Civl23403_Current.qxd 1. Introduction Stone mastic asphalt (SMA) is a gap-graded mix with a skeletal stone-to-stone arrangement that requires a high- er viscosity asphalt binder to keep the interlocked aggre- gates bound and intact. The interlocking nature of the mix is expected to increase the stability and to minimize the lateral displacement of the aggregates that tend to reduce __________________________________________ *Corresponding author E-mail: ratnas@eng.upm.edu.my permanent deformation or rutting of the mix. The concept was first developed in Germany in the early sixties. It was further developed in the United States in the early 1990s, (Selim et al., 1994). Whenever a conventional dense graded aggregate gra dation is altered to a gap graded matrix , the voids in min- eral aggregates (VMA) increases considerably. Such a void content must be filled with a mastic that has excellent shear properties so as to hold the gap graded coarse aggre- gate matrix over a long period of time. In order to achieve this, recycled ground tire rubber of #40 mesh ring Research 1 (2004) 53-58rnal of EngineeThe Jou Laboratory Evaluation of Ground Tire Rubber in Stone Mastic Asphalt R. Muniandy*1, A. A. Selim2, S. Hassim1, and H. Omar1 1Department of Civil Engineering, University Putra, Malaysia 2 Department of Civil & Environmental Engineering, South Dakota State University, Brookings, SD, USA Received 23 April 2003; accepted 5 November 2003 Abstract: Stone mastic asphalt (SMA) is a gap-graded mix whereby stiffer asphalt cement is required to bind the stone matrix or arrangement of stones together. Although various asphalt additives are tra- ditionally available, the use of rubber crumbs in SMA is still a new rresearch endeavor. Many coun- tries around the world are facing serious problems on what to do with reject or discarded tires. In the present study, commercial truck tires, containing 70% natural rubber, were ground and pre-blended in 80-100 penetration asphalt for use in SMA mixtures. An assessment was made of the laboratory per- formance of rubberized SMA in terms of stability, resilent modulus, dynamic creep and tensile strength ratio. It was observed that the performance of SMA with ground tire rubber was for superior as com- pared to SMA mix with unmodified asphalt. Sulfur and Styrene Butadeline Rubber (SBR) were used in rubberized SMA mixes as additives to test the sensitivity of SMA mixtures. As standard practice a 0.3% newly developed cellulose oil palm fiber was used in SMA to minimize the asphalt drain-down effects. Keywords: Stone mastic asphalt, Rubber, Tire, Recycling, Modified, Roads :¢ü∏îà°ùŸŸG) Aɵ£°üŸG …ôî°üdG â∏Ø°S’GSMA≈∏Y IQOÉb iƒbG ᫨ª°U ¢UGƒN hP ¿ƒµj É¡«a Ω~îà°ùŸG â∏Ø°S’G ¿G å«M èjQ~àdG áMƒàØe á£∏N ƒg ( äÉàa ΩG~îà°SG ¿G ’G ,â∏Ø°S’G ™e IOÉY Ω~îà°ùJ äÉaÉ°V’G øe ´GƒfG I~Y ¿G øe ºZôdG ≈∏Y .áe~îà°ùŸG Qƒî°ü∏d áØ∏àıG ΩÉéM’G ÚH ∂°SÉ“ øjƒµJ ) ∫G äÉ£∏N ‘ •É£ŸGSMAøe ¢ü∏îàdG á«Ø«c »gh ’G IÒ£N á∏µ°ûe ¬LGƒJ ⁄É©dG ∫ƒM ¿G~∏ÑdG øe ~j~©dG .¤h’G á«ãëÑdG ¬∏MGôe ‘ ∫Gõj ’ ( hP â∏Ø°SG ¤G É¡àaÉ°VGh É¡≤ë°S ~©H áÄŸÉH 70 áÑ°ùæH »©«Ñ£dG •É£ŸG øe ¿ƒµŸG äÉæMÉ°ûdG äGQÉWG äÉàa Ω~îà°SG áãj~M á°SGQO ‘ .áØdÉàdG äGQÉW’G ) á£∏N êÉàf ’ 100-80 ¥GÎNG πeÉ©eSMA,áfG~∏dG QG~≤eh ,á«JÉÑãdG ¢SÉ«≤H πª©ŸG ‘ äÉæ«©dG AGOG QÉÑàNG ≥jôW øY á£∏ÿG √òg º««≤J ” ~bh .( ) ∫G á£∏N AGOG ¿G áHôéàdG ∫ÓN øe ~Lh ~bh .~°ûdG Iƒb áÑ°ùfh ,»µ«eÉæj~dG ∞MõdGhSMA~«©H ~M ¤G ¥Éa ~b äGQÉW’G ¥ƒë°ùe É¡«dG ±É° ŸG ( ) øjOÉJƒ«H øjôjÉà°ùdGh âjȵdG •É£e Ω~îà°SG ɪc .áaÉX’G ∂∏J ¿h~H ÉgAGOG SBR) ∫G äÉ£∏N ‘ (SMAá«°SÉ°ùM ¢SÉ«≤d ±É° e πeÉ©c á«WÉ£ŸG ( ) ∫G äÉ£∏N ‘ áÄŸÉH 0^3 .áÑ°ùæH π«îædG ±É«dG Rƒdƒ«∏°S âjR ΩG~îà°SG ” »°SÉ«b AGôLÉch .äÉ£∏ÿG √ògSMA»∏Ø°ùdG ±Gõæà°S’G äGÒKÉJ øe ~ë∏d ( .â∏Ø°SÓd :á«MÉàØŸG äGOôØŸG.¥ô£dG ,(Qƒ£e) ∫~©e ,ôjh~àdG IOÉYG ,QÉWCG ,•É£e ,»¨ª°üdG ôé◊G øe â∏Ø°SCG from shredded tires was used in the pre-blending of unmodified 80-100 asphalt, (Lundy et al., 1987) (SMA in general in prone to asphalt drain-down during mix stor- age and transportation to the construction site). Even after the placement of the mix, the softening of the mastic dur- ing hot days tends to slowly drain down the asphalt. This slow drain down effect of the binder is further accelerated by the contact pressure from wheel loading which may result in the ultimate loss of bitumen from the top part of the asphalt layer. To minimize this problem cellulose fibers were used. 2. Material Characterization SMA is a high strength mix that requires good quality aggregates. Inferior quality aggregates may be crushed upon repeated loading that may in turn alter the stone matrix posture entirely. Therefore, aggregate quality must be controlled to ensure the superior performance of SMA mixtures, (Bukowski, 1991). Since granite is abundantly available in Malaysia, it was identified as a prime candidate for use in SMA mix. In our research, the Public Works Department specifica- tions (PWD), (Public Works Department, 1985), for aggregate properties were adopted in for mulating stone mastic asphalt, Table 1. A typical gradation was formulat- ed that would give higher stability and reliability with a maximum size of 14 mm. Approximately, 80% of the aggregates were larger than 2 mm, and more than 70 per- cent larger than 8mm, Figure 1. Traditional 80/100 penetration asphalt is considered to be too soft for use in SMA. It has to be modified to increase the viscosity. Polymers have been traditionally used to modify asphalts for specific applications. *UMA = Unmodified asphalt TRA2 = Tire rubber asphalt with 2% rubbe content by weight of asphalt TRA3 = Tire rubber asphalt with 3% rubber content by weight asphalt TRA4 = Tire rubber asphalt with 4% rubber content by weight asphalt TRA4-2.2S = Tire rubber asphalt with 4% rubber content and 2.2% butanol by weight of asphalt TRA4-5B = Tire rubber asphalt with 4% rubber content and 5% sulfur by weight of asphalt 0 10 20 30 40 50 60 70 80 90 100 0.0000 0.1000 0.2000 0.3000 0.4000 0.5000 0.6000 0.7000 0.8000 0.9000 1.0000 Sieve Number P er ce n t P as si ng LB DB UB MD ACTB Linear (MD) # 20 0 # 10 0 # 80 # 50 # 40 # 30 # 20 # 16 # 10 # 8 # 4 1/ 4" 3/ 8" 1/ 2" 3/ 4" 1" Aggregate more costly to produce. Surface easy to finish Critical mixture readily unstable with slight excess of asphalt or water Porous mixture - lack tensile strength when mixed with fuel oil best result with heavy bitumen Harsh mixture, inclined to segregate not critical Porous mixture - lack tensile strength when mixed with fuel oil best result with heavy bitumen Figure 1. Stone mastic asphalt (SMA) and conventional gradations No. Type of Test Result % PWD Requirement 1 L.A Abrasion 19.70 ≤ 30 % 2 Crushing Value 26.20 ≤ 30 % 3 Impact Value 12.6 ≤ 15 % 4 Soundness Test (Sodium Sulfate) 1.76 ≤ 15% 5 Polishing Stone Value 50.9 ≥ 49 6 Flakiness and Elongation Test ( 3:1 ratio) 17.1 ≤ 20% 7 Specific Gravity 2.62 ≥ 2.60 8 Water absorption 0.469 ≤ 2% Table 1. Granite aggregate properties used in SMA mix Asphalt Blend* Penetration (mm ) Softening Point (°C) Thin Film Oven UMA 85.7 52.3 0.003 TRA2 79.3 63.3 0.009 TRA3 71.5 68.2 0.010 TRA4 62.0 70.3 0.011 TRA4-2.2S 61.7 70.3 0.009 TRA4-5B 60.3 71.0 0.007 Table 2. Physical properties of rubberized asphalt 54 ring Research 1 (2004) 53-58rnal of EngineeThe JouMuniandy et al. / However, the use of polymers increases the overall cost of asphalt mix production. Recent experiences in polymer modified asphalt (PMA), (Bukowski, 1991) showed that it costs at least three times more compared to conventional asphalt. Tire rubber from recycled shredded tires were pulverized to #40, which is about 450 micron size. The tire rubber powder was checked for deleterious material like particles from the steel belting that could have been *mixed with the tire powder during the shredding, grind- ing, and separation process. This was done to ensure the purity of the selected material. The unmodified asphalt was blended with 2%, 3% and 4% ground tire rubber. The physical properties of the tire rubber modified asphalt are shown in Table 2. Sulfur and styrene butadiene random (SBR-Butanol) were also used in our study to check the sensitivity of SMA mixes. 3. Marshall Mix Design Marshal mix design was carried out in accordance with the American Standard for Testing and Materials (ASTM) D1559. Fifteen 100 mm diameter Marshall specimens were prepared for each tire rubber blended asphalt. Each speciment required approximately 1200 grams of granite aggregates. A total of 60 batches were used in the prepara- tion of SMA samples. Four sets of asphalt blended with 2%, 3%, and 4% tire rubber and one unmodified set were prepared for the mix design. The aggregate samples were heated in an oven to 170°C while the asphalt blends were heated to 165°C based on the viscosity-temperature of the asphalt. For each asphalt type, five sets of three specimens, each with an asphalt content of 4%, 5%, 6%, 7% and 8%, were prepared. A 0.3% cellulose fiber by weight of aggregates were used in all SMA samples. The compaction of the SMA mixtures was done using an automatic compactor with a standard 50 blow on each side. It is a common practice in Malaysia to use 50 blow Marshall compaction for hot mix asphalt mixtures. 4. Optimum Asphalt Content The optimum asphalt content for each rubber blend was determined at maximum bulk density, maximum stability, and at 4% air voids using the Asphalt Institute Method, Table 3. The average of four optimums was considered in the preparation of specimens for performance tests. Additional specimens were made with the optimums of each asphalt blend. Marshall properties tests were carried out in accordance with ASTM D1559. The performance of the rubberized and control samples are shown in Figures 2 to 7 and in Table 3. The resilient modulus test was carried out in accordance with ASTM D4123, Figure 4. Samples with more than 4 percent tire rubber blended asphalt were found to have non-desirable properties. Contents higher than 4% tire rubber were tried. However, it was found to be very difficult to blend the pulverized tire beyond 4% due to clumping and dispersion. Hence the study focused on a maximum tire rubber content of 4% (by weight of asphalt) only. Marshall tests with various proportions of tire blend showed that SMA with 4% ground tire rubber (TRA4) dis- played better properties compared to other SMA mixtures. However, SMA with a tire rubber blend of more than 4% did not display good Marshall properties. The workability was greatly affected. Hence, the results of SMA with greater than 4% tire rubber are not dis- Mix MR (MPa) Stability (kN) Flow Bulk Density VMA (%) VTM (%) VFA (%) Optimum AC UMA 3436.3 12.65 3.43 2.28 18.16 4.95 72.72 5.86 TRA2 3612.7 14.31 4.50 2.311 17.11 3.62 78.86 5.9 TRA3 3764.7 15.01 4.76 2.313 16.96 3.44 79.70 6.0 TRA4 3942.0 17.30 4.66 2.321 16.76 5.60 66.58 6.1 TRA4-2.2S 4275 17.88 4.60 2.35 17.82 4.95 72.31 - TRA4-5B 4885 18.31 4.69 2.35 17.79 5.04 69.87 - Table 3. Results of SMA mix performance Load Cycle Sample 1% Strain 3% Strain 1 2 3 Ave 1 2 3 Ave UMA 52 92 61 68.3 1711 1755 2100 1855.3 TRA4 79 71 90 80.0 2782 2712 2854 2782.7 TRA4-2.2S 180 105 158 147.7 3244 3004 3328 3192.0 TRA4-5B 125 100 150 125.0 8360 4641 4885 5962.0 Table 4. Permanent deformation of rubberized SMA 55 ring Research 1 (2004) 53-58rnal of EngineeThe JouMuniandy et al. / 8 10 12 14 16 18 3 4 5 6 7 8 9 Percent Asphalt St ab ili ty (k N) UMA TRA2 TRA3 TRA4 1 2 3 4 5 6 7 4 5 6 7 8 % Asphalt tire rubber blend Fl ow (m m ) UMA TRA2 TRA3 TRA4 2.230 2.250 2.270 2.290 2.310 2.330 4 5 6 7 8 % Asphalt tire rubber blend B ul k D en si ty UMA TRA2 TRA3 TRA4 15 16 17 18 19 20 4 5 6 7 8 % Asphalt tire rubber blend % V M A UMA TRA2 TRA3 TRA4 1800 2300 2800 3300 3800 4300 4800 3 4 5 6 7 8 9 Percent Asphalt R es il ei n t M o d u lu s (M P a) UMA TRA2 TRA3 TRA4 0 1 2 3 4 5 6 7 8 9 10 4 5 6 7 8 % Asphalt tire rubber blend VT M (% ) UMA TRA2 TRA3 TRA4 Figure 2. Marshall stability vs asphalt blend Figure 5. Flow vs asphalt blend content Figure 6. VMA vs asphalt blend Figure 7. VTM vs asphalt blend content Figure 4. Resilient modules vs % asphalt blend content Figure 3. Bulk density vs asphalt blend content 56 ring Research 1 (2004) 53-58rnal of EngineeThe JouMuniandy et al. / cussed in this paper. A sensitivity analysis was carried out on SMA-4 using sulfur and a styrene butadiene rubber product (Butanol). A total of fifteen samples were prepared and tested for Marshall properties, Table 3. The optimum sulfur and butanol contents were determined by preparing and testing 15 SMA specimens with sulfur proportions of 1,2,3,4, and 5 % and 15 specimens with 2,4,6, and 8 % butanol by weight of asphalt tire rubber blend. Figure 8 shows the optimum sulfur and butanol content determined from the resilient modulus test. 5. Dynamic Creep Performance SMA specimens were tested for permanent deforma- tion in the uniaxial repeated compressive load. Sets of three specimens were made for each asphalt tire rubber blend at their individual optimum asphalt content and trimmed to 50mm with a 2:1 ratio of the diameter. The specimens were then tested in accordance with ASTM D3497 to determine the number of load cycles to reach 1% strain , 3% strain and ultimate failure, Table 4 and Figure 9. An Australian IPC MATTA machine was used in the testing of the SMA specimens. The selected test parame- ters were as follows: Test temperature : 50°C Pulse Width : 500ms Pulse Period : 2000 Test Load Stress : 500kPa 6. Moisture Induced Damage Analysis Roads in tropical countries like Malaysia are exposed to extreme moisture conditions. As such, it was deemed appropriate to carry out a moisture induced damage test on SMA samples. The samples were prepared and tested using the Modified Lottman Test in accordance with AASHTO T283. Two sets of three specimens for each test were prepared. One set of samples was used as a control while the other set was vacuum saturated at 28 in Hg with water for 25 minutes. Both sets (control and conditioned) were then tested for indirect tensile strength at 25°C using a loading rate of 51 mm/minute. The tensile strength ratio (TSR) of the samples was determined as a ratio of condi- tioned strength over control strength, Figure 10. A mini- mum of 0.7 is generally specified. The higher the TSR , the higher the durability of the mix. 7. Discussion The Marshall mix properties of SMA with tire rubber blends of 2%, 3%, and 4% generally increased with a gradual increment in optimum asphalt content. The opti- mum asphalt contents ranged between 5.86 to 6.10%. 2600 2800 3000 3200 3400 3600 3800 4000 4200 4400 4600 4800 5000 0 1 2 3 4 5 6 7 8 9 10 11 % Sulfur/Butanol in 4% tire rubber blend R es ili en t M od ul us , M P a Sulfur Butanol Figure 8. Resilient modulus vs sulfur and butanol content 0 5000 10000 15000 20000 1% 3% rupture % Strain N um be r o f l oa d cy cl es SMA0 SMA4 SMA4+2.2 SMA+5 Figure 9. Strain vs number of load cycles 84 86 88 90 92 94 96 98 100 0% 2% 3% 4% 4+ 2.2 % 4+ 5% Type of asphalt blend TS R (% ) Figure 10. Tensile strength ratio of SMA mix 75 ring Research 1 (2004) 53-58rnal of EngineeThe JouMuniandy et al. / SMA with 4% tire rubber showed higher values in terms of stability and resilient modulus. A sensitivity analysis on SMA-4% tire rubber with 2.2% sulfur and 5 % SBR addi- tive greatly improved the SMA mixture’s permanent deformation or rut potential. It was found that SMA with 4% blend took 80 load cycles to reach 1% strain and 2782 load cycles to reach 3% strain while the control SMA mix displayed a much lower value of 68.3 cycles for 1% and 1855 cycles for 3% strain levels. The addition of a small amount of sulfur and butonal, separately, resulted in a marked increase in the dynamic creep performance of the mix. The addition of sulfur 2.2% by weight of rubber- asphalt blend increased the load cycles to 148 for 1% strain and to 3192 for 3% strain. However, the addition of 5 % butanol by weight of rubber-asphalt blend had the highest value. It took 5962 cycles to reach the 3 % strain level and almost 20,000 load cycles to failure. The control mix on the average failed at 5181 cycles. The indirect tensile test also showed that all SMA mixes displayed a tensile strength ratio of more than 70%. SMA with 4% tire rubber without any additives showed a remarkable 98% TSR. However, SMA with 4% tire rubber and additives did not improve the moisture damage of SMA. Instead lower TSR values of 95% and 96%, respec- tively, were displayed. 8. Conclusions In conclusion, the rubberized stone mastic asphalt mix has shown great promise. This mix may be able to super- sede the conventional mix. Since SMA requires a stiff binder the traditional 80/100 penetration asphalt can be blended with recycled ground tire rubber. This will give good binding properties for SMA performance. A cost effective SMA with reject or discarded tire rubber will give excellent pavement performance not only in tropical regions with high rainfall, but also in regions with high temperatures. The laboratory study has shown that there is a great potential for ground tire rubber for use in SMA, up to 4% with 0.3% cellulose oil palm fiber. However, a field trial needs to be carried out to evaluate its full potential. References Bukowski, J., 1991, “SMA comes to USA,” Journal of the National Asphalt Pavement Association Vol.6, No.2, p.5. Lundy, J.R., Hicks, R.G., Richardson, and E., 1987, “Evaluation of Rubber Modified Asphalt Performance Mt. St. Helens Project,” Asphalt Pavemen Technology, Proceedings Associations of Technologists, Vol.56, pp.573-598. Public Works Department, 1985, “Manual of Quartzite Based Stone Matrix Asphalt Mixtures (SMAM)”, Proceedings of the third Materials Engineering Conference 804, ASCE, p.635-642. 85 ring Research 1 (2004) 53-58rnal of EngineeThe JouMuniandy et al. /