Saeed Ghaffarpour Jahromi and Ali Khodaii

CARBON FIBER REINFORCED ASPHALT CONCRETE

and ** Ali Khodaii Department of Civil Engineering Amirkabir University of Technology Tehran, Iran

: 0591. . . : . . . .

* To whom correspondence should be addressed 380878221989+ :E-mail: [email protected], Tel 01034546)12(89+ :* * E-mail: [email protected], Tel
8002 Paper Received 27 January 2008; Revised 28 February 2008; Accepted 4 June

8002 October

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

553

Saeed Ghaffarpour Jahromi and Ali Khodaii

ABSTRACT Fibers are often used in the manufacture of other materials. For many years, they have been utilized extensively in numerous applications in civil engineering. Fiberreinforcement refers to incorporating materials with desired properties within some other materials lacking those properties. Use of fibers is not a new phenomenon, as the technique of fiber-reinforced bitumen began as early as 1950. In all industrialized countries today, nearly all concretes used in construction are reinforced. A multitude of fibers and fiber materials are being introduced in the market regularly. The present paper presents characteristics and properties of carbon fiberreinforced asphalt mixtures, which might improve the performance of pavements. To evaluate the effect of fiber contents on bituminous mixtures, laboratory investigations were carried out on the samples with and without fibers. During the course of this study, various tests were undertaken, applying Marshall test, indirect tensile test, creep test, and resistance to fatigue cracking by using repeated load indirect tensile test. Carbon fiber exhibited consistency in results, and as such it was observed that the addition of fiber does affect the properties of bituminous mixtures, i.e. an increase in its stability and a decrease in the flow value as well as an increase in voids in the mix. Results indicate that fibers have the potential to resist structural distress in pavement, in the wake of growing traffic loads, and thus improve fatigue by increasing resistance to cracks or permanent deformation. On the whole, the results show that the addition of carbon fiber will improve some of the mechanical properties like fatigue and deformation in the flexible pavement. Key words: modified asphalt mixture; reinforcement; carbon fiber; creep; deformation; fatigue

356

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

October 2008

Saeed Ghaffarpour Jahromi and Ali Khodaii

CARBON FIBER REINFORCED ASPHALT CONCRETE

1. INTRODUCTION Experts and engineers are eager to improve the performance of asphalt pavements and to this end they have applied different procedures, including modification of asphalt binder. At present, addition of polymers is a common method applied for binder modification, although various types of fibers have also been evaluated. It is widely believed that the addition of fibers to asphalt enhances material strength as well as fatigue characteristics while at the same time adding ductility. Likewise, carbon fibers may also offer excellent potential for binder modification due to their inherent compatibility with asphalt cement and superior mechanical properties. With new developments in production, carbon modified binder has become cost competitive with polymer modified binders. Further, it was expected that carbon fibermodified asphalt mixtures would increase stiffness and resistance to permanent deformation and, similarly, that the fatigue characteristics of the mixture would improve with the addition of discrete carbon fibers. Because of the high tensile strength of carbon fibers, cold temperature behavior of asphalt mixtures was also expected to improve. Finally, carbon fiber modified asphalt could produce a higher quality asphalt mixture for pavements. However, the performance of asphalt mixtures reinforced by these fibers is yet to be improved due to lack of understanding of reinforcing mechanisms and ways of optimizing fiber properties (i.e., fiber diameter, length, surface texture, etc.). Considering the characteristics of fibers, its addition to asphalt concrete could be very different. For example, if fibers are too long, it might create the so-called balling problem (i.e. some of the fibers may lump together) and the fibers may not blend well with the asphalt concrete. If the fibers are too short, they may not provide any reinforcing effect and may just serve as an expensive filler in the mix. 1.1. Literature Review Using fibers to improve the behavior of materials is not a new concept. Fibers are widely used as reinforcing agent in concrete [15], however, the modern ways of fiber reinforcement started in the early 1950s. Zube was the foremost researcher who studied the reinforcement of asphalt. In an attempt to prevent reflection cracking, the study evaluated various types of wire mesh placed under an asphalt overlay [6]. Metal wires were also used but they were found to be susceptible to rusting, with the penetration of water [7]. Different literatures indicate that there were attempts to use nonsynthetic fibers. Even asbestos was used but it was reported to be degradable and unsuitable as a long-term reinforcement [8] and a health hazard [9, 10]. As a reinforcement material, the principal function of the fiber is to provide additional tensile strength in the resulting composite. This could increase the amount of strain absorbed during the fatigue and fracture process of the mixture [11]. The results obtained from field studies show that the addition of fiber will help produce more flexible mixtures that are also more resistant to cracking [12]. A multitude of fibers and fiber materials are being introduced regularly in the market as new applications such as polyester fiber, asbestos fiber, glass fiber, polypropylene fiber, carbon fiber, cellulose fiber, etc. [13]. Polypropylene fibers are also used as a modifier in asphalt concrete in the United States. Ohio State Department of Transportation (ODOT) has published standards for using polypropylene fibers in high-performance asphalt concrete [14]. This particular fiber provides three-dimensional reinforcement of the concrete, making it more tough and durable [15, 16]. Glass fiber shows that the addition of fiber improves properties of SMA mixtures by decreasing stability and increasing flow value as well as decreasing the voids in the mix. Addition of fiber also improves fatigue properties by increasing resistance to cracking and permanent deformation of bituminous mixtures [17]. The degree of homogeneity of dispersion of the fibers determines the strength of the resulting mixtures [18]. Here, mixture containing fiber indicates a slight increase in the optimum binder content compared to the control mix. This is similar to the addition of very fine aggregates. The bitumen quantity is dependent on absorption and surface area of the fibers and is therefore affected not only by different fiber concentrations and also by different fiber types [19].

October 2008

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

357

Saeed Ghaffarpour Jahromi and Ali Khodaii

2. LABORATORY INVESTIGATION 2.1. Materials The materials applied in the present research include 60/70-penetration grade bitumen, aggregates with gradation characterized by 12.5 mm as shown in Figure 1 (in accordance with the Pavement Guidelines of Iran), limestone mineral filler and carbon fiber. Table 1 indicates the properties of the carbon fiber.
100 Pasing Percent 80 Upper Limit 60 40 20 0 0.01 0.1 1 Particle Size, mm 10 100 Lower Limit

Figure 1. Aggregate gradation Table 1. Properties of Carbon Fiber Diameter Tensile Modulus of Elasticity Tensile Strength Failure Strain Base Temperature of Carbonization Cost 2.2. Experiments This research has followed the basic experimental approach to investigate the properties of carbon fiber-reinforced bituminous mixtures. In due process, standard laboratory tests were carried out i.e. Marshall test (ASTM D 1559), indirect tensile modulus test (IDT-ASTM D 4123-95), dynamic creep test (AS 2891.12.1), and fatigue test using indirect repeated tensile load (BSI; 2nd draft-DD ABF, method for the determination of the fatigue characteristics of bituminous mixtures using indirect tensile fatigue)[20]. Standard Marshall Apparatus and UTM-25 equipment were applied for the above-mentioned tests. Creep tests were conducted at 40C, while IDT and fatigue tests were carried out at a maximum temperature of 25C. 2.3. Sample Preparation Bituminous mixture was prepared by graded aggregates of 60/70-penetration grade bitumen and carbon fibers. The dry blending method was applied where carbon fibers were blended with hot aggregate (160oC). Here, fiber contents were selected arbitrarily and without optimization. Previous researchers however had selected fiber content between 0.3%0.8% by weight of mixture [21]. Further, in the current research, the filler content was 4% while carbon fiber contents were 0.1%, 0.2%, 0.3%, 0.4%, and 0.5% by weight of mixture. The optimum binder content for the original mixture was 5.5% by weight, while the modified mixtures were found to have an optimum binder content of 5.6%, 5.7%, 5.8%, 5.9%, and 6% corresponding to each fiber content of 0.1%, 0.2%, 0.3%, 0.4%, and 0.5%, respectively. The fiber 7.5 7.8 m 200 GPa 3.2 GPa >2% Polyacrylonitrile 1400 oC 1535 $/kg

358

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

October 2008

Saeed Ghaffarpour Jahromi and Ali Khodaii

length in the mixture was kept constant at two lengths of 12.5 mm and 20 mm, corresponding to one and one and a half times the nominal size of the aggregates used. For reinforcement, sufficient fiber length must be maintained to have an impact on material behavior. By this way, preservation of fiber length in asphalt mixtures has been considered of paramount importance. Specimens were prepared using a Marshall Compactor machine, where compaction blows were 75 for the top and bottom specimens in accordance with the standard specified for traffic-congested roads. Mixing and compaction temperatures were designated at 160C and 140C, respectively. Specimens for the creep test were cut smoothly from both ends to 50 mm height and then were capped using grease and powder. This process minimized the friction with the loading plates, ensuring uniaxial stress conditions. 3. RESULTS AND DISCUSSION 3.1. Marshall Stability In the Marshall test, the heights of the samples were measured and specimens were immersed in a water bath at 60C for 355 minutes. Specimens were removed from the water bath and quickly placed in the Marshall loading head. The Marshall apparatus deformed the specimen at a constant rate of 50.8 mm per minute. Stability was identified as the maximum load sustained by the sample. Flow was the deformation at maximum load. The stability values were then adjusted with respect to sample height. Figure 2 shows an initial increase in stability values once the fiber content increased in the mixture, but it also decreases with higher fiber contents. To increase stability, it seems that there is optimum percentage of fiber content. A large amount of fiber in the mixture produces lower contact points between aggregates, hence resulting in lower stability. The above decrease also shows that fiber length has little effect on mixture stability.
13 Stability (KN) 12 11 10 9 0 0.2 0.4 Fiber Content (%) 0.6
12.5 mm Fiber Length 20 mm Fiber Length

Figure 2. Stability and fiber content 3.2. Flow Values Figure 3 shows that an increase in fiber content decreases the flow value and as such, when the fiber content is higher than 0.30% (i.e. 0.4% and 0.5%), the flow values start to increase. The effect of fiber length on flow is more pronounced than on the stability.
3 Flow (mm) 2.5 2 1.5 0 0.2 0.4 0.6 Fiber Content (%)
12.5 mm Fiber Length 20 mm Fiber Length

Figure 3. Flow and fiber content

October 2008

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

359

Saeed Ghaffarpour Jahromi and Ali Khodaii

3.3. Voids in Total Mix (VTM) Figure 4 shows consistent results concerning the effect of fiber content on the VTM. Accordingly, an increase in fiber content and length in the mixture followed an increase in the VTM. This property is significant in so far as the pavements of hot regions are concerned because asphalt may be prone to bleeding and mounting void ratio could prevent bleeding by providing more spaces for the binder to move into. This was probably due to greater surface areas to be coated. Fiber absorbs binder and then leads to increase the voids in the mixture. In addition, mixtures with higher fiber contents might experience lower compact ability, leading to higher air void values.
8

VTM (%)

4
12.5 mm Fiber Length 20 mm Fiber Length

2 0 0.2 0.4 Fiber content (%) 0.6

Figure 4. Total voids and fiber content 3.4. Resilient Modulus Results In the indirect tensile test (ITT) resilient modulus test, a cyclic half sine load is applied along the vertical diameter of the specimen and the resulting deformations along the horizontal diameters are measured. The pulse load selected for the test is a half sine with a 125 ms pulse width and 1250 ms rest period. Using the principle of plane stress elastic theory for homogeneous and isotropic materials, the resilient modulus is calculated as:

Mr =

P (0.27 + ) t.d

(1)

where: P is the peak force (15 percent of the tensile strength), is Poissons ratio (assumed 0.35), t is the thickness of sample, and d is the horizontal diametric deformation. However, the assumption of an elastic response is reasonable if the tests are conducted in the linear viscoelastic range using a loading rate which produces low permanent deformations. Five pulses of loading were applied and the average results express the resilient modulus. An addition of carbon fibers resulted in an increase of resilient modulus in the mixture regardless of fiber content (Figure 5). The findings show that the resilient modulus increases with the increase of fiber content up to 0.4% and remains constant when the fiber content increases above 0.4% (i.e. 0.4 % and 0.5%).
4500 Resilient Modulus, MPa 4000 3500 3000
12.5 mm Fiber Length

2500 0

20 mm Fiber Length

0.2 0.4 Fiber Content %

0.6

Figure 5. Resilient modulus and fiber content

360

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

October 2008

Saeed Ghaffarpour Jahromi and Ali Khodaii

It must be noted that the increase of resilient modulus may be due to high tensile modulus of elasticity and lower ability of extension of carbon fibers. In the proposed sample, fibers have a random orientation in a different direction, which firmly bind aggregate particles inside the matrix and prevent them from moving, hence making the mixture stiffer. However, large amount of fiber leads to higher surface area that must be coated by bitumen; consequently, the aggregate particles and fiber would not be fully coated with bitumen and thereby looser and less resilient modulus is obtained. Compared to the original sample, it was noted that the resilient modulus increased significantly: about 22% with the inclusion of 0.4% fiber content.
3.5. Dynamic Creep Test

A repetitive uniaxial compression test was used to evaluate the permanent deformation of the mixtures. In this test, a constant repetitive load is applied and the resulting permanent deformation is measured in relation to time (pulse counts). The pulse form selected is half sine, with a duration of 200 ms and a rest period of 800 ms; hence one pulse period is 1s. Two axial LVDTs are used to measure the axial displacement and the average of the two gives the average permanent deformation. Normally the permanent strain (accumulated permanent strain) is reported, which is computed as average permanent deformation divided by specimen thickness. Presented in the form of permanent strain under repeated load of 200 kPa, the creep test results indicated that the fiber content affects the creep properties of the bituminous mixtures. As illustrated in Figure 6, increasing fiber contents evidently decreases the permanent strains at 6500 cycles. Generally, the creep test showed that the addition of small amount of carbon fiber into the original bituminous mixture reasonably improves its deformation behavior, compared to the original bituminous mixture, which could deform more easily under the same load and temperature conditions. In contrast, higher amount of carbon fiber into the mixture may not have beneficial effect and might deteriorate its deformation properties due to the same reasons we mentioned the previous sections.
14000 Permanent Strain ( )

11000

8000
12.5 mm Fiber Length 20 mm Fiber Length

5000 0 0.2 0.4 Fiber Content (%) 0.6

Figure 6. Permanent strain at 6500 cycles and fiber content

3.6. Fatigue Test

An indirect tensile test with diametric compression load was used to evaluate the fatigue resistance of the mixtures. In this test, a constant repetitive load is applied and the resulting vertical deflection is measured in relation to time (pulse counts). In the indirect tensile fatigue test (ITT fatigue), the fatigue life of a material is defined as the number of load cycles to specimen fracture. Normally the result is reported as stress or strain versus the resulting fatigue life and commonly referred as SN curve. Here it is reported the number of cycle at which specimen fracture occurs. Tests were conducted using constant repeated stress of 350 kPa with half sine pulse of 5 Hz frequency, 150 ms loading period and 50 ms rest period. Figure 7 illustrates the results obtained through laboratory fatigue testing (repeated load indirect tensile test). It indicates that the use of carbon fibers significantly improves fatigue performance. Further, the mixtures reinforced with carbon fiber are considerably more fatigue resistant than the original mixes.

October 2008

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

361

Saeed Ghaffarpour Jahromi and Ali Khodaii

10000 Fatigue Life

8000

6000
12.5 mm Fiber Length 20 mm Fiber Length

4000 0 0.2 0.4 Fiber Content (% ) 0.6

Figure 7. Fatigue and fiber content

With the addition of 0.1%, 0.2%, and 0.3% carbon fiber, fatigue life increased to about 28.2%, 37.2%, and 44.4%, respectively. This was probably due to carbon fibers that were distributed in different directions of the bituminous matrix, which resisted the shear displacement and prevented aggregate particles from any movement, thus increasing fatigue life by efficiently delaying initial cracks and its propagation.
3.7. Correlation Between Resilient Modulus, Fatigue Life and Permanent Deformation

From the results mentioned in previous section and Figures 5, 6, and 7, it can develop correlation between resilient modulus, fatigue life, and permanent deformation. Figure 8 shows that fatigue life of reinforced bituminous mix with fiber has good correlation with resilient modulus properties of the mixture. It also indicates the linear relationships fitted for 12.5 mm and 20 mm long fibers with R2 values of 0.93 and 0.99, respectively. It was observed that, was an inverse correlation between permanent deformation and stiffness. The increase in resilient modulus properties of the specimens was followed by a decrease in the permanent deformation as shown in Figure 9. It was founded that these was a significant linear relationship between resilient modulus and permanent deformation with R2 values of 0.97 for 20 mm and 0.85 for 12.5 mm long fibers.

9000
y = 3.7488x - 6130.3 R2 = 0.9965

8000 Fatigue Life 7000

6000
12.5 mm Fiber Length 20 mm Fiber Length

5000 3200 3500 3800 Resilient Modulus / Stiffness (MPa) 4100

Figure 8. Fatigue life and stiffness

362

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

October 2008

Saeed Ghaffarpour Jahromi and Ali Khodaii

14000 Permanent Strain ( )

12000 10000 8000 6000 3200

y = -5.7009x + 31234 R2 = 0.9735 12.5 mm Fiber Length 20 mm Fiber Length

3500 3800 Resilient Modulus / Stiffness (MPa)

4100

Figure 9. Permanent deformation at 6500 cycles and stiffness

4. CONCLUSION

The use of carbon fiber showed consistency of results in the present study. It was observed that the addition of fiber favorably affects the properties of bituminous mixtures by increasing its stability and voids and decreasing the flow value. As such, it can be said that carbon fiber has the potential to improve structural resistance to distress occurring in road pavement due to traffic loads. Further, addition of fiber improves fatigue life and permanent deformation of bituminous mixtures by improving mix stiffness. Compared to the control mixture, the fiber content of 0.4% by weight of total mix resulted in highest performance in terms of stiffness, resistance to permanent deformation and fatigue; however, some mechanical properties of the same mix may be compromised when the fiber content exceeds 0.4% level. Since the length of the fiber is a critical factor affecting the performance of carbon fiber modified asphalt mixtures it must be ensured that individual fibers keep their linear configuration intact after the mixing process. To achieve these improvements, proper attention must be paid to ensure that the fibers are uniformly dispersed in the mixture. The results indicated that 20 mm fiber shows better mechanical behavior than the longer fibers, which may lead to a balling phenomenon in the mix and there by lose its beneficial effects. Modified carbon fiber mixtures show significant increases in stiffness, fatigue life, and rutting resistance, indicating good correlation between stiffness and fatigue life and permanent deformation. This increase in stiffness is directly related to the addition of fiber, and its contents and properties. However, due to the advanced technology involved in their production, use of carbon fibers can be expensive in developing countries like Iran and hence not economical, but they will become feasible once the chemical plant in Mahshahr starts production.
REFERENCES
[1] [2] [3] [4] [5] [6] [7] [8] American Concrete Institute, Report on Fiber Reinforced Concrete. ACI 544.1 R-96, 2002. P.S. Song, S. Hwang, and B.C. Sheu, Strength Properties of Nylon and Polypropylene-Fiber-Reinforced Concretes, Cement and Concrete Research, 35(8)(2005), pp. 154650. W. Yao, J. Li, and K. Wu. Mechanical Properties of Hybrid Fiber-Reinforced Concrete at Low Fiber Volume Fraction, Cement and Concrete Research, 33(1)(2003), pp. 2730. Y. Choi and R.L. Yuan, Experimental Relationship between Splitting Tensile Strength and Compressive Strength of GFRC and PFRC, Cement and Concrete Research, 35(8)(2005), pp. 15871591. A.M. Alhozaimy, P. Soroushian, and F. Mirza, Mechanical Properties of Polypropylene Fiber Reinforced Concrete and the Effects of Pozzolanic Materials, Cement and Concrete Composites, 18(2)(1996), pp. 8592. E. Zube, Wire Mesh Reinforcement in Bituminous Resurfacing, Highway Research Record, Bulletin 131, 1956, pp. 118. E. Tons and E.M. Krokosky, A Study of Welded Wire Fabric Strip Reinforcement in Bituminous Mixtures, Concrete Resurfacing. Proceedings of the Association of Asphalt Paving Technologists, 29(1960), pp. 4380. H.W. Bushing and J.D. Antrim, Fiber Reinforcement of Bituminous Mixtures, Proceedings of the Association of Asphalt Paving Technologists, 37(1968), pp. 629659.

October 2008

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

363

Saeed Ghaffarpour Jahromi and Ali Khodaii

[9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

J.H. Kietzman, The Effect of Short Asbestos Fibers on Basic Physical Properties of Asphalt Pavement Mixes, Highway Research Board, Bulletin No. 270, 1960, pp. 119. C. P. Marais, The Use of Asbestos in Trial Sections of Cap-Graded Asphalt and Slurry Seals, Proceedings of the Third Conference on Asphalt Pavements for South Africa, Durban, 1979, pp. 172175. S.F. Brown, R.D. Rowlett, and J.L. Boucher, Asphalt Modification, Proceedings of the Conference on the United States Strategic Highway Research Program; Sharing the Benefits. London: Thomas Telford, 1990, pp. 181203. Yi. Jiang and Rebecca S. McDaniel. Application of Cracking and Seating and Use of Fibers to Control Reflection Cracking, Transportation Research Record, 1388(1993), pp. 150159. J.P. Serfass and J. Samanos Fiber-Modified Asphalt Concrete Characteristics, Applications and Behavior, Journal of the Association of Asphalt Paving Technologists, 65(1996), pp. 193230. ITEM 400HS, Standard Specification for Asphalt Concrete-High Stress Using Polypropylene Fibers. Ohio Department of Transportation, Construction and Materials Specifications, 1998. A. Noumowe, Mechanical Properties and Microstructure of High Strength Concrete Containing Polypropylene Fibers Exposed to Temperatures up to 200 o C, Cement and Concrete Research, 35(11)(2005), pp. 21922198. S. Singh, A. Shukla, and R. Brown Pullout Behavior of Polypropylene Fibers from Cementations Matrix, Cement and Concrete Research; 34(10)(2004), pp. 19191925. M. Abdelaziz and R. K. Mohamed Fatigue and Deformation Properties of Glass Fiber Reinforcement Bitumen Mixes, Journal of the Eastern Asia Society for Transportation Studies, 6(2005), pp. 9971007. D.R. Mills and T. Keller Jr., The Effectiveness of Synthetic Fiber-Reinforced Asphalt Concrete Overlays in Delaware. Dover, Delaware: Delaware Department of Transportation, 1982, pp 46. J.W. Button and R.L. Lytton, Evaluation of Fabrics, Fibers and Grids in Overlays, Sixth International Conference on the Structural Design of Asphalt Pavements, Ann Arbor, Michigan, 1987. British Standards Draft for Development, 2nd Draft, Method for the Determination of the Fatigue Characteristics of Bituminous Mixtures Using Indirect Tensile Fatigue, 1997. C. M. Aren, Investigation of the Properties of Fiber Modified Asphalt Mixtures, Thesis, Master of Science in Civil Engineering, Michigan Technological University, 2000.

364

The Arabian Journal for Science and Engineering, Volume 33, Number 2B

October 2008