The effect of varied types of steel fibers on the performance of self-compacting concrete modified with volcanic pumice powder

This study used Ordinary Portland cement (OPC), type I of OPC utilized according to ASTM C150 [22] Tables 1 and 2 illustrate the physical properties and chemical composition of OPC, respectively.

VPP was used in this study at 30% partial replacement of cement mass. Moreover, VPP was obtained from a quarry to produce lightweight aggregates in the Jazan region, Saudi Arabia, as shown in Figure 1. The VPP was collected from waste materials resulted from crushing volcanic pumice. The collected VPP was passed through a sieve of 150 mm to remove the bigger particles. Then, the VPP was refined by balls mill to produce fine powder with an average diameter of 4.96 μm (Figure 2). The VPP physical properties and cement are displayed in Tables 1 and 2, which illustrate the chemical composition of VPP and cement. The fine VPP chemical properties met the ASTM C618 PM specification requirements as reported in a previous study [23].

Fig. 1

Fig. 2

The chemical composition of cement and VPP were measured by X-Ray Fluorescence (XRF) test, using the X-ray spectrometer Rigaku RIX 3000. The XRF analysis was carried out according to BS EN ISO 12677 [24].

A crushed limestone rock with a maximum size of 12.7 mm in the concrete mixture was used as gravel, while natural desert sand was used as sand. The gravel and sand showed 1.1% and 1.6% water absorption and 2.67 and 2.70 specific gravities, respectively. The gravel and sand met the specification of ASTM C33 [25]. Figure 3 shows the grading of fine and coarse aggregate.

Fig. 3

Five fiber types were utilized for the preparation of SCC samples. The types of fibers included a hook-end of 60 mm in length (SF1), a hook-end of 30 mm in length (SF3), a flat-end of 50 mm in length (SF2), as well as microfibers measuring 21 mm (SF4) and 13 mm in length (SF5). In addition, there was application of hybrid fibers at 0.2% for each of the five types. The SF volume fraction used in this research was 1% of concrete volume; this percentage has been suggested based on several studies that recommend adding 0.5% to 2% of SFs as the ideal percentage of the concrete volume fraction, so that decline in the fresh concrete properties and strength can be avoided [26]. Table 3 reports the characteristics of these fibers.

Superplasticizer (SP) Sicka 1050 was used in this study for SCC production and SCFs. The SP properties are provided in Table 4.

Table 7 presents the absorption and porosity of the concrete mixtures studied. As shown in Table 7, the absorption of concrete mixtures increased by adding SF to concrete. An average increase in absorption in the entire mixtures ranged approximately between 1.23% and 23.19%, with the inclusion of fibers in the mixtures. As shown in Table 7, adding SF to concrete increased the porosity between 0.38% and 13.53%, compared with SCC. Besides, the small length fibers (i.e., SCF3 and SFC5) had high absorption and porosity compared with the large length fibers (i.e., SCF1 and SFC4), which can be attributed to the surface area of the short fiber, which is higher than the long fibers, and thus more spread in the concrete mixture. This increased the internal friction forces in the mixtures, thereby increasing the pores and the absorption in the mixtures of fiber concrete [40]. The use of VPP as a partial substitute for cement, on the other hand, significantly contributed to reducing the permeability of concrete, and this result has been reported by Zeyad and Al-Maliki in their study [41].

The compressive strength at different ages of SCC, SCF1, SCF2, SCF3, SCF4, SCF5, and SCF6 mixtures are shown in Figure 7. All concrete mixtures containing fiber were higher than the SCC mixture. The increase in compressive strength of concrete containing fibers may be attributed to arresting of the growth of cracks, depending on the bond strength (i.e., confinement effect) of SF and the concrete matrix [42]. Also, SCF3 (i.e., the smaller hooked fiber) exhibited the highest compressive strength compared with all the mixtures, while SCF6 (i.e., hybrid fiber) had the lowest compressive strength in all fiber mixtures as displayed in Figure 7. The average compressive strength of SCF3 was 22.47%, 6.16%, 5.56%, 4.49%, 8.99%, and 15.62% higher than the SCC, SCF1, SCF2, SCF4, SCF5, SCF6 mixtures, respectively. The results showed that the SCF3 mixture at 7 days was 29.78%, 5.17%, 4.45%, 5.90%, 8.54%, and 19.14% higher than the SCC, SCF1, SCF2, SCF4, SCF5, SCF6 mixtures, respectively. A similar trend was detected for the compressive strength of the SCF3 mixture at 90 days, that is, it was 18.66%, 6.20%, 5.08%, 3.12%, 8.53%, and 14.55% higher than the SCC, SCF1, SCF2, SCF4, SCF5, and SCF6 mixtures, respectively. Furthermore, the results revealed that there exists a slight difference in compressive strength between the SCF1, SCF2, and SCF5 mixtures.

Fig. 7

The indirect tensile strengths (splitting tensile) of various concrete samples were determined after 7, 14, 28, and 90 days; the average of three-cylinder in each type of concrete mixture was measured and is displayed in Figure 8. The indirect tensile strength was enhanced significantly when using SF in producing the SCF mixtures versus SCC. Therefore, the mixes’ indirect tensile strength were 3.2 MPa, 3.93 MPa, 3.82 MPa, 4.06 MPa, 4.35 MPa, 4.16 MPa, 4.08 MPa at 28 days for SCC, SCF1, SCF2, SCF3, SCF4, SCF5, and SCF6, correspondingly. Nevertheless, there was an increase in the indirect tensile strength by 6.56%, 5.59%, 4.97%, 5.67%, 3.68%, 4.57%, and 2.94% at 90 days for concrete mixtures of SCC, SCF1, SCF2, SCF3, SCF4, SCF5, and SCF6, respectively. The addition of SF to SCC significantly increased the indirect tensile strength at all test ages, as shown in Figure addition, SCF4 (i.e., larger Microfiber) exhibited the highest indirect tensile strength compared with all the other mixtures. As shown in this research, all fibers increase their indirect tensile strength capabilities by >19.5%. The existence of the fiber content enhances the characteristics of tensile strength due to the fibers’ action in arresting and bridging the concrete matrix cracks; the forces of bridging are extremely dependent on the fiber type [43].

Fig. 8

The compressive strength effects on the indirect tensile strength of SCC, SCF1, SCF2, SCF3, SCF4, SCF5, and SCF6 are shown in Figure 9. With increase in compressive strength, the indirect tensile strength also increased. Furthermore, the correlation between the compressive strength and indirect tensile strength results are shown in Eq. (2).

Fig. 9

According to ASTM C 1018 [35], the performance of SCF beams can be obtained under third-point loading, by calculating the amount of energy required for first-crack deflection (δ_cr) and then to further deflect to specific multiples of the first-crack deflection (i.e., 3δ_cr, 5.5δ_cr, 10.5δ_cr, etc.) based on Figure 10.

Fig. 10

The toughness index is determined by dividing the complete area under the load–deflection (P − δ) curve up to the given service level deflection by the area under the load–deflection curve up to the first-crack deflection. Toughness indices I₅, I₁₀, I₂₀, and so on, are determined on three service levels at 3, 5.5, 10.5, times of the first-crack deflection, respectively, as described in Eqs. (3)–(5) [35].

The residual strength factors (R) can evaluate the average post-crack load retained over a specific deflection (i.e., 3 times, 5.5 times, and 10.5 times of the first-crack deflection) as described in Eqs. (6) and (7) [35].

The index of toughness and the factors residual strength of plain concrete was equivalent to one and zero, correspondingly, due to the given brittle response. Thus, the energy at first-crack is very important because the entire performance parameters, i.e., (I and R) were later calculated according to it. An accurate determination of the beams’ deflection is also very important after the first crack. In this study, Eqs. (3)–(5) were used to calculate the toughness indices I₅, I₁₀, and I₂₀ through the areas under P − δ curves (Figure 11) related to 3, 5.5, and 10.5, times of the first-crack deflection (i.e., 3δ_cr, 5.5δ_cr, 10.5δ_cr, etc.,). In addition, I₅, I₁₀, and I₂₀ were used to determine the residual strength factors R_5,10 and R_10,20 as described in Eqs. (6) and (7). Figures 12 and 13 show the areas under the curves of 3δ_cr, 5.5δ_cr, and 10.5δ_cr, toughness indices, and residual strength factors for all concrete mixtures.

Fig. 11

Fig. 12

Fig. 13

As shown in Figure 12A, the highest first-crack loads value was recorded in the SCF3 mixture, while the lower first-crack loads value was recorded in the SCF6 mixture. The results revealed that the first-crack loads of the SCF1, SCF2, SCF4, SCF5, and SCF6 mixtures were different from the results of the SCF3 mixture, reaching 39.4%, 32.1%, 30.5%, 22.1%, and 43.9%, respectively. Moreover, the SCF6 exhibited the highest first-crack deflections compared with all the other mixtures, as shown in Figure 12B. The increase ratios were 86%, 10%, 8.5%, 46.2%, and 70.4% for the SCF1, SCF2, SCF3, SCF4, and SCF5 mixtures, respectively.

According to the ASTM C 1018 [35] method (Figure 12C–F), the maximum increase in area under P − δ curves (Figure 11) up to first-crack deflection (δ_cr) is provided by the SCF3 mixture with an increased ratio of 3.94 times. A similar phenomenon can be observed in 3δ_cr, 5.5δ_cr, and 10.5δ_cr for the SCF3 mixture compared with all the other mixtures. Regarding toughness indices, the highest values of I₅ were recorded in the SCF2 mixture, and the increase ratios were 15.3%, 70%, 16.9%, 9.9%, and 12% for SCF1, SCF3, SCF4, SCF5, and SCF6 mixtures, respectively. Also, a similar result can be observed for I₁₀ and I₂₀ in the SCF2 mixture, with various increase ratios ranging from 15.8% to 116.1% and from 22.2% to 179.8% for I₁₀ and I₂₀, respectively, compared with all the other mixtures Figure 13A–C. Regarding residual strength factors (R_5,10 and R_10,20), the same as toughness indices, the highest R_5,10 and R_10,20 value was recorded in the SCF2 mixture, while the lower value was recorded in the SCF3 mixture, as shown in Figure 13D,E.

Regarding the length of hooked fibers effect (i.e., SF1 and SF3), the increasing length of the hooked-end fibers (i.e., SCF1) resulted in increasing toughness significantly (ranging from 47.53% to 128.96%) compared with SCF3; the results were well consistent with the results in [44]. As evidenced from Figure 13, the smaller hooked fiber (i.e., SCF3) showed a high area under curves compared with SCF1. Despite this increase, the toughness indices and residual strength factors for SCF1 (i.e., longer hooked fiber) were higher than SCF3. On the other hand, the long straight fiber (i.e., SCF5) showed a high first-crack load compared with SCF4 (i.e., smaller straight fiber), with slight differences of over 6.9% only, while the highest first-crack displacement was recorded in SCF4 with increased ratio 24.4% compared with SCF5. Thus, it must be observed that the straight fibers’ aspect ratio can affect the performance of hardened and fresh concrete considerably. It is not practical, however, to assess the values of the aspect ratio via comparing the flat and hooked end fibers’ performance versus straight fibers.

According to ACI 544 [45], the toughness index can be defined as the ratio of the measure of the energy required to deflect a fiber concrete beam by a given amount of the energy required to bring the fiber beam to the point of the first crack. Figure 14 shows the definition of the toughness index based on ACI 544 [45], which can be measured by the area under the load–deflection curve up to a deflection of 1.9 mm (δ_{1.9 mm}) divided by the area under the load–deflection curve up to the first-crack deflection δ_cr of fiber concrete beam (as described in Eq. (8)), which is affected by the properties of concrete and the properties of fiber geometry. The calculated values of toughness indices indicated that the highest value was recorded in the SCF1 mixture, while the lower value was recorded in the SCF3 mixture, as shown in Figure 15. Also, SCF5 shows a slight difference of 6.44% only compared with SCF1. Moreover, SCF2 and SFC4 show close results with a slight difference of about 2.76% only.

Fig. 14

Fig. 15

The performance of concrete mixtures with and without fibers had significant differences which can be observed in the load–deflection curves, first-crack load, and toughness indices. Thus, a standardized method for evaluating the beams’ energy absorption capacity can be conducted by calculating specified areas under the load-induced flexural deflection curve. Load–deflection curves were obtained from the result of the flexure test of beams under three-point loading, as shown in Figure 11. As shown in Figure 11, all the concrete mixtures containing fibers (i.e., SCF1 to SCF6) showed more ductile behavior and more deformation (i.e., the failure propagation has been more uniform and gradual) compared with conventional concrete (SCC), which fails in a brittle at the formation of the first crack. Primarily, the behavior of load–deflection is perfectly elastic up until the first-crack load. The curve growth in this segment is steeply linear. The slope of the load–deflection behavior decreased due to concrete cracking and hence exhibited a nonlinear behavior. The presence of fibers led to increase in first-crack strength through the bridging effect for all concrete mixtures compared with SCC as shown in Figures 11 and 12. After this load level, the deflection of all beams continued with a considerable decrease in the applied load. After the maximum load capacities, as shown in Figure sudden drop was observed in the resisting force for SCC due to the significantly continued extent and widening of the first-crack, which ended once the crack achieved its maximum extent at the end of the beam’s cross-section. Fiber only sustained load after cracking via mechanical interlock and the friction at the fiber–matrix interface. Moreover, the curve’s descending part showed that SF were pulled out gradually from the matrix, which is consistent with [46]. Also, the load–deflection curves’ comparison showed a substantial improvement regarding all concrete mixtures’ inelastic behavior, which contain fibers, whereby fibers can carry tensile forces as in the conventional reinforced concrete. In the latter case, therefore, there was an obvious post-cracking response through delaying the decline rate of the load–deflection curve up until the fibers’ rupture or slipping. This showed that concrete with fibers is highly flexible, and it is more able to absorb energy than the conventional concrete mixture [47]. The load–deflection curve of SCF3 declines slowly and covers a significant area greater than the other fiber types. It also means that the hooked end fiber with a lower aspect ratio can provide higher energy absorption capacity than the other fiber types. This is attributed to the distinct bonding performance between different-shaped steel fiber and concrete matrix sustaining the load at cracks. The results were consistent with those mentioned in [48], while SCF2 shows the highest deformation capacity and the best flexural toughness of all the concrete mixtures. In the meantime, hybrid fiber concrete (SCF6) provided good flexural toughness, as well as better anti-crack performance when compared to single-fiber concrete. This is because it contains different fiber types, which showed the greatest crack bridging.