Impact Energy Consumption of High-Volume Rubber Concrete with Silica Fume
Adding rubber to concrete aims to solve the environmental pollution problem caused by waste rubber and to improve the energy absorption and impact resistance of concrete. In this paper, recycled rubber particles were used to replace fine aggregates in Portland cement concrete to combine the elasticity of rubber with the compression resistance of concrete. Fine aggregates in the concrete mixes were partially replaced with 0%, 20%, 40%, and 60% rubber by volume, and the cement in the concrete mixes was replaced with 0%, 5%, and 10% of silica fume by mass. The properties of the concrete specimens were examined through compressive strength, splitting tensile strength, flexural loading, and rebound tests. Results show that the compressive strength of concrete and the splitting tensile strength decreased to 11.81 and 1.31 MPa after adding silica fume to enhance the strength 37.8% and 23.7%, respectively, and the dosage of rubber was 60%. With the addition of rubber, the impact energy of rubberized concrete was 2.39 times higher than that of ordinary concrete, while its energy absorption capacity was 9.46% higher. The addition of silica fume increased its impact energy by 3.06 times, but the energy absorption capacity did not change significantly. In summary, the RC60SF10 can be used on non-load-bearing structures with high impact resistance requirements. A scanning electron microscope was used to examine and analyze the microstructural properties of rubberized concrete.
Widespread automobile use has increased the number of discarded rubber tires, and the low recovery rate of waste tires has aggravated environmental pollution. Burning, burying, or piling up waste tires in landfills is not only wasteful but also harmful to the environment [1, 2]. Given the dense connection among concrete aggregates, wastes such as rubber, plastic, broken ceramic, broken glass, and recycled aggregate can be added to concrete [3–6]. Among concrete materials, rubberized concrete exhibits good tenacity and energy absorption properties. Although rubberized concrete is not recommended for high-strength structures, its high tenacity, impact resistance, and sound absorption features have prompted studies on its application in sidewalk pavements , railway subgrades and sleepers , road construction barriers , and sound absorption structure . Similarly, rubber can also be added to a brick masonry structure. Al-Fakih et al. [11, 12] studied the addition of rubber and fly ash to a masonry structure to meet the requirements of ASCE C90-09.
The fluidity and strength of concrete decrease as rubber content increases [13–16]. Eldin and Senouci  found that when all coarse aggregates in concrete are replaced with rubber, up to 85% of compressive strength and up to 50% of splitting tensile strength are decreased; furthermore, when all fine aggregates in concrete are replaced with rubber, up to 65% of compressive strength and up to 50% of splitting tensile strength are decreased. Khatib and Bayomy  summarized previous studies and developed the characteristic function SRF, which quantifies the reduction in strength for rubberized concrete mixes and is useful in describing the strength of concrete with high rubber content. Khatib also suggested that rubber contents must not exceed 20% of the total aggregate volume. Notably, although a large reduction in the strength of rubberized concrete is observed, such concrete shows good tenacity. Gerges et al.  and Atahan and Yücel  examined the properties of rubberized concrete at failure and found that the rubberized concrete does not demonstrate brittle failure but rather a ductile failure with good impact resistance.
The cause of the strength decreases of rubberized concrete is the weak bond between rubber and cement, which limits the development of concrete with high rubber content. Therefore, methods that limit the strength decreased of concrete while increasing rubber content have been explored. Xue and Cao  and Chen et al.  studied the impact of polyvinyl alcohol (PVA) solution on improving the connection between rubber and cement in concrete. Findings indicate that the strength of modified rubber concrete was not obviously improved by adding PVA solution, but good tenacity and impact resistance of the concrete were achieved. Nanosilica, silica fume, and slag have been used to remedy the negative effect of rubber by improving the interfacial transition zone between rubber and cement paste [21–25]. Adamu et al.  and Mohammed and Adamu  found that nanosilica could improve the refinement of pore structure, strengthen the interface transition zone (ITZ), and enhance the bonding effect, thereby partially reducing the strength loss of rubber concrete. Silica fume is the best choice for economic and practical considerations. Gupta et al.  found that the use of silica fume in rubberized concrete mixtures not only reduces the amount of cement but also increases compressive strength and dynamic and static moduli. The use of silica fume also in these instances reduces water permeability and chloride ion diffusivity.
The impact resistance of concrete added with plastic fiber and steel fiber is increased by 10–15 times [27, 28]. Ramakrishna and Sundararajan  tested cement mortar slabs reinforced with natural fibers, such as coir, sisal, jute, and hibiscus, and showed that the addition of such fibers increases the impact resistance of the slabs by 3–18 times. Gupta et al.  found that the impact resistance of concrete with 25% rubber content is increased by five times. Zongcai et al.  improved the test method for determining the impact resistance of fiber-reinforced concrete beams to dynamic loads and showed that the layered steel fiber-reinforced concrete exhibits impact resistance, which was measured as the number of blows from first cracking to failure, six times that of the reference concrete. In Zhang’s  test, impact energy was unchanged and the impact energy of steel fiber-reinforced concrete beams decreased with increased falling height of the ball and increased with increased weight of the falling ball.
As mentioned above, the addition of silica fume limits the strength decrease of rubberized concrete. This feature prompts further study on the impact resistance and energy absorption properties of rubberized concrete with high rubber content. In this study, the water-cement ratio of concrete was maintained at 0.5. The fine aggregates in the concrete mixes were partially replaced with 0%, 20%, 40%, and 60% of rubber by volume, and the cement in the concrete mixes was replaced with 0%, 5%, and 10% of silica fume by mass. Three basic mechanical properties of concrete—unit weight, compressive strength, and splitting tensile strength—were tested. The relationship between compressive strength and splitting tensile strength was established, and the failure modes under compression were analyzed. The impact resistance of concrete beams was measured in the falling ball impact test, and the energy absorption ability of concrete test cubes was measured in the rebound test. The developing course and mechanism of beam cracks were elaborated, and the relationship between impact resistance and energy absorption ability was analyzed and established.
2. Test Materials and Specimen Preparation
2.1. Raw Materials
Ordinary Portland cement (P.O 42.5) from Huainan Bagongshan Cement Factory was used; its physical parameters and chemical composition contents are shown in Table 1. Coarse aggregates were graded crushed stone with a maximum particle size of 16 mm and a specific gravity of 2.63. Fine aggregates were medium sand collected from Huaihe River with a fineness modulus of 2.6 and a specific gravity of 2.41. The water used in the test was tap water obtained from the laboratory. Rubber particles with a particle size of 20–40 mesh and a specific gravity of 1.03 were ordered from the Internet. Silica fume was produced in a factory in Sichuan Province, and its composition is shown in Table 1. The particle size distribution of the silica fume and cement measured using the Mastersizer laser particle analyzer is presented in Figure 1.
2.2. Mix Ratio
Concrete mixes containing 0%, 20%, 40%, and 60% of partial replacement of fine aggregates with rubber particles were prepared. A water-cement ratio of 0.5 was adopted to prevent rubber particles from floating during vibration due to the addition of a water-reducing agent. The cement in the concrete mixes was replaced with 0%, 5%, and 10% of silica fume by mass. The mix ratio and unit weight of concrete mixes are shown in Table 2. RC20SF5 indicates that the rubber content in plain concrete is 20%, and the silica fume content in cement is 5%.
First, the dry concrete mixture was stirred for 3 min, the weighed water was evenly added into the mixture, and the mixture was stirred again for approximately 2 min. Second, the fully mixed concrete mixture was placed into plastic molds with dimensions of 100 × 100 × 100 mm3, 100 × 100 × 400 mm3, and 150 × 150 × 150 mm3. Third, the mixtures together with the molds were shaken on a table vibrator for 20–30 s to allow full compaction, and the surface of the specimen was troweled. Fourth, the surface of the mold was covered with a thin film to prevent water evaporation, and the specimens were cured under room temperature. Finally, on the following day, the specimens were removed from the molds and placed in a water tank containing saturated Ca(OH)2 solution for water bath curing for 7 and 28 days.
3. Test Method
3.1. Compressive Strength and Splitting Tensile Strength
The 7-day and 28-day compressive strength and splitting tensile strength of the specimens were tested in groups with each group containing three specimens. Loading speed was measured in accordance with the provisions of the Standard for Test Method of Mechanical Properties on Ordinary Concrete (GB/T 50081-2002). The testing device is shown in Figure 2.
3.2. Impact Resistance under Flexural Loading
The schematic of the drop-weight impact test for concrete beams is shown in Figure 3. In the test, the drop hammer periodically impacted the concrete beam at a fixed drop height and position. The change in the resistance value of the strain foil at the bottom of the specimen was measured by a dynamic resistance strain gauge. When a sudden change or overload sign was observed on the strain gauge, which indicates the generation of a microcrack at the bottom surface of the concrete beam specimen, the accumulated number of blows by this moment was recorded and expressed as N1, and the impact energy at this moment was Ef,1. The test was continued until the beam specimen was forced to separate. The total number of blows and impact energy at failure were recorded as N2 and Ef,2, respectively. The relation between impact energy and the number of blows is as follows:where m is the hammer weight, which is 1.4 kg; h is the drop height of the hammer, which is 300 mm; is the local acceleration of gravity, which is 9.81 m/s2; N1 is the number of blows at first crack; and N2 is the number of blows at failure. The air resistance in the free fall of the drop hammer was ignored.
To allow even spreading of the impact force of the drop hammer to the beam and ensure the fixed position of the steel gasket, we evenly applied Vaseline under the steel gasket whose side length and thickness were 100 and 6 mm, respectively. The strain foil used in the test was 8 cm long with a sensitivity coefficient of 2.0. The foil was connected with a CS dynamic resistance strain gauge and a TST3406 dynamic test analyzer through wires. The moment when the first microcrack was produced on the beam bottom was determined according to the resistance value and waveform diagram displayed on the analyzer. Equations should be provided in a text format, rather than as an image.
3.3. Rebound Test
In this test, the rebound height of the steel ball falling on the concrete test cube whose side length was 150 mm was recorded using a NAC Memrecam HX-5E high-speed camera. The test was conducted in groups with each group containing three cubes, and the average value of the rebound height in each group was calculated. The test device is shown in Figure 4. The initial energy and energy after rebound were recorded as Er,1 and Er,2, respectively. The relation between energy and falling height is described as follows:where m is the mass of the steel ball, which is 0.5 kg; is the acceleration of gravity, which is 9.81 m/s2; h1 is the initial falling height of steel ball, which is 1.0 m; and h2 is the height of the first rebound of the steel ball recorded by the high-speed camera.
The rebound energy absorbed by the test cube with different rubber and silica fume contents is described as Er. The formula is Er = Er,2 − Er,1. The influence of air resistance was ignored.
4. Test Results
4.1. Unit Weight
The unit weights of concrete mixes after 28 days of curing are shown in Table 2. The fine aggregates in the concrete mix were replaced with rubber by volume; thus, the unit weight of the concrete mix decreased with increased replacement ratio. When the silica fume content in cement was 0% and the fine aggregates were replaced by rubber particles by 0% to 60%, the unit weight of concrete was reduced from 2,317 kg/m3 to 2,070 kg/m3. This reduction is due to the similarity between the particle size of rubber and the fine aggregate, but the specific gravity of rubber is smaller than that of fine aggregate. The cement in the concrete mix was replaced with silica fume by mass. The particle size of silica fume was smaller than that of cement. Hence, it can fill the gap of cement paste. The apparent density of concrete increased slightly as the silica fume replacement ratio increased. When the rubber content was 40% and the cement was replaced with silica fume by 0%, 5%, and 10%, the unit weights of concrete mixes were 2,120, 2,137, and 2,157 kg/m3, respectively. In addition, the rubber content is 60%, and the weight increased with the addition of silica fume concrete, along with the light quality of NC60SF0.
4.2. Compressive Strength
The average compressive strength of each group of specimens whose side length was 100 mm is presented in Figure 5, showing an increasing trend of compressive strength as curing age increased. When the silica fume content was 0% and the rubber contents were 0%, 20%, 40%, and 60%, the growth rates of compressive strength with age were 1.33, 1.23, 1.17, and 1.13, respectively. However, the compressive strength of concrete decreased as rubber content increased. When the silica fume content was 0% and the rubber contents were 0%, 20%, 40%, and 60%, the compressive strengths of concrete were 38.92, 23.15, 13.49, and 11.81 MPa, respectively, and reductions of 41%, 65%, and 70% compared with the reference group NC0SF0 were observed. The interfacial transition zone between the cement paste and the rubber aggregate was often porous due to the side wall effect and is the weakest zone in concrete, thus leading to strength decrease. Ozbay et al.  confirmed that the specific gravity of rubber was lower than that of fine aggregate, which leads to a decrease in density and compressive strength.
Similarly, compressive strength increased with increased silica fume content, as shown in Figure 5. When the rubber content was 0% and the silica fume contents were 0%, 5%, and 10%, the compressive strengths of concrete mixes were 38.92, 48.42, and 55.13 MPa, respectively. Increases of 24.4% and 41.6% compared with the reference group were also observed. Similarly, when the silica fume content was 10% and the rubber contents were 0%, 20%, 40%, and 60%, the growth rates of concrete compressive strength with age were 1.75, 1.59, 1.29, and 1.27, respectively. The compressive strength of rubberized concrete cube greatly improved with increased silica fume content, and the compressive strength of concrete in the later curing period also improved by silica fume. This result was due to the stronger activity, smaller particle size, and larger specific surface area of silica fume compared with those of cement. When water was added to the mixture of silica fume and cement, the cement initially reacted with water (i.e., hydration) and SiO2 in the silica fume reacted with Ca(OH)2 precipitated in hydrated cement to generate calcium silicate hydrate colloid, which filled the voids around the hydrated cement particles and improved the compressive strength of the concrete.
4.3. Failure Mode
The failure modes of concrete test cubes under compression are shown in Figure 6.
As shown in Figure 6, after the NC0SF5 specimen reached ultimate failure load under compression, the upper and lower surfaces of the cube broke and peeled off from the edge due to the hoop effect. However, Figure 6 demonstrates that after the RC40SF5 specimen reached the ultimate failure load, vertical cracks formed on the cube, although the cube was still intact as a whole and no breakage or crush was apparent. This condition occurred because rubber as an elastomer in concrete can inhibit the development of microcracks in concrete and enhance the resistance of concrete against compressive deformation, thus preventing most microcracks from evolving into macrocracks. Thus, adding rubber can change the failure mode of concrete from brittleness to ductility. Gerges et al.  and Atahan and Yücel  also confirmed this view.
4.4. Splitting Tensile Strength
The splitting tensile strengths of concrete mixtures with three different silica fume contents changed with rubber contents in different curing ages, as shown in Figure 7. When the content of silica fume remained unchanged, the splitting tensile strength of concrete decreased as the rubber content increased. For example, when the content of silica fume was 5%, the 7- and 28-day splitting tensile strength of concrete decreased from 2.02 MPa to 1.43 MPa and from 3.36 MPa to 1.45 MPa, respectively. In specimens with the same rubber content, the splitting tensile strength increased as the silica fume content increased. For example, when the rubber content was 40%, the 7- and 28-day splitting tensile strength increased from 1.52 MPa to 1.79 MPa and from 1.66 MPa to 1.93 MPa, respectively.
Figure 7 demonstrates that the splitting tensile strength of concrete increased as curing age increased, but this effect was not evident in specimens with high rubber content. For example, when the silica fume contents were 0%, 5%, and 10% with increased rubber content, the splitting tensile strength of concrete decreased from 1.67, 1.66, and 1.60 to 1.04, 1.02, and 1.06, respectively. Adamu et al.  believed that the weak adhesion between rubber particles and cement matrix, as well as the increase and weakness in the thickness of the ITZ, led to the failure of the gel powder cementitious matrix when concentrated load was applied; subsequently, the tensile strength was reduced. Gergrs et al.  confirmed that the concrete without rubber was splitting into two parts, and the concrete with rubber was not substantially damaged after the splitting failure.
4.5. Tension-Compression Ratio
The ratio of the splitting tensile strength to the compressive strength of the test cube is the tension-compression ratio, an important index reflecting the brittleness of concrete. The 28-day tension-compression ratio of the specimens is shown in Figure 8. When the silica fume content was 0% and the rubber contents were 0%, 20%, 40%, and 60%, the tension-compression ratios of rubberized concrete were 0.079, 0.115, 0.123, and 0.111, respectively, which were all greater than that of the reference group whose tension-compression ratio was 0.079. This result was due to the addition of rubber improving the plastic deformation ability of concrete and reducing the brittleness of concrete. This result could also be confirmed by the compression failure mode of the rubberized concrete. As shown in Figure 6, when the rubberized concrete test cube reached the limit load, the cube could still bear a certain load and display good integrity and ductility.
As the amount of silica fume replacement increased, the tension-compression ratio of concrete decreases considerably. Silica fume has the advantages of filling the gap between cement and rubber in the form of small particles and promoting cement hydration. Thus, the addition of silica fume increases the density of rubberized concrete, leading to the increased brittleness and reduced plasticity of the concrete. When the rubber content was 0%, the tension-compression ratios of the concrete mixes with silica fume contents of 0%, 5%, and 10% were 0.079, 0.069, and 0.063, respectively. When the rubber content was 20% with silica fume contents of 0%, 5%, and 10%, the tension-compression ratios of concrete mixes were 0.115, 0.096, and 0.082, respectively. However, when the rubber contents were 40% and 60%, the tension-compression ratio of concrete showed a downward trend as silica fume content increased, but the decreasing rule was not evident. This finding may be attributed to high rubber content or an improper experimental operation.
4.6. Impact Resistance under Flexural Loading
Figure 9 shows the relationship between rubber content and impact energy under impact loading.
Figure 9(a) shows the impact energy exposed to the rubberized concrete beam when a microcrack occurred for the first time under impact load. The impact energy of the beam increased as rubber content increased. When the rubber contents were 0%, 20%, 40%, and 60%, the impact energies of the beam at first crack were 21.6, 22.6, 28.8, and 47.3 J, respectively. Silica fume addition enhanced the impact energy at the first crack. When the silica fume content was 10% and the rubber contents were 0%, 20%, 40%, and 60%, the impact energies at first crack were 23, 24.7, 35, and 56.6 J, respectively.
Figure 9(b) shows the impact energy exposed to the rubberized concrete beam when the beam was forced to break under the impact load. Similarly, the impact energy of the beam increased as rubber content increased. When the rubber contents were 0%, 20%, 40%, and 60%, the impact energies of the beam at failure were 34.0, 42.2, 50.4, and 81.3 J, respectively. Silica fume addition enhanced the impact energy of the beam at failure. When the rubber content was 60% with silica fume contents of 0%, 5%, and 10%, the impact energies at failure were 81.3, 90.6, and 104.0 J, respectively.
4.7. Failure Process
Ordinary concrete is a brittle material. The difference between the impact energy at first crack and at failure is not evident, and the total impact energy consumption is small. Thus, judging when a microcrack appears in the specimen for the first time is important. When the strain foil breaks, the resistance measured by the dynamic strain gauge exceeds the measuring range, and the number of blows delivered by the falling ball is recorded as N1, which is also the basis for judging the occurrence of the first crack. Figure 10 shows a part of the crack development in the RC40SF10 specimen under impact load. Figure 10(a) shows the occurrence of the first microcrack and the fracture of the strain foil at 150x magnification. The rubberized concrete beam was damaged at this time. Figure 10(b) shows that as the number of blows continued to increase, and the microcrack on the side surface of the beam began to extend upward along the bottom. At this time, the beam damage continued to expand, and the crack width was small. As the test continued, a second crack appeared on the beam, as shown in Figure 10(c). The second crack was adjacent to the first crack and appeared discontinuous from the surface. Both cracks continued to extend upward, with a small increase in crack width. As illustrated in Figure 10(d), as the beam continued to be subjected to impact load, the cracks continued to develop upward until they penetrated through the side surface of the beam, and the crack width increased. At this time, the fracture surface was basically formed. After another blow, the rubberized concrete beam would break.
The crack of ordinary concrete developed rapidly, and the crack width was larger than that of the rubberized concrete. Thus, the number of blows borne by the ordinary concrete was less than that of the rubberized concrete. This finding was mainly attributed to the addition of rubber particles that could be bonded with surrounding voids to form a structural deformation center with certain strength under the binding effect of cement. The elasticity advantage of the rubber also enabled the concrete to bear, buffer, and absorb some of the stress when the concrete was subjected to impact load. The rubber particles in the concrete mixture could also eliminate stress concentration around the voids and restrict the generation and development of microcracks, thus improving the impact resistance of the concrete beam.
4.8. Rebound Test
The energy absorbed by concrete depends on the content of both rubber and silica fume, as shown in Figure 11.
The energy absorbed by rubberized concrete gradually increased as rubber percentage increased. For example, when the silica fume content was 0% and the rubber contents were 0%, 20%, 40%, and 60%, the energy absorbed by concrete was 4.17, 4.41, 4.5, and 4.52 J, respectively. Compared with rubber, silica fume exhibited no evident influence on the energy absorption effect of concrete. When the rubber content was 0%, the energies absorbed by the specimens with silica fume contents of 0%, 5%, and 10% were 4.17, 4.14, and 4.11 J, respectively. When the rubber content was 60%, the energies absorbed by the specimens with silica fume contents of 0%, 5%, and 10% were 4.52, 4.51, and 4.50 J, respectively. Rubber is compressible and bendable, so rubber particles as a replacement for aggregates can add ductile mechanism to conventional concrete. In addition, the low rigidity of rubber instills relatively high flexibility in the rubber-cement composite material. Thus, the energy absorption ability of rubberized concrete is stronger than that of ordinary concrete.
4.9. Relation between Impact Resistance and Energy Absorption Ability
Figure 12 shows the relationship between the impact resistance and energy absorption ability of concrete.
The impact resistance increased as the energy absorbed in the rebound test increased. This finding indicates that high rubber content corresponds to improved impact resistance and energy absorption ability of concrete. The fitted equations of the two different forms of energy are shown in Table 3. The equation differed according to the silica fume content, but the correlation coefficient R2 showed a good correlation between impact resistance and energy absorption ability.
5. Microstructure Analysis
The microstructures of RC40SF10 concrete observed by SEM are shown in Figure 13. As shown in Figures 13(a) and 13(b), microcracks and microgaps occurred between rubber and cement paste, resulting in considerable reduction in the unit weight and the strength of rubberized concrete. As illustrated in Figures 13(c) and 13(d), the bonding surface between rubber and cement paste was weak, but the addition of silica fume could improve a part of this problem. As presented in Figure 13(d), the silica fume surrounded with cement paste filled a part of the weak bonding surface. Therefore, the impact resistance of the rubberized concrete beam improved.
The effects of different contents of silica fume and rubber on the unit weight, compressive strength, splitting tensile strength, impact resistance, and energy absorption of concrete were experimentally studied. The following conclusions were derived from this study:(1)The replacement of fine aggregates with an equal volume of rubber particles considerably reduces the unit weight of concrete, whereas the addition of silica fume slightly increases the unit weight of concrete.(2)The compressive strength and splitting tensile strength of concrete decrease as rubber content in the concrete mixture increases, but the addition of silica fume can partially compensate for the strength decreased. With increased rubber content, the concrete behaves in a ductile manner rather than in a brittle one at ultimate failure under compression.(3)With increased rubber content, the impact resistance of the rubberized concrete beam at first crack and at failure is 2.19 and 2.39 times higher than that of the reference group, respectively. Silica fume addition increases the impact energy exposed to concrete at first crack and at failure by 2.62 and 3.06 times, respectively. The width and speed of crack development in the rubberized concrete beam are smaller than those in the reference group, indicating that rubber has the effect of blocking crack development.(4)Similarly, with the increase in rubber content, the energy absorbed by the rubberized concrete in rebound test increases by 9.46% compared with the reference group. However, the addition of silica fume does not affect the energy absorption of concrete.(5)The microstructure analysis result proved that the addition of silica fume can promote cement hydration and improve the weak bond between rubber and cement paste.
The datasets generated and analyzed during the current study are available from the corresponding author on reasonable request.
Conflicts of Interest
The authors declare no conflicts of interest.
This work was financially supported by the National Natural Science Foundation of China (51728201) and Natural Science Foundation of Anhui University (KJ2018A0074).
A. Al-Fakih, B. S. Mohammed, M. S. Liew, and W. S. Alaloul, “Physical properties of the rubberized interlocking masonry brick,” International Journal of Civil Engineering and Technology, vol. 9, no. 6, pp. 656–664, 2018.View at: Google Scholar
A. Al-Fakih, B. S. Mohammed, M. S. Liew et al., “Mechanical behavior of rubberized interlocking bricks for masonry structural applications,” International Journal of Civil Engineering and Technology, vol. 9, no. 9, pp. 185–193, 2018.View at: Google Scholar
D. Zongcai, L. Jinahui, S. Hongjun, and F. Zhi, “Flexural impact performance of steel and glass fiber reinforced concrete beams,” Journal of Highway and Transportation Research and Development, vol. 22, no. 6, pp. 24–27, 2005.View at: Google Scholar