Effect of different PVA and steel fiber length and content on mechanical properties of CaCO₃ whisker reinforced cementitious composites

M. Cao, C. Xie*, L. Li, M. Khan

School of Civil Engineering, Dalian University of Technology, Liaoning (People’s Republic of China)

*xie_chaopeng@mail.dlut.edu.cn

1. INTRODUCTIONTOP

The inherent brittle behavior of cementitious composites leads to a low tensile strength, toughness, impact resistance and cracking (1–3). Therefore, fibers are added to enhance the crack resistance, tensile strength and ductility of fiber reinforced cementitious composites(FRCC)(3,4).The hybrid fiber system with multi-scale characteristics is widely used to reinforce cementitious composites. In comparison with normal FRCC, hybrid fibers can make the effect of toughening and strengthening at multi-level (5,6). Usually, long fibers and short fibers are mixed together in FRCC for controlling cracking at different levels (7–11). However, the metallic fiber, polymeric fiber or natural fiber only can restrict macroscopic or mesoscopic cracks in cementitious composites. Therefore, it is necessary to add a low cost microscale fiber to arrest cracks at microscale.

CaCO₃ whiskers (CW) is an inorganic microfiber like needle having characteristic of high strength, high elastic modulus, low price (approximately U.S.$230 per ton), large aspect ratio and better crack resistance at micro level (12–13). Cao et al. (12) added CW into cement-based materials firstly and CW could effectively restrict the generation and propagation of cracks at microscopic level. Cao et al. (13) also reported that adding CW into cement mortar not only improved the compressive and flexural strength of cement mortar but also enhanced the flexural toughness. Furthermore, the microstructure of CW could fill the pores of cement mortar and make it denser. On the other hand, the behavior of whisker pullout, crack deflection, whisker bridging and whisker breakage could increase flexural performance. Thus, CW may be one of the ideal reinforced material at microscale in cementitious composites. Zhang et al. (14) and Cao et al. (15–18) studied CW into PVA-steel fibers reinforced cementitious composites to form a multiscale hybrid fiber reinforced cementitious composites (MHFRCC). The result showed that both compressive strength and flexural performance of MHFRCC are significantly improved showing the multiple cracking behavior.

Flexural behavior is an important mechanical behavior of FRCC influenced by many material parameters, such as strength, elastic modulus of matrix, fiber length, type, content, dispersion and interaction between fiber and matrix (11, 19). Therefore, large numbers of researchers have focused on studying the material parameters of FRCC to minimize cost and maximize flexural behavior of FRCC. Kim et al. (19) compared the flexural performance of four fibers (twisted steel, hook steel, polyethylene spectra and polyvinyl alcohol) in reinforced cementitious composites. The matrix consisting of twisted steel fibers showed best flexural behavior of equivalent flexural strength, flexural toughness and multiple cracking behaviors. However, matrix with PVA fiber presented the worst flexural performance as compared to that of other fiber reinforced cementitious composites without PVA fiber. The result indicated that different fiber type and fiber length have prominent impact on flexural behavior of FRCC. Banthia and Soleimani (7) investigated the flexural performance of hybrid fiber reinforced concrete. The result showed that combination of steel and polypropylene fiber reinforced concrete (SPFRC) has higher post-cracking strength (PCS) value than that of single steel fiber reinforced concrete (SFRC). However, the addition of mesophase carbon fiber to SPFRC showed a higher PCS value as compared to that of SPFRC. It was found that there existed synergistic effect among steel fiber, polypropylene fiber and carbon fiber.

The flexural behavior of cementitious composites is mainly determined by many parameters and the interaction between fiber and matrix. Thus, different fiber type, length and content should be studied in depth to minimize cost and maximize flexural behavior of FRCC. To the best of author’s knowledge, no study has been reported on flexural behavior of MHFRCC with combination of calcium carbonate whisker, PVA fiber and steel fiber having various type, content and length. Thus, two type of steel fibers (short steel fiber(SS)and long steel fiber (LS)), two lengths of PVA fibers (short PVA fiber (SP) and long PVA fiber(LP)) and CW are considered to form different MHFRCC. The purpose of this study is to explore the effect of steel-PVA fiber lengths and contents on the overall flexural performance of CaCO₃ whisker reinforced cementitious composites. Thus, many tests were employed to characterize the measured physical properties of MHFRCC. The mechanical parameters of the MHFRCC are investigated to determine the influence of steel fibers, PVA fibers and CW at different flexural stages.

2. FLEXURAL PARAMETERS OF MULTISCALE HYBRID FIBER REINFORCED CEMENTITIOUS COMPOSITESTOP

The flexural behavior of MHFRCC was evaluated by different flexural parameters according to ASTM standards (20–21), post-crack strength (PCS) (1, 22-23) and reinforcing index (RI) (22-25).

2.1. ASTM standardsTOP

The point that becomes first nonlinear on the load-deflection curves is taken as the first cracking point according to ASTM C1018.The first-crack strength, deflection and toughness are important parameters to evaluate pre-cracking behavior. The peak-load point of load-deflection curve described in ASTM C1609 is another important point and its corresponding peak strength, deflection and toughness are used to evaluate flexural behavior (11, 19). Furthermore, the residual strength and flexural toughness at specific point such as 600/span, 150/span are also used to analyze the flexural behavior of MHFRCCs in this study (21). Besides this, the 100/span point is further studied to describe the flexural behavior of MHFRCC (11, 19). The schematic diagram of these specific points is shown in Figure 1.

The flexural toughness is the energy absorption capacity of the test samples and can be obtained by calculating the areas under the load-deflection curve up to a specified deflection. The values of strength at these specific points are used in the following Equation [1] according to ASTM C1609/1609M-12.

Where f is the strength at specific point; P is the applied load at these specific points; b and h are the width and height of specimen, respectively.

2.2. Post-crack strength techniqueTOP

The flexural toughness could be obtained through two standard methods, i.e. ASTM C1018 and JSCE-SF4. However, Banthia and Trottier (1) reported thatASTM C1018 and JSCE-SF4 test methods have some imperfection. The main problem of these two test methods was discussed and their susceptibility to human judgment errors was shown. The calculation of toughness index requires an accurate location of first cracking point but ASTM C1018 standard has occasionally in determining the location of first crack point. Similarly, the deflection point of span/150 which described in JSCE- SF4 is always be criticized for having higher value than that of acceptable deformation limit. Based on above reasons, post-crack strength (PCS) technique was proposed to evaluate the flexural toughness of FRCC. The schematic diagram of PCS technique is shown in Figure 2 and is calculate by Equation [2] and Equation [3].

Where E_post,m is equal to the total energy (E_total,m) minus the pre-peak energy (E_pre); L is span length of specimen; δ_peak is the value of deflection at ultimate load; the value of m can be chosen on the basis of the considered application. Where b and h are the width and height of the specimen, respectively.

2.3. Reinforcing indexTOP

The fibertype, lengthand content are significant factors for evaluating the cracking behavior of FRCC. Ezeldin and Balaguru (26) proposed the reinforcing index to evaluate the effect of hooked end steel fiber in concrete. Then reinforcing index was extended to evaluate the effect of other type of steel fiber in concrete (22, 23, 27). CECS38:2004 (28) defined a characteristic value of steel fiber which had relationship with fiber content and aspect ratio. However, these stipulations were only for one type of fiber and could not be applied to hybrid fibers. Meanwhile, the simple superposition method of reinforcing index for each fiber is not a good choice because different fibers present different characteristics. Thus, a comprehensive reinforcing index(RI_v) was developed to describe the effect of hybrid fiber system (24, 25, 29). The formula of comprehensive reinforcing index is presented as follow in Equation [4]:

Where RI_v is comprehensive reinforcing index of hybrid fiber and k_i is the mechanical anchoring coefficient between fiber and matrix. In this study, SSF, SPF, LPF and CW are taken as 1, and HSF is taken as 1.66 (28). The v_fi, l_i, d_i represent the fiber content, fiber length and fiber diameter, respectively. f_i is the tensile strength of different fiber type and f_s is the tensile strength of steel fiber. The m is fiber type index, and for steel fiber it is 1and for PVA fiber and CW both are taken as 0.5 (29).The suffix i represent different fiber type. The value of i is taken as 1, 2 and 3 for steel fiber; PVA fiber and CW, respectively.

3. EXPERIMENTAL PROGRAM TOP

3.1. MaterialsTOP

The raw materials were Portland cement (P·O 42.5R), silica sand, ordinary tap water and superplasticizer. The chemical compositions of cement is shown in Table1. The physical properties and mix proportion of matrix are presented in Table 2. The water/ cement ratio was kept as 0.3, and sand/ cement ratio was controlled to 0.5 according to previous study (16, 18).The superplasticizer of 0.5wt %-1.5wt% was used to ensure workability of fresh mixture because of addition of PVA-steel fibers and CaCO₃ whisker. The steel fibers were from Bekaert Co. and PVA fibers were acquired from Wanwei High-Tech Material Co. (Chaohu, China). The CaCO₃ whiskers were provided by Youxing Technology Co. (Changde, China), and their chemical composition is also shown in Table 1.The appearance of these fibers are shown in Figure 3. The mechanical properties of these fibers were provided by manufacturer and are presented in Table 3.

The previous findings (16) showed that 1.5 vol % of steel fibers, 0.5 vol % of PVA fibers and 1.0 vol % of CaCO₃ whiskers and 1.25 vol % of steel fibers, 0.55 vol % of PVA fibers and 2.0 vol % of CaCO₃ whiskers had a better flexural behavior for MHFRCC. Therefore, these two fiber contents are selected in this study and the detailed mix proportions of CaCO₃ whisker-PVA-steel fiber are shown in Table 4.

3.2. Mixing, casting and test procedureTOP

Model HJW-60 concrete blender was used to mix the raw material. First of all, Portland cement, silica sand and CW were added into blender and rotated for about 30 seconds to ensure homogeneity of dry material. Secondly, water blended with superplasticizer were divided into three parts and poured into mixer three times during next 60seconds mixing. The fibers were added gradually until the mortar showed a good workability. Since PVA fiber was easier to agglomerated than steel fiber, so the sequence of addition was steel fiber followed by PVA fiber. To ensure the dispersion, steel fiber was divided into three parts and added into mortar for three times during next 120 seconds. Then, PVA fiber was gradually added and mixed for next120 seconds, so that dispersion can be improved and agglomeration can be reduced. Finally, defoaming agent (10 ml tributyl phosphate) was added into mixture and blended for another 15 seconds to eliminate the bubbles caused by addition of fibers and whiskers. The flow chart of mixing process of raw materials is shown in Figure 4.

After that, fresh mixture was carefully placed in the middle of plastic molds one time by a spoon to let fresh mixture flow from the middle to both ends of plastic molds. The molds were then put on vibrator for 30 seconds to improve compactness of fresh mixture which guarantee the consolidation between fibers and mortar (30). Three specimens were casted from each mixture. Later, all the specimens were covered with plastic sheet and placed at laboratory temperature. After 24 hours, the specimens were demolded and cured for 28 days at 20±2ºC temperature with more than 95% relative humidity according to GB/T 50081-2002 (31).

The dimension of beam was 100 mm × 100 mm × 400 mm for performing flexural properties. At the same time, 100 mm × 100 mm × 100 mm cubes were also cast for testing compressive strength according to CECS13:2009(30). Before testing, two-line displacement sensors GA-10 (LVDT) made in Beijing King Sensor Technology Co. Ltd. were fixed with a special device to measure mid-span deflection on both sides of specimen. The application of this special device could avoid additional deformations came from support, loading points and twisting of the specimen. The BLR-1/10T load cell was employed to measure load value. The four-point flexural test was performed to determine flexural parameters of beams by electro-hydraulic servo universal testing machines. The displacement control method with a loading rate of 0.05mm/min according to ASTM C1609. The testing data were collected using DH3820 high speed static strain test analysis system at 5 Hz. The loading setup of beam is shown in Figure 5 and the schematic diagram of data collection is presented in Figure 6.

4. RESULT AND ANALYSISTOP

4.1. Compressive strengthTOP

The compressive strength of MHFRCCs are presented in Table 5 and Figure 7. It can be seen that samples cosisiting of fibers showed a higher compressive strength than that of plain sample. In comparison with plain group, compressive strength of SS2.0 and LS2.0 were increased by 39.27% and 11.58%, respectively. The increasement values in compressive strength can be attributed to crack resistance effect of steel fibers in cementitious composites. The compressive strength of SS1.5SP0.5, SS1.25SP0.75, SS1.5LP0.5 and SS1.25LP0.75 were reduced by 12.93%, 10.62%, 4.24% and 14.84%, respectively, as compared to that of SS2.0. The reduction of compressive strength may be due tolow steel fiber content which was replaced by PVA fiber ultimately brought new interfaces. However, a rising trend were observed in LS1.5SP0.5, LS1.5SP0.4CW1.0, LS1.25SP0.75, LS1.25SP0.55CW2.0, LS1.5LP0.5, LS1.5LP0.4CW1.0, LS1.25LP0.75 and LS1.25LP0.55CW2.0 specimens. The reason for increase in compressive strength is very likely the length and content of long steel fibers in LS2.0 which resulted in non-uniform dispersion of long steel fiber in cubes. The compressive strength of SS1.5SP0.4CW1.0, SS1.25SP0.55CW2.0, SS1.5LP0.4CW1.0, SS1.25LP0.55CW2.0, LS1.5SP0.4CW1.0, LS1.5LP0.4CW1.0, LS1.25LP0.55CW2.0 were increased by 1.32%, 11.44%, 5.42%, 20.44%, 3.48%, 14.91% and 2.28%, as compared to that of SS1.5SP0.5, SS1.25SP0.75, SS1.5LP0.5, SS1.25LP0.75, LS1.5SP0.5, LS1.5LP0.5 and LS1.25LP0.75, respectively. This indicated that addition of CW into PVA-steel fiber specimens could increase compressive strength because micro-sized CW filled pores of matrix and improve the compactibility of cubes (12, 17). However, compared with LS1.25SP0.75, the compressive strength of LS1.25SP0.55CW2.0 was decreased by 16.4%which is probably related to discreteness of compression test. Furthermore, the cubes of series-II showed overall higher compressive strength than that of series-I, series-III and series-IV. The SS1.25LP0.55CW2.0 showed highest compressive strength of 62.0 MPa.

4.2. Ultimate flexural strengthTOP

The results of ultimate flexural strength of beams are presented in Table 5 and Figure 8. The plain group had lowest flexural strength and its ultimate flexural strength was only 6.65 MPa. The ultimate flexural strength of SS2.0 and LS2.0 were increased by 51.89% and 68.63%, respectively,as compared to that of plain group. This is because presence of steel fiber could effectively control generation and development of crack by bridging effect. Compared to SS2.0, the ultimate flexural strength of SS1.5SP0.5, SS1.25SP0.75 and SS1.5LP0.5 were decreased by 4.69%, 2.32% and 2.16%, respectively, but decreasing trend was not significant. The reason for decrease is weak interface between fibers and matrix caused by addition of PVA fibers. However, SS1.5SP0.4CW1.0, SS1.5LP0.4CW1.0 and SS1.25LP0.55CW2.0 leaded upto 16.94% increment in ultimate flexural strength. The reason may be that CW could improve compactibility of matrix which improves bonding capacity between fibers and matrix.Similarly, ultimate flexural strength of LS1.5SP0.4CW1.0, LS1.25SP0.55CW2.0, LS1.5LP0.4CW1.0 and LS1.25LP0.55CW2.0 were increased by 12.52%, 38.39%, 17.14% and 38.27%, respectively, as compared to that of LS2.0. The results further verified that multiscale hybrid fiber system could improve ultimate flexural strength. The improvement effect came from cracking resistance of CaCO₃ whisker-PVA-steel fibers at multi scales such as micron-sized CW, meso PVA fibers and macro steel fibers could delay generation, propagation and development of cracks at micro, meso and macro level, respectively (16, 18). However, SS1.25SP0.55CW2.0 and LS1.5SP0.4CW1.0 were decreased by 4.81% and 4.07%, respectively, as compared to SS1.25SP0.75 and LS1.5SP0.5, respectively. The decrement in ultimate flexural strength of SS1.25SP0.55CW2.0 and LS1.5SP0.4CW1.0could be caused by poor dispersion of some of CW which agglomerate and caused matrix defects. Moreover, it can be seen that the ultimate flexural strength of series-III andseries-IV were higher than that of series-I and series-II. This indicates that long steel fiber had morecontribution to ultimate flexural strength, as compared to that of short steel fiber. The LS1.25SP0.55CW2.0 showed highest ultimate flexural strength of 15.4 MPa.

4.3. Flexural behaviorTOP

4.3.1. Flexural responseTOP

The load-deflection curves of all composites are shown in Figure 9. The plain group presented typical brittle fracture characteristic and had lowest flexural load and deflection capacity. On the other hand, samples with fibers showed significant ductile failure characteristics and had better load carrying and deformation capacity than that of plain group. The samples consisting of PVA-steel fibers showed extended softening behavior than that of samples with single steel fiber. This behavior could be attributed to cracking resistance effect of PVA and steel fibers at meso- and macro-scales, respectively. The flexural response of PVA-steel fiber specimens was further increased with addition of CW which provided cracking resistance effect at micro-scale.

The flexural parameters at five specific points (first crack, peak load, L/600, L/150 and L/100 points) described in ASTM standards were used to describe flexural behavior of MHFRCC. It can be seen from Table 6 that samples consisting of fibers had better flexural behavior than plain group. The first crack deflection (δ₀) of CW-PVA-steel fibers specimens were higher than that of respective PVA-steel fibers specimens. This indicated that CW contribute to inhibit the generation of microscale crack and improve crack resistance before first cracking. The peak load deflection (δ_m) of samples having CW-PVA-steel fibers were about 1.5 to 12.0 times higher than plain group because CW-PVA-steel fibers system could arrest cracks at multiscale before peak load. For instance, micron-sized CW could inhibit generation and evolution of small cracks at micro-scale, PVA fiber played an effect role by preventing extension of cracks and decentralized cracks at mesoscale and steel fiber could control cracks by fiber bridging at macroscale. After peak load, with increase in deflection, the residual flexural strength (f_L/600, f_L/150 and f_L/100) began to decrease gradually. It was worth noting that order for residual flexural strength of samples were series-IV followed by series-III, series-II and series-I. The result showed that long steel fibers play a key role after peak load stage because long steel fibers could arrest macro-cracks and hooked ends shape could improve the mechanical anchoring capacity between long steel fibers and matrix. Furthermore, this also indicated that long PVA fiber exhibited a better crack arresting capacity than short PVA fiber in post-peak stage because length of long PVA fiber will be more effective as crack width increases.

4.3.2. Flexural toughnessTOP

Toughness is defined as energy absorption capacity of test specimens.The flexural toughness could be obtained by calculating area under the load-net deflectioncurve up a specified deflection according to ASTM C1018-97. The results of flexural toughness at five specific points are presented in Table 6. The flexural toughness of FRCC groups wereremarkably higher than that of plain. Compared to single steel fiber specimens, the samples consising of hybrid fibers brought dramatically improvement in flexural toughness. Figure 10 show the flexural toughness of different samples. It can be seen that flexural toughness of CW-PVA-steel fiberspecimens were higher than that of their respective PVA-steel fiberspecimens at first crack point. This may bedue to addition of CW which act as a filler and improve the compactibility of matrix, thereby enhancing crack resistance before first cracking.Furthermore, CW and PVA fiber of high aspect ratio can delay generation and propagation of mircocracks by mechanism of fiber bridging. Also, the same resultswereobserved at peak point (see Figure 10) and LS1.25SP0.55CW2.0 exhibited best flexural toughness at peak point and reached to 11.898 N∙m. However, SS1.25SP0.55CW2.0 showeda little difference after peak load point. Compared to SS1.25SP0.75, flexural toughness of SS1.25SP0.55CW2.0 at L/600, L/150 and L/100 points were decreased by 25.5%, 42.0% and 45.2%, respectively. In comparison with other flexural parameters of SS12.5SP0.55CW2.0 before peak load, it could be concluded that poor dispersion of hybrid fibers resulted in weak flexural behavior. In addition to this, the main energy absorption capacity wastaken from peak load to L/150 deflection as shown in Figure 10. This demonstrated that strain hardening behavior was beneficial to increase energy absorption capacity of samples.

Furthermore, PCS technique was also used to evaluate flexural toughness of samples. In this study, m wastaken as 600, 150 and 100 to describe flexural toughness. So, the value of L/mfor L/600, L/150 and L/100 were 0.5, 2 and 3 mm, respectively. The aimto select these L/mvalues wereto maintain consistency with the evaluation of flexural toughness based on ASTM standards. It can be seen from Table 7 and Figure 11 that PCS values at small deflection werehigher than that of large deflection. The tendency of fluctuations at 0.5 mm were different from that of 2 mm and 3mm which is due tocrack bridging effect of PVA-steel fibers at 0.5mm. However, only steel fibers affected cracking behavior at 2 mm and 3 mm. The LS1.25SP0.55CW2.0 and LS1.25LP0.55CW2.0 hada higher PCS values than others when L/m equal to 0.5 mm. Moreover, PCS loss rate was also relatively higher when L/m equal to 2 mm and 3 mm. The resaon maybe due to low content of long steel fibers. For instance, samples consising of 1.5 vol % of steel fiber, 0.4 vol % PVA fiber and 1.0 vol % of CW (SS1.5SP0.4CW1.0, SS1.5LP0.4CW1.0, LS1.5SP0.4CW1.0 and LS1.5LP0.4CW1.0) had highest PCS in their respective series. This indicated that steel fibers with 1.5 vol% contents improved post-peak behavior at a larger deflection more efficiently. With increasein crack width, long steel fibers had a better bridging capability than that of short steel fibers. However, it is also need to pointed out that decreased PCS was not only related to the geometrical shape size and content of fibers but also in connection with location and characteristic of cracks.

4.3.3. Influence of different fiber combination on flexural behaviorTOP

The flexural parameters with different fiber combination are illustrated in Figure 12. At first cracking points, the hybrid fiber combination of LS+SP+CW significantly improved deflection capacity (δ₀), flexural strength (f₀) and flexural toughness (T₀) of cement-based materials. The LS1.5SP0.4CW1.0 and LS1.25SP0.55CW2.0 exhibited a better flexural behavior than that of other samples. The excellent flexural behavior at first cracking points may be due to hybrid effect of short PVA fiber and CW which could inhibit cracks at micro-meso scales before first cracking. Also, the long steel fiber could disperse easily at the same content ascompared to short steel fiber due to its less quantity. When the flexural strength reach peak, the samples consising of LS+SP+CW still showed a good flexural behavior. The deflection capacity (δ_m), flexural strength (f_m) and flexural toughness (T_m) of LS1.25SP0.55CW2.0 was 0.299 mm, 15.423 MPa and 11.898 N∙m, respectively. However, the crack resistance capacity of CW and short PVA fibers decreased gradually with increase in crack width which is likely due to the limition of its small length. Moreover, at the same time long steel fibers and long PVA fibers started to play a bridging role to control cracks at macro level. Thus, it can be seen from Figue 12 (c), 12 (d) and 12 (e) that the hybrid fiber combination of LS+LP+CW showed a higher flexural properties, especially LS1.5LP0.4CW1.0 presented highest flexural toughness.

Moreover, the flexural behavior is also in connection with hybrid fiber contents. The LS1.25SP0.55CW2.0 showed higher flexural parameters before L/600 deflection as compared to that of others samples. The 1.25 vol % of long steel fiber, 0.55 vol % of short PVA fiber and 2.0 vol % of CW was considered as the optimium fiber combination before L/600 deflection. However, LS1.5LP0.4CW1.0 exhibited a better flexural behavior after L/600 deflection and 1.5 vol % of long steel fibers, 0.4 % of long PVA fibers and 1.0 % of CW was taken as the optimium fibers combination.

4.3.4. Crack behaviorTOP

The crack behavior of all composites are presented in Figure 13. It can be seen that fractures of all samples demonstrated one major crack. The plain specimen was broken into two parts but FRCC was still connected by fibers dueto the effect of fiber bridging. However, more secondary cracks were observed in CW-PVA-steel fiber specimensas compared to that of PVA-steel fiber specimens. This is consistent with the results of flexural toughness. The resaon maybe due to the crack arresting mechanismof steel fiber, PVA fiber and CW at macro-, meso- and micro-scales, respectively. Thus, it could be concluded that multiple cracking behaviors may contributes towards improved energy absorption capacity.

4.4. Correlation between reinforcing index and flexural behaviorTOP

The hybrid fiber type, length and content are significant material parameters affecting the flexural behavior of beams. Thus, it is necessary to synthesize these materials parameters to describe characteristics of hybrid fibers. The comprehensive reinforcing index (RI_v) was a suitable parameter used by Almusallam et al. (24) and Cao (29). In this study, RI_v values were calculate by Equation. (4) and the resultsare shown in Table 7. The relationship between RI_v and flexural parameters of beams are shown in Figure 14. A linear relationship between RI_v and flexural parameters was developed and the trend of fitted linesshowed that the flexural parameters increases with the increase in RI_v. Furthermore, a slightly increase in first cracking deflection was observed with the increase in RI_v (see Figure 14(a). There as on is because the samples seems to be more elastic with the decreased modulus of elasticity asRI_v increases (22, 23). The fitted curve of δ_m showeda significant increment of deflection at ultimate load as RI_v increases. This indicates that addition of hybrid fiber brought significantly improvement in deformability of cement-based materials. The linear fitting results for deflection can reasonably reflect the regularity of deflection variation with increasing of RI_v.Figure 14 (b) demonstrated that theincreased RI_vcan provide a better load carrying capacity. In particularly, as RI_v increases, the increment in trendency of f_L/600 were more obvious than others ( f₀, f_m, f_L/150 and f_L/100).Figure 14 (c) showed that the flexural toughness were plotted with increasing RI_v. The increamentin RI_v results in increased flexural toughness and all of the fitted curves for flexural toughness was found to be linear. The slope of fitted curves for T₀ was almost parallel to x-axis. The reason maybe that only micron-sized CW and some of short-sized PVA fibers were playedan effectiverole in crack arresting before first cracking. Later, the cracks was extensively propagated as loading increased and the hybrid fibers began to work there by improving energy absorption capability. Thus, the increase in flexural toughness at large deflection was more evident than that at first cracking deflection. It can be seen from Figure 14 (d) that the general trend of these results seems to be directly proportional to RI_v value according to the linear fit of PCS. The increased PCS indicated a better energy absorption capability of samples as RI_v increased. The resluts showed a good consistency with the flexural parameters described in ASTM C1609.

5. DISCUSSIONTOP

The PVA-steel fibers-CW specimens showed a better compressive strength and flexural behavior than that of their respectivesteel-PVA fiber specimens. The SS1.25LP0.55CW2.0 exhibited highest compressive strength.The compressive strength of SS1.25LP0.55CW2.0 was increased by 42.9%, as compared to that of plain specimen. This is because 2.0 vol % of CW could maximum fill poresand improved compactibility of matrix.The ultimate flexural strength of LS1.25SP0.55CW2.0 was increased by134% than that of plain specimen.Meanwhile, LS1.25SP0.55CW2.0 exhibited highest flexural toughness before L/600 deflection. This can be attributed to the crack arresting effect of 1.25 vol % LS, 0.55 vol % of SP and 2 vol % of CW atmacro-, meso- and micro-scales before peak load. Moreover, the LS1.5LP0.4CW1.0 showed highest residual strength and flexural toughness at L/150 and L/100 deflections with increase in crack width which is probably because of long PVA fiberbridging effect at large deflection. Compared to LS1.5LP0.4CW1.0, the residual strength and flexural toughness of LS1.25LP0.55CW2.0 wereless at L/150 and L/100 deflections. The reason maybe due to more long steel fibers and long PVA fibers contents which arrest cracks at large deflection. The PCS results also indicated that LS1.25SP0.55CW2.0 had a better energy absorption capability, i.e.13.6 MPa than other specimens at L/600 deflection. The LS1.5LP0.4CW1.0 showed highest PCS value at L/150 and L/100 deflections.Thus, the evaluation results on flexural toughness based on PCS technique had good consistency with the calculated results based on ASTM standards.The crack pattern of PVA-steel fibers-CW specimens showed more amounts of cracks on the side of beams(especially, LS1.25SP0.55CW2.0 and LS1.5LP0.4CW1.0) which indicated that the occurrence of multiple cracks could consume more energy. Moreover, poor dispersibility of fibers can account for the weak flexural behavior.The relationship between comprehensive reinforcing index (RI_v) and flexural parameters indicated that there existed a linear relation. Also, the trend of fitted lines between RI_vand flexural parameters showed that the flexural parameters increases with increase in RI_v. The comparative flexural behavior of hybrid FRCCs with different CW, PVA-steel fiber content and length can further comprehensively understand in structural applications. The optimized CW and PVA-steel fiber content and length are favoring it utility for improving performance of flexural members.

6. CONCLUSIONSTOP

The flexural deflection capacity (δ), strength (f), toughness (T) and post-crack strength (PCS) were determined to evaluate flexural behavior of multiscale fibers reinforced cementitious composites (MHFRCC).The inflence of different PVA-steel fiber length and contenton CaCO₃ whisker reinforced cementitious composites were discussed. The following conclusions were made:

Hence, the optimized CaCO₃ whiskers, PVA and steel fiber length and content are favoring its utility for structural application. The hybrid fibers combination of LS1.25SP0.55CW2.0 and LS1.5LP0.4CW1.0 can be helpful in improving flexural performance of beams.

ACKNOWLEDGMENTTOP

The authors would like to acknowledge the support of this work by Natural Science Foundation of China under Grant No.51678111 and No.51478082. The authors are also thankful to China Scholarship Council (CSC) for providing financial support for PhD studies of Engr. Mehran Khan at Dalian University of Technology, Dalian, China.

REFERENCESTOP