Electrical resistance and self-sensing properties of pressure-sensitive materials with graphite filler in Kuralon fiber concrete

This study used Type I Portland cement from Taiwan Cement Corporation. The physical and chemical properties of this cement met the specifications of American Society for Testing and Materials (ASTM) C150. The content of CaO, SiO₂, and Al₂O₃ in cement was 68.0%, 20.7%, and 5.1%, respectively. The fine aggregate came from the Lanyang River, and its fineness modulus equaled 2.86. The retention percentages of fine aggregates in No. 8 sieve, No. 16 sieve, No. 32 sieve, No. 50 sieve, and No. 100 sieve were 25.4%, 21.3%, 13.7%, 10.3%, and 9.4%, respectively. The particle size distribution of the fine aggregate met the specifications of ASTM C33. The maximum size of coarse aggregate was 12.5 mm.

Graphite powders (C-B01-P3000-999) manufactured by Ceramet Inc. (Taipei, Taiwan) were used. The purity of the powders was >99.9% (carbon), and the particle size varied from 5 μm to 1 mm (D50 was 10 μm).

The Type G superplasticizer (a water-reducing ratio of 12%) produced by Poplar Company (Taipei, Taiwan) was used to enhance the fluidity of all mixtures. The electrodes used in this study comprised a double-sided conducted aluminum foil tape manufactured by Tai Puu Tape Company (Taipei, Taiwan). The width and surface resistivity of the tape are 2.5 cm and 0.05 ω, respectively, and it displayed good adhesiveness with concrete specimens.

Kuralon (K) fiber is a new polyvinyl alcohol fiber manufactured by the Japanese company Kuraray. The diameter, length, and aspect ratio of fiber are 0.75 mm, 30 mm, and 40 mm, respectively (see Figure 1). Table 1 presents the basic properties of the fibers. The fibers bond well with cement substrates and are widely applied in industry and construction.

Fig. 1

Graphite powder was used to replace 0%, 4%, 8%, 12%, and 16% by weight of the cement in the concrete. The cementitious material was the sum of cement and graphite powders, and water included the superplasticizer. Considering that the inclusion of graphite reduces the strength of the concrete, we used a lower water-to-cementitious ratio of 0.45 so that the concrete would reach basic engineering strength requirements. We also compared mixtures containing an extra 0.5% of the total volume of K fibers with those not containing K fibers. Table 2 displays the various mixture proportions that we experimented to understand the differences between the properties of concrete with different graphite contents and whether adding fibers can enhance its performance. The addition of the super-plasticizer was quantified by the weight of the cement. In addition, the graphite powder was taken in the mixer with the cement for mixing during the mixing process. The specimens were placed in water curing after demolding and taken out of the curing pool before the tests. The results of each test were averaged from three specimens, and the standard deviation was controlled to within 5%.

The first letters K and N in the mixture number distinguish the specimens with and without K fibers, referred to as fiber and normal concrete in this study. The second letter G stands for graphite, and the number behind G indicates the volumetric proportion of cement replaced with graphite powder. Thus, the graphite contents of 0%, 4%, 8%, 12%, and 16% are denoted as G0, G4, G8, G12, and G16, respectively. Table 3 shows the tests performed in this study. The test results of each test were averaged by using three specimens.

The measurement of resistivity in concrete requires electrodes. Electrodes may be embedded or attached to the outside of the specimen. For the sake of costs and convenience in application, it used a conductive aluminum foil tape (width: 2.5 cm), which can be applied to the surface of specimens with a conductive adhesive. The standard resistivity measurement methods included the two-electrode and the four-electrode approaches. The latter can more effectively eliminate nonohmic contact resistance than the former. For this reason, the four-electrode method was chosen to measure resistivity. In addition, the specimens kept were under saturated surface-dry conditions to avoid the influence of humidity on the resistance measurement of the specimen [29].

Figure 2 displays the schematics of resistivity measurement; we used a 7V DC power source and placed the specimen in a series with a resistor of known resistance. As the currents passing through, the specimen and the resistor are equal in magnitude according to Ohm's law, V1R1=V2R2 {{{V_1}} \over {{R_1}}} = {{{V_2}} \over {{R_2}}} means that R1=V1V2R2 {R_1} = {{{V_1}} \over {{V_2}}}{R_2} , where R₁ and R₂ denote the resistance of the specimen and the resistor (a known resistance), respectively. After obtaining V₁ and V₂ using a data collector, we can observe the resistance variations in the specimens. Subsequently, the resistivity can be calculated using the formula ρ=RAL \rho = {{RA} \over L} , where ρ is the resistivity (kOhm-cm), R is the resistance of the specimen (Ohm), A denotes the cross-sectional area of the specimen (cm²), and L symbolizes the distance (cm) between the two electrodes (V₁ and V₂). Figure 3 shows a photo of the actual experiment.

Fig. 2

Fig. 3

Figure 4 displays the results of the compressive strength histograms. Except NG4, the normal concrete specimens without K fibers (NG series specimens) showed a decline in compressive strength as the graphite content increases. Comparing the strength reduction rates between the normal and fiber concrete specimens revealed that the fiber concrete specimens with a graphite content <8% presented significantly lower strength reduction rates than the normal concrete specimens with the same graphite content. By contrast, the strength reduction rates of the two types of concrete with graphite contents >12% were roughly the same. It was speculated that this may have been because the fibers generated a binding force in the specimens that counteracted the friction reduction caused by graphite [34]. The addition of graphite compound fibers enhanced the electrical conductivity of the concrete, but the excessive usage had a negative impact on the compressive strength [35]. However, once the graphite content reached 12%, the internal binding force could not rival the overly low friction, which increased the reduction in strength. It was concluded that the corresponding relationship between conductivity and strength must be considered when using graphite as a conductive material to produce conductive concrete.

Fig. 4

The addition of graphite decreases the elastic modulus of the concrete specimens (see Figure 5). Among the normal concrete specimens, the elastic modulus of the specimen with 4% graphite content (specimen NG4) was not significantly different from that of the control specimens (specimen NG0). By contrast, those with 8%, 12%, and 16% graphite contents were lower by 11%–15%. The decrease is likely due to the directly proportional relationship between compressive strength and the elastic modulus (see Figure 6), which means that the elastic modulus declines when the compressive strength decreases. Among the fiber concrete specimens, the elastic modulus of those with 4% and 8% graphite contents was roughly 4% lower than that of specimen KG0. In comparison, the specimens with 12% and 16% graphite contents were lower by 9%–12%. A comparison of the normal and fiber concrete specimens showed that the elastic modulus of the specimens containing K fibers did not decrease as much, thereby indicating that K fibers can raise the upper limit of the linear stage of elastic modulus development.

Fig. 5

Fig. 6

Figure 6 compares the regression curve of the elastic modulus and compressive strength in this study with that in ACI 318; clearly, the regression curve of this study is considerably lower. This is likely because the strength of foreign aggregate is higher than that in Taiwan cement, which reduces the aggregate-to-cement ratio in the cement and the elastic modulus.

Figure 7 shows the results of the flexural strength test, in which the assessed index is the modulus of rupture. The normal concrete specimens presented declining flexural strength as the graphite content increased. The specimens with 8%, 12%, and 16% graphite contents displayed reductions of 8.8%, 12.4%, and 22.9%, respectively, compared with the control group. Only the specimen with a 4% graphite content had higher flexural strength than the control group, which may be because of its higher compressive strength.

Fig. 7

Flexural strength also declined among the fiber concrete specimens as the graphite content increased. Regarding the magnitude of reduction in flexural strength, the specimen with a 4% graphite content presented a reduction of approximately 4.7%. By contrast, the specimens with 8%, 12%, and 16% graphite contents displayed reductions of 11.6%, 15.4%, and 23.8%, respectively. A comparison of the normal and fiber concrete specimens, as presented in Figure 7, showed that the specimens containing K fibers possessed greater flexural strength than the normal concrete specimens. Figure 8 shows the regression curve of compressive strength concerning the modulus of rupture. As can be seen, the fiber concrete curve is slightly higher than the normal concrete curve, which means adding K fibers enhanced the flexural strength of the concrete specimens.

Fig. 8

The saturated water absorption rate of concrete shows how the porosity and density of the concrete progress with time. Figure 9 displays the results of the saturated water absorption test. Except for NG4, the normal concrete specimens presented lower saturated water absorption rates than the control group. Furthermore, the saturated water absorption rate increased significantly with the graphite content. A comparison of the normal and fiber concrete specimens showed roughly the same saturated water absorption rates in both the groups, which indicates that K fibers cannot reduce the internal porosity of the concrete. The use of graphite powder to replace the amount of cement in the specimen would increase the internal porosity of the mortar, reducing the compactness of the concrete. The gradually increased water absorption and decreased compressive strength tests showed no significant hydration reaction between the graphite powders and the cement particles. Wang et al. [36] found that the specimens with functional fillers produced more hydration crystals than the control ones. Thus, accelerated hydration growth led to reduced porosity, better compactness, pore refinement [37], and an improved microstructure [38].

Fig. 9

The ISA test assesses the capillary adsorption of outside water molecules in the pores of cement-based materials. According to the specifications of BS 1881, the ISA rate is defined as the amount of water absorbed by a fixed surface area of the cement-based material in contact with water within a specified period. This test was performed after a curing period of 56 days. We used the parameters established by Kumar and Bhattacharjee [39] to determine the permeability of the concrete specimens (see Table 4).

As shown in the test results in Figures 10 and 11, all specimens showed high permeability at 30 s, except NG0, NG4, and KG0. Figure 10 indicates that surface water absorption rates increase with the graphite content. Greater surface water absorption rates result from higher porosity, causing the adhesive force of the water to the pore walls to be greater than the cohesive force within the water. It demonstrates that adding graphite increases the internal porosity of concrete. The results, as displayed in Figure 11, show the same trend among the fiber concrete specimens. The lack of differing increases or decreases shows that adding K fibers does not significantly change the internal porosity of the concrete.

Fig. 10

Fig. 11

Figure 12 presents the scanning electron microscopy (SEM) images of specimens (NG series specimens) with 0%, 4%, 8%, and 12% graphite (at a magnitude of 3,000 times). In terms of porosity, the images show that porosity increases with the graphite content. Comparing the three images containing graphite displays an even increasing distribution of graphite particles. The specimen with a 12% graphite content also presents graphite clusters. Graphite can evenly mix with the cement in specimens with a lower graphite content. However, as the graphite content increases, the quantity of water required increases, thereby preventing some particle clumps from dispersing. Furthermore, the surface of graphite particles has some magnetism, which causes some graphite particles to cluster when the graphite content is high.

Fig. 12

Figures 13 and 14 show the resistivity coefficients of the concrete specimens derived using the four-electrode approach. As shown in Figure 13, the resistivity coefficients of the normal concrete specimens containing graphite show the same trend as the mortar [40], in which resistivity decreases as the graphite content increases. The specimen with a 4% graphite content presents approximately 7.4% lower resistivity than the control group, while the specimens with 8%, 12%, and 16% graphite contents show resistivity lower by 37.8%, 46.7%, and 47.4%, respectively. The resistivity coefficients of the fiber concrete specimens in Figure 14 show that the specimen with a 4% graphite content has approximately 14.4% lower resistivity, whereas the specimens with 8%, 12%, and 16% graphite contents show resistivity lower by 35.5%, 35.9%, and 43.2%, respectively.

Fig. 13

Fig. 14

The results show that resistivity did not significantly decrease until the graphite content reached 8%, indicating a slight difference from that of the graphite content of 5%, corresponding to the percolation threshold of the mortar [22]. The cause of this difference may be because the media within the mortar [40] is evenly distributed with small particles, enabling the graphite to form conductive pathways within the specimens. By contrast, the concrete specimens contain coarse aggregates, consisting of larger particles and requiring electrons to travel longer paths than they would in a mortar. It caused the percolation threshold to increase to a graphite content of 8%.

The resistivity of the concrete specimens was observed in terms of age development, with a rapid increase between 7 days and 28 days, followed by slower growth at 56 days. It is because the resistivity was related to the porosity of the specimens and the concentration of ions in the water. When the age of the specimens was 7 days, the internal pores were large and the internal electron conduction was mainly carried out by ions such as Ca²⁺, Na⁺, K⁺, OH⁻, and SO42− {\rm{SO}}_4^{2 - } in the pore water, so the resistance coefficient was lower than that of the specimen at later age [26]. As the age increased, the internal pores gradually decreased. As a result of the hydration reaction, most of the ions are reacted, resulting in a significant reduction in conductivity. At this point, graphite in the specimens began to act as the main bridge between the electrons. This conclusion was consistent with previous research findings by Lee et al. [23] and Le et al. [41]. The addition of conductive powders could reduce the resistivity of the sensing concrete. When the addition amount reached the critical filler value (8% carbon black) [26], that is, the percolation threshold was reached, the trend of resistance reduction began to slow down [42].

Figure 14 presents the resistivity coefficients derived using the DC loop resistance test. As can be seen in Figure 14(A), the normal concrete specimens with 4%, 8%, 12%, and 16% graphite contents show resistivity lower than that of the control group by 40.7%, 53.7%, 61.1%, and 64.8%, respectively. In Figure 14(B), the fiber specimens with 4%, 8%, 12%, and 16% graphite contents show resistivity reduction of 39.3%, 65.6%, 70.5%, and 75.4%, respectively.

The results indicate that the resistivity of the concrete dropped significantly with a 4% graphite content. When the graphite content was 8%, it declined by another 20%. Adding more graphite decreased the resistivity further, which gradually stabilized. The results show resistivity beginning to stabilize when the graphite content was between 8% and 16%, which means that an 8% graphite content is the percolation threshold of the concrete. Although the resistivity of the concrete dropped significantly with a 4% graphite content, it declined even more significantly as more graphite was added. It means that a 4% graphite content has not reached the percolation value of the concrete yet. The results of the pressure sensitivity test support these results.

Figures 15 and 16 display the relationships among resistance variation, stress, and strain in the normal and fiber concrete specimens during the compressive strength tests. As can be seen, the normal and fiber concrete specimens containing no graphite presented poorer regularity as the load increased. During the earlier phase, the resistivity declined slightly. However, as the load increased and reached the fracture strength, the resistivity remained volatile and could not properly reflect the damage conditions within the specimens. The two normal and fiber concrete specimens with a 4% graphite content displayed similar circumstances in the earlier phase, with resistivity showing a significant decrease right before failure. Although the resistance variations in these two specimens were more apparent than those in the two specimens with no graphite, the percolation threshold was not reached. The two normal and fiber concrete specimens with an 8% graphite content also exhibited a considerable decline as the load increased; when the stress reached 50% of the ultimate load, the resistivity decreased less until failure, at which point the magnitude of the decline in the resistivity increased again. It indicates that the specimens have reached the percolation threshold. The normal and fiber concrete specimens with 12% and 16% graphite content showed similar trends in resistance variation, and resistivity decreased as stress increased until failure.

Fig. 15

Fig. 16

In terms of the range of variation in resistivity, the maximum variation ranged from 25% to 35% in the specimens with no graphite, from 50% to 60% in the specimens with a 4% graphite content, around 70% in the specimens with an 8% graphite content, from 80% to 85% in the specimens with a 12% graphite content, and from 85% to 90% in the specimens with a 16% graphite content. The greater variations in resistivity indicate that the specimens containing graphite formed better conductive networks during the compression process. These conductive powders (graphite) increased the sensitivity [43] and repeatability of the sensing concrete [44]. There was no significant difference in the electrical conductivity of the addition of fibers to sensing concrete. The fiber group had better sensitivity when the fiber concrete contained a low amount of graphite (4%).

Under uniaxial compression testing, the distance between conductive particles within the specimen was reduced, improving the conductive network within the concrete; when larger cracks appeared in the compressed specimen, the extension of the cracks led to the destruction of the conductive network. Although not significantly contributing to the conductivity, the addition of fibers was effective in mitigating damage to the conductive network caused by the extension of cracks. The test results also showed that the recovery rate of resistance variation was higher than that of the normal concrete group when the specimens exceeded the compressive failure strength, which is consistent with previous studies by Ding et al. [26] and Bekzhanova et al. [27].

According to the study [40], fine cracks exist within most concrete materials. When the stress in the concrete equals 0%–30% of the ultimate load, the concrete is linear, and the fine cracks within do not develop. When the load applied to the concrete reaches 30%–40% of the ultimate load, the fine cracks begin to spread and increase with stress. At 70% of the ultimate load, visible cracks start to appear inside and on the surface of the concrete and worsen until failure.

Figure 17 presents the changes in resistivity in the ordinary specimens under cyclic loading at 50% of the ultimate load. When the control group specimen is in the linear phase, resistance variations are minor (10%) and do not exhibit clear correspondence to the changes in loading. The resistance variations in the concrete specimen with a 4% graphite content somewhat increase during the linear phase, ranging from 10% to 15%, but still show no apparent correspondence to the changes in loading. By contrast, the maximum resistance variations in the ordinary concrete specimens with 8% and 16% graphite contents range from 20% to 25% and show clear reactions to the changes in loading. Adding graphite powders to micro-scale secondary fillers effectively controlled the sensing performance of the sensing concrete, thereby increasing the maximum resistance variation and sensing sensitivity range when subjected to load.

Fig. 17

Figure 18 presents the changes in resistivity in the fiber specimens under cyclic loading at 50% of the ultimate load. When the specimen containing no graphite is in the nonlinear phase, the resistance variations range from 15% to 25% and do not exhibit clear correspondence to the changes in loading. The resistance variations in the concrete specimen with a 4% graphite content range from 20% to 30% but still show no noticeable correspondence to the changes in loading like in the linear phase. By contrast, the maximum resistance variations in the specimens with 8%, 12%, and 16% graphite content range from 30% and 40%, from 40% to 50%, and from 50% to 60%, respectively, and show clear reactions to the changes in loading. The test results show that the concrete specimens with a 8% graphite content or more present clearer resistance variations in the linear phase.

Fig. 18

Excluding the specimens containing no graphite, whose resistance variations were too slight, only the resistivity of the specimens with a 4% graphite content could not return to their initial values. It speculated that the 4% graphite content was insufficient to form a complete conductive network in the specimens. Although loading reduced the pores within the concrete and enabled graphite particles to come into contact with one another, reducing the resistivity, the pores could not return to their original state upon unloading, thereby producing this result. Despite the pores being unable to return to their original condition in specimens with a higher graphite content, the better conductive networks in the concrete enabled resistivity to return to its initial value.

According to Mindess et al. [45], numerous cracks begin to form in specimens when subjected to loading at 40% of the ultimate strength. Comparing the specimens in the nonlinear phase revealed that resistivity presented slight increases with greater cycles. This phenomenon is most apparent in specimens NG8 and KG16. The increase in resistivity is possible because of the damage inflicted on the specimens during each loading process. In the form of minor cracks, the damage affected the conductive network within the specimens, causing the resistivity to increase with each loading irreversibly. The difference between normal and fiber concrete specimens can be seen in the specimens with 8% and 12% graphite concrete. After different loading processes, the normal concrete specimens show significant increases in resistivity. By contrast, the rising trend is not apparent in the resistivity of the fiber concrete specimens. It shows that adding K fibers to conductive concrete can reduce the speed at which cracks develop within concrete. The use of fibers can significantly reduce the trend of increasing resistivity [27] and has great potential for monitoring fatigue damage in different concrete structures [46].