the LMS value change. FTIR results asphalt binder production process, the rubbers absorb the asphalt binder’s resin, oil Figure 9-3 shows the infrared (IR) spectrum analysis of four asphalt binders extracted

Figure 9-4. Bonding ratio (C=O) of asphalt binder extracted from rubberized PEM photo-oxidation, and other uncontrolled natural phenomena influence the nature and depending on weathering time and asphalt binder type. Evaporation, leachins 9.3.2 Effect of weathering on asphalt binder in SMA

Figure 9-5, G*sin(8) (at 19 °C) of asphalt binder extracted from SMA and wet- processed rubberized CRM asphalt binder was almost identical. GPC results

FTIR results

onhancement after weathering 1,000 hours.

Figure 9-7 shows the ratio change (C=O) in the asphalt binder extracted from four

Figure 9-7. Ratio of bonding (C=O) of asphalt binder extracted from SMA

stringline using a hollow steel pipe (Fig. 10-2).

Table 10-1. Field inspection test results

Figure 10-4. Dry-processed rubberized OGFC surface

The Hamburg test using a retrofitted APA was performed on the core samples

Figure 10-8. Field core specimens forA PA Hamburg testing Mixtures showing excessive susceptibility to moisture damage tend to undergo

Figure 10-9. Two different outputs of the HWTD test (Bhasin et al. 2004) rutting depth of 12.5 mm indicate less susceptibility to moisture damage.

Figure 10-10. Stripping point determination

Table 10-2. Rut depth per number of wheel passes

Table 10-3. Summary of test results

‘igure 10-13 present the values of post-compaction consolidation and creep slope.

The core samples with dry- or wet-processed rubberized OGFC had significantly

Table 10-4. Rates of rutting at the two-cycle interval