1 Introduction
Rubberized concrete, as a new type of relatively environmentally friendly material, is widely applied in construction projects. Waste rubber cannot be decomposed, leading to environmental pollution and even fire risks. The processing of waste tires into crumb rubber (CR) mixed with concrete solves the industrial pollution problem caused by the waste rubber, which is significant to the economic cycle and sustainable development worldwide. Rubberized concrete has advantages, such as light weight, high ductility [
1], better fatigue resistance, and durability [
2,
3]. However, the strength of rubberized concrete can significantly deteriorate due to the poor bond between the waste rubber and the matrix [
4,
5], which limits its use in engineering. Incorporating fibers can efficiently compensate for the loss of mechanical properties. Eisa et al. [
6] reported that adding 5% and 10% CR reduced compressive strength by 10% and 24%, respectively. However, the combination of 1% steel fiber (SF) and 10% CR increased compressive strength by 11% and splitting strength by 41%. Zhang et al. [
7] found that adding polyvinyl alcohol (PVA) fiber to concrete with a 25% rubber content increased flexural strength by 12.7%–25.0%. In addition, calcium carbonate whiskers (CW) show good performance in enhancing the strain hardening of concrete. It has been reported that mixing CaCO
3 whiskers, SF, and PVA fibers in rubberized concrete can increase the tensile strength and the ultimate strain greatly, and can improve the strain-hardening capacity [
8]. However, concrete structures may be subjected to various extreme environments, such as fires, during their service life. High temperatures can cause severe damage to concrete structures, significantly deteriorating their mechanical properties and durability. Yu et al. concluded that rubberized concrete suffered a significant reduction of ultimate strength after being subjected to 400 °C and then allowed to cool, because the number of pores greater than 1μm in rubberized concrete was 3.4 times higher than that of ordinary concrete [
9]. Su and Xu [
10] reported that the static and dynamic compressive strengths gradually decreased with increasing temperature, but incorporating 1% basalt fibers and 1.5% polypropylene fibers effectively improved the performance of rubberized concrete under heating. Zhang et al. [
11] concluded that incorporating multi-scale fibers could prevent explosive spalling and had efficient resistance to thermal damage. Therefore, it is important to study the fracture process and the damage evolution of fiber-reinforced rubberized concrete after it has been subjected to high temperatures.
The above research on the damage process of rubberized concrete is based on residual mechanical properties at the macroscale. However, crack propagation is a significant indicator of damage evolution and damage mechanism of concrete. Recently, digital image correlation (DIC), laser speckle, and acoustic emission (AE) technology have been employed to monitor the crack development of concrete under loading [
12], but these technologies cannot reveal the mesoscale damage. Liu et al. [
13] studied the surface strain fields of coral concrete mixed with carbon fibers after fracture using DIC and analyzed the flexural damage process of the specimen. Pang et al. [
14] investigated the deformation and flexural behavior of concrete sleepers under different support conditions using non-destructive laser speckle imaging technology. However, those two techniques mainly reflect the crack propagation on the surface of specimens. It is important to understand the internal damage evolution of concrete at the microscopic scale. AE is a non-destructive monitoring technique to detect the cracking of concrete and the evolution of damage inside the concrete at the microscopic scale based on the elastic wave emitted by the concrete itself due to the release of stored strain energies. The AE monitoring system processes the collected signals to evaluate the degree of damage [
15–
17]. AE parameters, such as peak frequency, amplitude, energy, etc., are often used to reflect the degree of damage to concrete, as shown in Fig.1. Behnia et al. [
18] reported the characteristics of AE from reinforced concrete beams under torsion. They concluded that the peak frequency decreased with increasing damage to the beam. Liu et al. [
19] determined the fracture mode of basalt fiber reinforced reactive powder concrete (RPC) under flexural loading based on the ratio of rise time to amplitude (RA) values and average frequency (AF). It was reported that AE amplitude can reflect the size of internal cracks in concrete [
20]. Liu et al. [
21] investigated that the peak amplitude changed with the decreasing load. In addition, the AE signal gradually weakened when the rubber content increased. Xu et al. [
22] analyzed the damage process of specimens with added ceramic waste aggregate and CR, in the three-point bending test based on AE technology. The cumulative energy, b-value, and RA vs AF values of different types of composite mortars were investigated. It was concluded that the AE energy was relatively low in the early stage of loading, but specimens exhibited macroscopic cracks during subsequent failure, resulting in an increase in AE energy. However, the AE characteristics in the fracture process of fiber-reinforced rubberized concrete under uniaxial compression and tension after exposure to high temperatures remain to be revealed.
Inspired by these investigations, in this paper, uniaxial compression and tension tests were applied to multi-scale fibers reinforced rubberized concrete (MSFRRC) with added CaCO3 whiskers, steel and PVA fibers, after being subjected to high temperatures. The AE technique was used to monitor the damage evolution of MSFRRC. Specifically, the AE event location, peak frequency, b-value, AF-RA value, and AE energy were analyzed to discuss the impacts of high temperatures, CR, and multi-scale hybrid fibers on the AE characteristics of concrete.
2 Experimental setup
2.1 Raw materials and specimen preparation
MSFRRC is the combination of P.O.52.5 Portland cement, Grade-I fly ash, silica fume, quartz sand, water, CR with a density of 1100 kg/m3, CW, SF, and PVA fiber. The CR with particle size ranging from 0.4 to 0.8 mm is substituted for quartz sand by a 30% volume ratio in MSFRRC. The morphologies and physical properties of the various fibers are displayed in Fig.2 and Tab.1, respectively. Tab.2 lists the mixture proportions of MSFRRC used in this work. Prismatic test specimens with sizes of 100 mm × 100 mm × 300 mm and specimens with dog-bone shapes were cast for each mix. After demolding, the mixtures were placed in a standard curing box at the conditions of (20 ± 2) °C and 90% ± 5% humidity for 28 d. Three test specimens were investigated for each group to ensure the validity of the test results.
2.2 Heating and loading test
The test specimens were heated to 400 and 800 °C respectively from room temperature at a heating rate of 1 °C/min in the furnace and then kept for 2 h at the target temperature to ensure that each entire specimen attained the target temperature uniformly [
23,
24]. After cooling to room temperature, uniaxial compression and tensile tests were performed on heated specimens and on others that had been held at room temperature, taken to be 25 °C.
The uniaxial compression test was conducted using a servo-controlled compression machine, with a loading rate of 0.05 mm/min. The uniaxial tensile test was carried out using a universal testing machine MTS CMT5105 at the displacement rate of 0.2 mm/min. Two dial gauges were employed symmetrically to get the average displacement for the stress–strain curves. Meanwhile, the AE signal was collected throughout the loading process. The test setup is displayed in Fig.3.
2.3 Acoustic emission monitoring
The AE detection device used in this study was DS5-16B. Four AE sensors (RS-2A) were placed symmetrically on the surface of the specimen, and the sensors were coated with petroleum jelly as a coupling agent for better signal reception. The pre-gain was set as 60 dB, and the sampling frequency was 3 MHz. According to the actual environment, the threshold was 50–100 dB. As shown in Fig.1, the main AE parameters include amplitude, duration, rise time, and energy, which can help evaluate the damage evolution of the concrete during the compression and tension tests.
3 Test results and discussion
3.1 Mechanical properties
3.1.1 Compressive behaviors
The uniaxial compressive stress–strain curves of different specimens are displayed in Fig.4. It can be seen that the peak stress significantly reduces after incorporating CR. As shown, the stress decreases sharply after reaching its peak, which is due to the high strength of the concrete matrix. Adding 30% CR cannot change its brittleness property. However, incorporating multi-scale hybrid fibers can improve the brittleness of concrete and increase the peak stress of rubberized concrete. Fig.4(a) shows that the compressive stress–strain curve at room temperature for C3SPR30, for example, can be divided into three stages: (I) elastic stage, (II) elastic–plastic stage, and (III) post-peak stage. It is noted that there is no post-peak stage for Reference and R30, as the specimens without multi-scale fibers are destroyed completely at the peak stress point. Fig.4(b) shows the stress–strain curves for material exposed to high temperatures. It is important to note that explosive spalling occurs in specimens without fibers after heating to 400 °C. This is mainly caused by the internal vapor pressure inside the dense matrix. As the temperature rises, the elastic modulus and the peak stress decrease clearly. However, the peak strain increases, and the stress declines slowly at the post-peak stage, showing plastic failure. After being subjected to 800 °C, the specimen almost loses its load-bearing ability. It is divided into two stages: (I) pre-peak stage, and (II) post-peak stage. Interestingly, these stress–strain curves are related to the subsequent analysis of AE characteristics.
Fig.5 exhibits the average compressive strengths without and with being subjected to high temperatures. When 30% CR replaces quartz sand, the compressive strength of specimens decreases from 80.25 to 67.03 MPa. This is caused by the poor bond of the rubber to the cement matrix [
25]. After adding CW, steel and PVA fibers (as is the case for C3SPR30), the strength loss is compensated due to the bridging effect between the multi-scale hybrid fibers. After being subjected to 400 °C, the compressive strength decreases by about 39%, but remains at a high value, with a strength of 46.43 MPa for C3SPR30. After being heated to 800 °C, it reduces to 4.15 MPa.
3.1.2 Tensile behaviors
Fig.6 displays the tensile stress–strain curves without and with being subjected to high temperatures. As is the case for uniaxial compression, the incorporation of CR leads to a significant reduction in ultimate stress and elastic modulus. The specimen without multi-scale fibers fractures after reaching the ultimate stress, so there is no descending section of the curve. As shown, SPR30 and C3SPR30 exhibit better strain-hardening ability at room temperature. The stress can continue to increase after initial cracking. CW are the most effective in improving the strain-hardening of specimens. This is due to the fact that CaCO3 whiskers, together with steel and PVA fibers, inhibit cracking at both micro and macro scales. After reaching the ultimate stress, the stress decreases slowly with increasing strain, and it is accompanied by the sound of fibers pull-out. As shown in Fig.6(a), tensile stress–strain curves of MSFRRC can be classified into three stages: (I) elastic stage, (II) strain-hardening stage, and (III) strain-softening stage. The Reference and R30 have two stages: (I) elastic stage, and (II) elastic-plastic stage, similar to that under compression. Fig.6(b) illustrates the stress–strain relationship of test specimens after high temperatures. As in the case of compression, Reference and R30 experience explosive spalling when heating to 400 °C. In addition, the tensile ultimate stress significantly decreases with increasing temperatures. After being heated to 400 °C, the strain-hardening ability of specimens almost disappears. The stress–strain curve exposed to high temperatures is classified into two stages: (I) pre-peak stage, and (II) post-peak stage.
The average tensile strengths of specimens subjected to the three different maximum temperatures, 25, 400, and 800 °C, are shown in Fig.7. For unheated specimens (subjected to 25 °C only) incorporating CR reduces the tensile strength from 4.04 to 3.05MPa. However, adding multi-scale hybrid fibers significantly enhances the tensile strength, even exceeding that of the Reference. The tensile strength of SPR30 is the largest at 5.6 MPa. After being subjected to 400 °C, the tensile strength of SPR30 decreases by about 47% to 3.04 MPa, while that of C3SPR30 decreases by about 48.5% to 2.58 MPa. Multi-scale hybrid fibers can still play a role in providing tensile strength. However, after being heated to 800 °C, the tensile strength is almost completely lost, and is less than 1 MPa. Multi-scale fibers no longer have an effect, because of the oxidation and softening of SFs exposed to 800 °C, the decomposition of CaCO
3 whiskers at around 780 °C, and the melting of the PVA fiber at 230 °C [
11,
26].
3.2 Damage evolution based on acoustic emission events location
Fig.8 and Fig.9 display the damage evolution of specimens exposed to various temperatures under uniaxial compression and tension tests based on AE location. According to the stress–strain curves in Subsection 3.1, the damage evolution of different specimens can be regarded as having several typical stages.
3.2.1 Damage evolution under uniaxial compression
As shown in Fig.8(a), at room temperature, the number of AE events at the elastic stage is relatively small, and the damage to the specimen is light at this time. As the loading proceeds, AE signals increase significantly at the elastic–plastic stage, indicating an intensification of damage to the specimen. Meanwhile, the AE signals weaken with the incorporation of CR, especially at the elastic stage. This is because rubber has a lower elastic modulus and a better ability to absorb energy [
27,
28]. It can be seen that although adding 30% volume ratio of rubber does not change the brittleness of specimens, the addition can significantly reduce the degree of damage. Meanwhile, the numbers of AE events for SPR30 and C3SPR30 are larger than that for Reference and R30, mainly because the multi-scale fibers prevent the generation and development of cracks; those processes release large amounts of energy. In addition, the specimens without hybrid fibers exhibit brittle fractures after reaching peak stress, accompanied by vertical cracks. However, MSFRRC does not fracture immediately at the peak stress point due to the addition of multi-scale fibers, but the stress gradually decreases. The AE sources of MSFRRC continue to increase at the post-peak phase due to the action of fibers. Eventually, the specimens mixed with fibers show ductile failure, accompanied by a main diagonal crack.
After exposure to high temperatures, as shown in Fig.9(b), the number of AE events at stage I is significantly higher than is the case for specimen held at room temperature. Many micro-cracks appear in the specimen after heating to 400 °C, which gradually expand with the increase of stress. Furthermore, the elastic modulus of the specimen decreases significantly after high temperature experience, resulting in less energy storage and a large amount of release. In addition, the fibers can still play a role at the post-peak stage, although the effectiveness is reduced because PVA fibers melt at about 200 °C, resulting in a decrease in toughness [
29,
30]. After being heated to 800 °C, fibers are not efficient due to the oxidation of SF and the decomposition of CW [
26,
31], and damage to specimens occurs once loaded. Meanwhile, AE sources do not change significantly at the pre-peak stage and the post-peak stage. The damage evolution proceeds throughout the whole loading. Moreover, the fracture of MSFRRC after high temperature experience is relatively ductile, with more cracks appearing than in the unheated specimen. It is no longer a major diagonal crack, but multiple cracks at the ends.
3.2.2 AE location under tension
Fig.9(a) reflects the AE location of specimens under tension at room temperature. Compared to under compression, the AE events of specimens are relatively fewer under uniaxial tension. This is mainly because the loading time is relatively short, and the damage mainly occurs after the main crack appears. According to Subsubsection 3.1.2, the damage evolution process of specimens without fibers is divided into the elastic stage and the elastic-plastic stage. As is the case for behavior under compression, AE sources increase at stage II, indicating that the damage mainly occurs in this stage. The specimen without hybrid fibers is directly pulled apart when it reaches the ultimate stress. The effect of CR is not significant under uniaxial tension. After incorporating multi-scale fibers, the specimens show better strain-hardening ability. As seen, the increase in the AE signals at the elastic stage of SPR30 and C3SPR30 is caused by the multi-scale fibers bridging cracks which can prevent the generation and development of cracks, improving the tensile strength of specimens. In addition, the AE sources continue to increase at the strain-hardening stage and the strain-softening stage due to the pullout of fibers. As a result, the MSFRRC is not completely pulled apart.
As shown in Fig.9(b), after exposure to 400 °C, due to the disappearance of the strain-hardening phenomenon, the fracture process is subdivided into a pre-peak stage and a post-peak stage. The distribution of AE event locations of MSFRRC is wider, indicating that multiple micro-cracks are generated and developed, as occurs that under compression. Besides, the MSFRRC is not immediately pulled apart when the ultimate stress is reached. The AE sources increase significantly at the post-peak stage because multi-scale hybrid fibers still play a positive role in restraining cracks. After being subjected to 800 °C, the specimen is rapidly pulled apart once loaded, and fibers lose efficacy. In addition, the number of AE events decreases, and the damage occurs throughout the entire loading process.
3.3 Peak frequency
3.3.1 Peak frequency analysis under compression
Fig.10 and Fig.11 exhibit the peak frequency distribution of specimens and their ratios relative to all frequencies in different stages under compressive loading, exposed to room temperature and high temperatures. As shown, the AE peak frequency is mainly distributed in the six ranges: < 20 kHz (ultra-low frequency), 20–70 kHz (low frequency), 70–120 kHz (medium-low frequency), 120–170 kHz (medium frequency), 170–190 kHz (medium-high frequency), and > 190 kHz (high frequency) [
32]. At room temperature, the peak frequency is mainly concentrated in the medium-low frequency and medium frequency range at the elastic stage. As the loading proceeds, the AE signal increases in the low frequency and ultra-low frequency ranges at the elastic–plastic stage, indicating matrix cracking and crack propagation until the whole specimen fractures. Compared with the Reference, the peak frequency of R30 tends to show a wider frequency distribution. This may be due to the debonding of the CR. In addition, before the stress reaches its peak, the peak frequency ratio of MSFRRC in the high-frequency range is greater than that of specimens without hybrid fibers, which is due to the fibers bridging each other and inhibiting the development of cracks. Moreover, the peak frequency ratio of the ultra-low frequency range of MSFRRC increases significantly at the post-peak stage. This is because specimens with multi-scale fibers do not fail at the peak stress point, and the cracks continue to expand until they penetrate the entire specimen.
After having been heated, as the damage to the specimen intensifies, the AE sources increase significantly, releasing large amounts of energy. Meanwhile, the peak frequency ratio of the medium frequency range increases clearly, mainly due to the debonding of fibers and the extension of initial cracks caused by high temperatures. After having been subjected to 800 °C, as there is more damage to the specimen, the peak frequency of the low-frequency range increases at the beginning of loading. The damage occurs throughout the entire loading process until the specimen fractures. In general, the distribution of peak frequencies increases in the low-frequency range as the cracks develop gradually and the degree of damage deepens.
3.3.2 Peak frequency analysis under tension
The AE peak frequency distribution under tensile test at various temperatures is shown in Fig.12 and Fig.13. The peak frequency under uniaxial tension is mainly concentrated in the following six ranges: < 50 kHz (R1), 50–150 kHz (R2), 150–200 kHz (R3), 200–250 kHz (R4), 250–350 kHz (R5), and > 350 kHz (R6), which is different from behavior under uniaxial compression [
33,
34]. It should be noted that the peak frequency in R6 (> 350 kHz) may be primarily affected by the tensile testing machine. The electromagnetic waves from electronic testing machines interfere with the AE signals. As can be seen, with the loading proceeding, the damage to the specimen aggravates, and the peak frequency distribution develops toward a lower frequency. At room temperature, after adding the CR, the proportion of peak frequency in R5 slightly increases at the elastic stage, possibly due to the delamination of CR, which is similar to behavior under compression. In addition, after adding multi-scale fibers, specimens exhibit better strain-hardening ability, especially for the specimens with added CaCO
3 whiskers. The proportion of peak frequency in R4 increases significantly, which is caused by the pulling out of fibers. In addition, the AE peak frequency ratio in R5 further increases. At the strain-softening stage, the AE peak frequency in R1 and R2 increases clearly due to the matrix cracking, the extension of the main crack, and the creation of several microcracks. The specimens are not eventually pulled apart because of fibers.
As shown in Fig.13, after being subjected to high temperatures, the peak frequency ratio in the lower frequency range significantly increases. The strain-hardening phenomenon of SPR30 and C3SPR30 almost disappears. Meanwhile, at the pre-peak stage, the distribution of AE peak frequency tends to become wider in R1, which is due to severe damage inside the specimen caused by high temperatures. After 800 °C, the specimen can no longer bear the load and is rapidly pulled apart, resulting in a decrease in AE signals. In summary, as is the case for behavior under uniaxial compression, the development of increasing damage correlates with increase of the proportion of peak frequency that is at a lower frequency range. However, the difference is that the frequency under uniaxial tension is generally higher than that under uniaxial compression. On the one hand, this may be caused by the influence of electromagnetic waves from the machine under tension. On the other hand, the pulling apart of fibers during the tensile process produces a higher peak frequency [
34].
3.4 Analysis of b-value
The b-value is related to crack propagation in concrete [
35]. After the b-value is modified in seismology, it can be defined in AE analysis as follows [
36–
38]:
where AdB is the amplitude of the AE event; N is the number of AE events with amplitudes greater than AdB.; a and b (b-value) are the intercept and slope of the linear fitting, respectively.
3.4.1 Analysis of b-value under compression
Fig.14 exhibits the b-value of specimens under uniaxial compression. As the loading progresses, the b-value gradually decreases, indicating that the microscopic cracks gradually develop into macroscopic cracks and the damage inside the specimen intensifies. This is similar to the results of the peak frequency analysis. In the early stage of loading, the b-value is high, reflecting that the damage is mainly caused by micro-cracks. After reaching the peak stress, the b-value suddenly decreases, and the b-value of the reference specimens decreases to the minimum. The damage tends to be caused by macroscopic cracks. In addition, after incorporating the rubber and the multi-scale fiber, the maximum b-value increases, indicating that rubber and fibers play a positive role in preventing microcrack propagation. Moreover, the b-value of MSFRRC decreases relatively slowly at the post-peak stage, and the minimum b-value is greater than that of Reference and R30. This indicates that multi-scale fibers reduce the damage to concrete. After heating followed by cooling, the maximum b-value decreases significantly, which is due to the large number of initial cracks inside the specimen caused by being subjected to high temperature. Furthermore, the decrease in the b-value becomes slower, and the b-value changes more smoothly. On the one hand, cracks inside the specimen expand steadily. On the other hand, exposure to high temperature significantly deteriorates the elastic modulus of specimens, resulting in a decrease in energy release.
3.4.2 Analysis of b-value under tension
Fig.15 displays the b-value of specimens under tension. Compared to behavior under uniaxial compression, the b-value fluctuates significantly throughout the entire loading process. For specimens not heated above room temperature, is the case under compression, the minimum b-value of reference specimens is the smallest, implying the most severe damage. When reaching peak stress, the b-value declines suddenly, and macroscopic cracks appear. The b-value of MSFRRC increases slightly after reaching peak stress due to the pulling-out of the fibers, which inhibits cracking. However, as the macro-cracks further develop, the b-value continues to decrease. After being subjected to high temperatures, at the beginning of loading, the b-values of concrete are relatively low, probably due to the damage to the specimen caused by the high temperature exposure. Subsequently, the b-value first increases and then decreases, attributed to the influences of multi-scale fibers. As shown, the sudden drop in the b-value occurs before the peak stress point, as macroscopic cracks already appear before reaching peak stress. In addition, the minimum b-value is significantly lower than for the unheated specimens. This reflects the intensified damage to the concrete caused by high temperatures, resulting in more severe cracking under loading.
3.5 The ratio of rise time to amplitude-average frequency value
Cracks are generally classified as tensile cracks and shear cracks. These two types of cracks can be identified using AF and RA values in AE technology [
39]. The AF and RA values are regarded as the ratio of ringing count to duration (kHz) and the ratio of rise time to amplitude (μs/V), respectively [
40]. According to previous research, the AE events of tensile cracks are short in duration and release large amounts of energy very rapidly, with high AF values and relatively low RA values. In contrast, the AE events of shear cracks have a longer duration and the AE signals increase slowly. The RA value of shear cracks is high [
41–
43].
3.5.1 Analysis under compression
The AF and RA relations of different specimens heating to various temperatures under compression are shown in Fig.16. Fig.17 reflects the variation of RA values with the loading process. At room temperature, the RA value gradually increases as loading progresses, reflecting the evolution of micro-cracks of concrete toward macro-cracks. Before the major crack is generated, the cracks are mainly caused by matrix cracking, manifested as tensile cracks. The RA value of Reference increases slightly at the elastic–plastic stage, indicating that the crack type evolves toward shear cracks. After adding the rubber and multi-scale fibers, the RA values are smaller at the elastic and elastic–plastic stages and increase rapidly at the peak stress point. This can be attributed to the effects of the rubber and hybrid fibers in inhibiting crack propagation and delaying the formation of macroscopic cracks, indicating that the formation of shear cracks is delayed. In addition, the b-value of MSFRRC remains high at the post-peak stage because the MSFRRC can continue to be loaded and the cracks continue to develop. In addition, there is friction between the multi-scale fibers and the matrix. After exposure to high temperatures, before reaching peak stress, the RA value increases significantly compared to that for specimens not heated above room temperature, implying the earlier formation of shear cracks. This is because the specimen is damaged by high temperature, with more internal cracks and early formation of the major crack. During the post-peak stage, the RA values decrease slightly, probably due to the failure of the fibers caused by high temperatures.
Fig.18 exhibits the classification of tensile cracks and shear cracks. The ratio of the maximum AF value to the maximum RA value, regarded as k1, proposed by Nguyen-Tat et al. [
44], is used to classify tensile and shear cracks. The ratio of AF to RA greater than k1 is considered as a tensile crack, otherwise, it is a shear crack. As shown, the specimen eventually fractures mainly in the form of shear cracks. The incorporation of CR reduces shear cracks. However, after adding multi-scale fibers, shear cracks increase because of the interaction of fibers with the concrete matrix. Furthermore, the ratio of shear cracks decreases slightly with increasing temperature, possibly due to the weakened interaction between the fibers and the matrix after high temperatures.
3.5.2 Analysis under tension
Fig.19 exhibits the RA and AF relations under uniaxial tension at different temperatures. The variation of RA values with the tension process is displayed in Fig.20. The change in RA values under uniaxial tension is not significant compared to that under uniaxial compression due to the interference of the electronic testing machine. However, it can still be seen that adding CR leads to an overall increase in the RA value, but the maximum RA value is slightly reduced. This is because the rubber has a better ability to absorb energy, and the rise time of the AE signal becomes longer. After adding multi-scale fibers, the maximum RA value further decreases. This indicates that the fibers and CR inhibit the creation and development of cracks during the loading process, and slow down the damage evolution of concrete. In addition, the RA values of the specimens without fibers decrease at the peak stress due to the pulling apart of the specimen. However, the RA values of MSFRRC increase slightly after reaching the peak stress, which can be attributed to the pulling out of the fibers and the extension of macroscopic cracks. After being subjected to 400 °C, the AF values increase significantly and the RA values decrease relatively. Specimens exhibit more tensile cracks and are eventually pulled off as the loading progresses. After being subjected to 800 °C, the specimen is pulled off once loaded. The maximum RA value increases due to the severe damage to the specimen caused by high temperatures.
In the uniaxial tension test, k2 = RA/AF = 50, is applied to identify tensile and shear cracks [
36]. A ratio of RA to AF less than 50 is regarded as identifying a tensile crack. As shown in Fig.21, after adding multi-scale fibers, the tensile cracks significantly increase mainly due to the effect of fiber pulling out, which inhibits the development of macroscopic cracks. As the temperature increases, shear cracks increase mainly because of the internal damage to the specimen caused by the exposure to high temperatures. Overall, as the damage evolves, cracks develop toward the macroscopic scale, and shear cracks of concrete increase.
3.6 Acoustic emission energy analysis
3.6.1 Acoustic emission energy of specimens under compression
AE energy is the area under the envelope of the AE signals [
45]. The AE energy is closely related to the AE characteristics. Fig.22 shows the AE energy of specimens under compression, exposed to room temperature, 400 and 800 °C. For specimens not heated above room temperature, in the elastic and elastic–plastic stages, the AE energy is relatively low. When the specimen is damaged, large amounts of energy are released, reaching a maximum at the peak stress point. Adding CR reduces the maximum value of AE energy due to its low elastic modulus and good energy absorption ability. In addition, the AE energy of MSFRRC still remains at a high level at the post-peak stage because multi-scale fibers inhibit the extension of cracks, releasing large amounts of energy. After being subjected to high temperatures, at the beginning of loading, the release of AE energy increases. This is because the number of internal cracks inside the specimen increases when exposed to high temperatures, and these internal cracks expand continuously with uniaxial compressive loading until macro-cracks are formed, resulting in specimen failure. After being subjected to 800 °C, the maximum value of AE energy is significantly reduced, probably due to the significant deterioration of the uniaxial compressive strength and the reduction of elastic modulus.
3.6.2 Acoustic emission energy of specimens under tension
Fig.23 displays the AE energy of concrete under the uniaxial tension test after being subjected to various temperatures. For room temperature specimens, similar to behavior under compression, relatively little energy is released in the early stages of loading. The AE energy increases significantly at the peak stress point. The maximum value of AE energy for R30 decreases due to CR. In addition, the AE energy of MSFRRC increases significantly during the strain-hardening stage and remains high at the strain-softening stage due to the pulling-out of the fibers, which inhibits the crack extension. After being subjected to 400 °C, the release of AE energy during the pre-peak stage is reduced, which is different from the behavior observed under compression. This may be caused by the significant degradation of tensile strength and elastic modulus. Compared to behavior of unheated specimens, after reaching the peak stress, the AE energy relatively decreases, which may be due to the damage caused to the fibers by high temperature exposure, resulting in a weakened effect of fibers in suppressing cracks. After being subjected to 800 °C, the AE energy further decreases because of the severe damage to the specimen. The specimen is pulled apart as soon as it is loaded, resulting in fewer AE events.
4 Conclusions
In summary, the influences of CR and multi-scale fibers on AE characteristics of specimens under uniaxial compression and tension test after exposure to room temperature, 400, and 800 °C are analyzed. The mechanical properties, AE event location, peak frequency, b-value, RA and AF values, and AE energy are discussed. The main conclusions can be drawn as follows.
1) Under compression and tension, adding CR significantly deteriorates concrete strength, while multi-scale hybrid fibers can compensate for the strength loss and enhance the ductility of concrete. MSFRRC shows better strain-hardening ability under tension when unheated than specimens without fibers. However, after being subjected to high temperatures, the strain-hardening ability disappears and the compressive strength and tensile strength decrease significantly, although the fibers can still play a positive role in cracking resistance after exposure to 400 °C.
2) AE events localization can better reflect the damage evolution in concrete. As the loading proceeds, the AE events increase significantly at the elastic–plastic stage, indicating that the damage to the specimen intensifies and cracks develop rapidly during this stage. Multi-scale fibers effectively improve the brittleness of concrete, leading to ductile failure. After high temperature exposures, there is a significant increase in the number of AE events at the elastic stage due to the initial cracks of the specimen caused by being subjected to high temperatures.
3) As cracks develop from the micro scale to the macro scale, the peak frequency tends to be increasingly distributed toward a lower frequency range. When CR and multi-scale fibers are incorporated, the peak frequency ratio in the medium frequency and high frequency range increases, indicating the inhibition of crack generation and development. After high temperature exposures, the ratio of medium frequency increases significantly. This is due to the debonding of fibers and the extension of initial cracks caused by high temperatures.
4) At the peak stress, the b-value decreases significantly, and macroscopic cracks extend. The incorporation of CR and multi-scale fibers increases both the maximum and the minimum b-value, restraining crack propagation. In addition, the high temperature exposure significantly reduces the maximum b-value due to the increase in initial cracks inside the specimen. The variation in b-value is more pronounced under tensile loading than under compression.
5) The RA and AF values can be used to classify tensile and shear cracks. Under uniaxial compression, the RA values gradually increase as the damage evolves, and specimen fracture involves shear cracks. CR and multi-scale fibers inhibit shear crack propagation. After high temperature exposures, the maximum RA value increases before the peak stress. Under uniaxial tension, the change in RA value is not significant, and the fracture of MSFRRC is mainly accompanied by tensile cracks. However, after being subjected to high temperatures, the proportion of shear cracks increases.
PDF
(30184KB)
