1 Introduction

The high energy consumption of producing cement and recycling of industrial waste has been widely concerned by researchers [1], promoting the development of environmental-friendly and sustainable construction materials, including novel concrete and reinforcement bar. Self-compacting concrete (SCC) prepared by replacing cement with industrial by-products, such as fly ash, limestone powder and ground granulated blast furnace slag, is a green concrete that can flow and fill by its own weight without any vibration [2, 3]. In these reported studies [4, 5], SCC has been adopted in actual engineering due to its advantages, such as low energy consumption, low-cost, vibration-free and fast speed in construction. It has been well reported that the performance of reinforced concrete structure is deteriorated due to the corrosion of steel material [6,7,8]. The durability problem of reinforced concrete structures has become a major concern for researchers and engineers. Fiber-reinforced polymer (FRP) reinforcement, which has the advantages of corrosion resistance, high tensile strength and light weight, has been studied as substitutes for traditional steel bars [9,10,11]. The structures constructed by combining FRP and SCC materials not only achieve sustainable construction, but also have a better service performance [12, 13].

The bond behavior is a significant factor affecting the serviceability and failure mechanism of FRP reinforced concrete structures [14]. Mazaheripour et al. [15] pointed out that the accurate bond relationship of GFRP-concrete was necessary to develop a reasonable design equation of concrete members reinforced with GFRP. A series of studies on the bond behavior of concrete and FRP bars have been carried out [16,17,18,19]. Baena et al. [20] investigated the bond behavior of FRP and concrete, and the test results revealed that the surface treatment of FRP bars was a key factor affecting bond mechanisms and bond strength. Gu et al. [14] found that the glass-FRP (GFRP)-reinforced specimens provided poor bond strength compared to steel reinforced specimens. Due to the influence of binding materials and aggregate size on the bond behavior, some scholars have conducted research on the bond behavior of SCC and FRP bars in recent years. Helincks and Pop et al. [21, 22] showed that the bond strength of SCC was the same as or higher than that of vibrated concrete. The bond behavior of steel and GFRP reinforcement in self-compacting concrete was studied by Golafshani et al. [23], and the results indicated that the bond behavior of steel bars was better than GFRP bars in both normal and self-compacting concrete. However, the bond behavior between FRP reinforcement and SCC materials has not been studied fully. In addition, it is difficult to identify the damage mechanism and failure mode of interface between reinforcing bars and concrete materials by using the traditional test methods. As a result, it is difficult to establish an accurate theoretical model for the prediction of the bond-slip relationship and to investigate the compatibility between FRP reinforcement and SCC materials. Therefore, a novel test method using piezoceramic transducers and wavelet energy analysis was proposed to identify the damage mechanism and failure mode of the bond between FRP bars and SCC.

Piezoceramic materials, which have the advantages of strong piezoelectric effect [24], high sensitivity [25], fast response [26], wide frequency range [24], energy harvesting [27] and low cost [28], have been actively adopted in structural health monitoring (SHM) in recent years [29]. The experimental studies on the monitoring of bolt loosening were carried out [28, 30, 31], and the test results indicated that the looseness of bolted connection can be monitored by using piezoceramic transducers. Besides, the investigations of using piezoceramic transducers on the damage monitoring of reinforced concrete structures show that the damage of the structures can be effectively identified [32, 33]. The test results of the study of the bond behavior between steel plates/tubes and concrete demonstrated that the debonding can be detected by the amplitude variation of the received signal using piezoceramic transducers [34,35,36]. Xu et al. [37] found that the bond slip between the GFRP bar and the concrete can be accurately captured by using piezoceramic-based active sensing approach. Recently, it was reported that the debonding failure between FRP reinforcement and concrete was identified and the damage condition could be detected by using piezoceramic transducers [38]. However, the damage mechanism and failure modes of concrete structures reinforced with steel/FRP bar have not been investigated and evaluated in these reported study [38]. In addition, the bond mechanism of reinforced concrete structures is affected by the surface treatment, type of reinforcing bars and concrete strength [20]. Thus, it is meaningful to identify the bond-slip damage mechanism and failure modes of concrete, especially for SCC structures reinforced with different reinforcements by using piezoceramic transducers.

To investigate the bond behavior between FRP and SCC, a wavelet energy ratio index (WERI) is proposed in this paper to identify the damage mechanism and failure modes of the interface between reinforcing bars and surrounding SCC by using piezoceramic transducers. In this test, two piezoceramic transducers were used for an active sensing approach. One piezoceramic patch transducer acting as an actuator was bonded on the outer surface of reinforcing bars, and the other piezoceramic smart aggregate (SA) transducer acting as a sensor was embedded in the SCC. The SCC specimens with different reinforcement materials were prepared in the experimental investigation. The test variables involved the type of reinforcing bar (GFRP bars, Basalt-FRP bars and deformed steel bars) and the type of concrete materials (fly ash content accounts for 0%, 50% and 65% of binding materials). A series of pull-out tests were carried out in this study and the active sensing approach was implemented. The test results reveal that the damage mechanism of bond behavior between reinforcing bars and concrete can be identified effectively, and the peak of bond stress can be captured accurately based on the proposed WERI values. This study produces a feasibility to identify the FRP–SCC bond behavior by using the piezoceramic transducers and wavelet energy analysis.

2 Piezoceramic transducers and wavelet energy analysis

2.1 Piezoceramic transducers and active sensing approach

Piezoceramic materials have been developed into various types of transducers, such as smart washer and smart aggregate (SA), for Structural Health Monitoring (SHM) and damage detection due to their direct and converse piezoelectric effects [28, 39]. Lead zirconate titanate (PZT) is a commonly used piezoelectric material [33], and a PZT (d33) disk is shown in Fig. 1. The SA is designed by sandwiching a PZT disk between copper sheets and then embedding in two machined marble blocks [40, 41], as shown in Fig. 2. Due to the mechanical protection and waterproof capability offered by the marble blocks, the SA transducers can be embedded inside the concrete structure to study the damage detection [42, 43].

Fig. 1
figure 1

Piezoelectric transducers

Full size image
Fig. 2
figure 2

Smart aggregate

Full size image

The active sensing approach, as a common way to use SAs in SHM and structural damage detection, can be simply achieved by using at least a pair of SAs: one acting as an actuator to generate the stress wave that propagates along the structure and the other one acting as a sensor to detect the stress wave [44]. The stress wave is emitted by exciting the SA actuator and then propagates along the structure [33]. The SA sensor detects the propagating stress wave. Since the propagation path and dissipation of the energy of the stress wave are affected by the structural damage level, the healthy state can be interpreted by the variation in the waveform and the amplitude of the received signals by using different signal processing techniques, including time reversal methods [45], and wavelet-based methods [29].

2.2 Wavelet energy analysis

Wavelet analysis is a common method for analyzing the received signals by piezoceramic transducers in SHM, and this method has been used in these reported studies [38, 46]. In a wavelet analysis [29], the signal is decomposed into a detail containing high frequency information and an approximation containing low frequency information. The approximation obtained by the first decomposition is decomposed into a second-level detail and approximation, and the decomposition process is repeated. The 3-level wavelet decomposition diagram is shown in Fig. 3. The advantage of wavelet decomposition is that it can clearly reflect the information in different frequency and time bands. In addition, the suitable frequency and time bands information can be selected for more in-depth analysis.

Fig. 3
figure 3

The 3-level wavelet decomposition diagram

Full size image

The received signal S is decomposed into n + 1 frequency bands by a n-level wavelet decomposition [47], and the signal set \( X_{j} \) can be expressed as:

$$ X_{j} = \left[ {X_{j,1} ,X_{j,2} , \ldots ,X_{j,m} } \right] $$
(1)

where j is the frequency band (j = 1, 2,……, n + 1) and m is the amount of sampling data. The energy \( E_{i,j} \) of each signal set can be calculated as:

$$ E_{i,j} = \left\| {X_{j} } \right\|_{2}^{2} = \sum\limits_{k = 1}^{k = m} {X_{j,k}^{2} } { = }X_{j,1}^{2} + X_{j,2}^{2} + \cdots \cdots + X_{j,m}^{2} $$
(2)

where i is the time index. Based on energy \( E_{i,j} \) of each signal set, the total energy \( E_{i} \) of the ith received signal can be defined as:

$$ E_{i} = \sum\limits_{j = 1}^{j = n + 1} {E_{i,j} }\, { = }\,E_{i,1} + E_{i,2} + \cdots \cdots + E_{i,n + 1} $$
(3)

To accurately reflect the damage mechanism of the interface between concrete and reinforcing bars in the pull-out test, a wavelet energy ratio index (WERI) K is proposed as follows:

$$ K = {{E_{i} } \mathord{\left/ {\vphantom {{E_{i} } {E_{0} }}} \right. \kern-0pt} {E_{0} }} $$
(4)

where \( E_{0} \) represents the total energy of the received signal in the initial state, i.e., the healthy state.

In this paper, the received signal is decomposed into 4 frequency bands using db4 wavelet by using 3-level wavelet decomposition. The analysis results find that more noise components exist in the high frequency band and other non-interface damage components present in the low frequency band [48], and the secondary low frequency band (D3 frequency band in Fig. 3) is adopted to analyze the damage mechanism of the bond-slip between concrete and reinforcing bars.

3 Experimental program

3.1 Principle of damage detection of bond-slip behavior

The bond strength between reinforcing bars and SCC is mainly derived from friction, chemical adhesion and mechanical interlock resistance of the interface [49]. The failure of mechanical interlock bond resulting from the steel rebar and concrete leads to the concrete being split (low strength of concrete or insufficient thickness of protective layer) or the steel bar being pulled out (the concrete around the interface being shear damage due to the high elastic modulus and the shear strength of steel bar) [50, 51]. Muñoz [52] showed that the bond behavior and damage mechanism of FRP bars and concrete were different from that of traditional steel bars owing to the low elastic modulus and shear strength of FRP bars. The main bond failure models of FRP bars-to-concrete include: (a) shearing off the surface deformations of the bar. (The bond strength is controlled by the shear strength of continuous interlayer fibers.) This is due to the lower shear strength of continuous interlayer fibers comparing to concrete materials; (b) concrete shear failure (the bond strength depends on concrete shear strength). This is attributed to the lower shear strength of concrete materials comparing to continuous interlayer fibers; (c) combined mode of the first two cases. This is caused by the fact that the shear strength of continuous interlayer fibers is similar to that of concrete materials; (d) squeeze through (the bond strength is determined by the stiffness in the radial direction of the FRP bars).

As a result, three major damage modes can be found in the bond test and are shown as: (a) concrete damage around the interface; (b) interface damage of the reinforcing bars; (c) debonding damage between reinforcing bars and concrete. To accurately monitor the interface damage in the bond-slip process using piezoceramic transducers, a PZT patch transducer is bonded on the outer surface of reinforcing bars and a SA transducer is embedded in concrete (see Fig. 4). The stress wave is generated by the actuator propagates through the interface between the reinforcing bar and the concrete. As damage in the interface occurs, the received stress wave energy is attenuated. The attenuation degree of the received stress wave is related to the damage level of the interface. The propagation of stress wave under different interface damages is shown in Fig. 4.

Fig. 4
figure 4

The propagation of stress wave under different interface damages

Full size image

3.2 Material properties and specimen fabrication

The aim of this study is to investigate the influence of concrete and reinforcing materials on the bond behavior and to identify the damage mechanism and failure modes using wavelet energy ratio index and the active sensing technology. The effect of other structural variables, such as FRP bar diameter and embedment length, will be investigated in the future study. As shown in Table 1, the concrete materials used in the test were prepared by using different binding materials. The good workability was presented in prepared concrete, and the flowing property of SCC can meet the requirement of corresponding standards [53]. Three cube specimens with a size of 150 mm for each group of concrete materials are used for the compressive test, and the test results are shown in Table 2. Figure 5 shows the geometric shape of the adopted reinforcing materials. Three specimens for each group of reinforcing bars are used for the monotonic tensile test to obtain the corresponding material properties (see Table 3).

Table 1 The compositions of concrete materials
Full size table
Table 2 Mechanical properties of concrete materials
Full size table
Fig. 5
figure 5

The geometric shape of reinforcing bars

Full size image
Table 3 Mechanical properties of reinforcing bars
Full size table

Ten cube specimens with a size of 200 mm, divided into five series, were prepared in the test to study the bond behavior between reinforcing bars and concrete. The test variables include the type of reinforcing bars and the type of concrete materials. The pull-out specimen characteristics of the five series are detailed in Table 4. As shown in Fig. 6, the embedment length for all the test specimens is 65 mm, which is determined by five times of the reinforcing bar diameter [54]. The non-bonded area is protected by two PVC pipes to isolate the concrete materials. The concrete was poured with the FRP bars in position inside the mold, in the middle of the specimen. All the test specimens were demolded after pouring concrete mixture for 24 h and then placed in a curing room at a temperature of 20 ± 2 °C and a humidity of 95% for 56 days [55].

Table 4 The characteristics of pull-out specimens
Full size table
Fig. 6
figure 6

The embedment length and position of reinforcing bars

Full size image

3.3 Transducer arrangement and test setup

To identify the bond-slip behavior of interface between reinforcing bars and SCC, a PZT patch transducer was bonded on the outer surface of reinforcing bars as an actuator and an SA transducer was embedded in SCC for acting sensor (see Fig. 7). The SA was placed adjacent to the interface between reinforcing bars and SCC for effectively monitoring interface damage. However, it is worth noting that the SA does not touch the surface of the reinforcing bar.

Fig. 7
figure 7

The arrangement of piezoceramic transducers

Full size image

The supporting equipment of SA transducers consists of a data acquisition device (NI USB 6366), a power amplifier and a personal computer (PC), as shown in Fig. 8. The device of data acquisition is used to generate or receive signals, and the power amplifier is used to amplify the generated or received signals. In the test, the start frequency and end frequency of sweep sine signals are corresponding to 100 Hz and 100 kHz, respectively [37, 38, 56]. The voltage amplitude and sweep period are 5 V and 1 s [57]. The force datum was acquired synchronously with the displacement datum collected by TDS-530 device.

Fig. 8
figure 8

The data acquisition system

Full size image

The direct pull-out test setup is shown in Fig. 9. The loading device used in the test is a microcomputer-controlled electro-hydraulic servo universal testing machine with a capacity of 1000kN. The test scheme was carried out according to ACI 440.3R-04 [54] and CSA S806-02 [58] standards. All the monotonic pull-out tests were performed under displacement control (1.2 mm/min) in order to obtain post-peak behavior of bond-slip curves. Two linear variable differential transformers (LVDTs) were symmetrically placed at the unloaded end of the test specimens with the aim of avoiding eccentric effects during loading.

Fig. 9
figure 9

Loading of the test specimen

Full size image

4 Experimental results and discussion

4.1 Bond behavior analysis of the test results

The average bond stress, namely \( \tau \), under different loads, is calculated by,

$$ \tau = \frac{F}{\pi dl} $$
(5)

where F is the applied tensile force, d and l are the diameter and the embedded length of the reinforcing bar. The bond-slip curves of the average bond stress versus the free end displacement of reinforcing bar are plotted in Fig. 10. It can be summarized that the two specimens of the same series are labeled name-1 and name-2. The failure mode of all specimens is pullout failure. The bond-slip curves are presented by three phases: (a) a linear (elastic) ascending phase; (b) a nonlinear (damaged) ascending phase; (c) a softening phase followed the peak bond stress.

Fig. 10
figure 10

Bond-slip curves

Full size image

As shown in Fig. 10a, the stiffness and the bond strength of the SCC-50%-steel series in this test are higher than that of the SCC-50%-GFRP and SCC-50%-BFRP series. This is mainly attributed to the high elastic modulus of steel reinforcement and the high mechanical interlock resistance of the interface between steel bars and SCC. However, the lower bond strength is shown in the SCC-50%-GFRP and SCC-50%-BFRP series, which is mainly related to the elastic modulus and surface treatment of BFRP/GFRP bars. Although the similar peak bond stress is presented in the SCC-50%-BFRP and SCC-50%-GFRP series, the SCC-50%-BFRP series exhibits a larger slippage compared to the SCC-50%-GFRP series at the same stress (before peak bond stress). This is caused by the difference of synthetic raw materials and surface treatment between BFRP and GFRP reinforcement. It is remarkable that the minimal slippage is present under the peak bonding stress for the SCC-50%-GFRP, SCC-65%-GFRP and NC-GFRP series.

As shown in Fig. 10b, the difference in bond behavior before the peak bond stress is not significant in the SCC-50%-GFRP and SCC-65%-GFRP series. This indicates that the bond behavior before the peak bond stress and bond strength is independent of the replacement of fly ash in SCC. It can be inferred that the debonding damage appears in the reinforcing bars when the interface bond is failed in the SCC-50%-GFRP and SCC-65%-GFRP series. However, the load required for the SCC-65%-GFRP series is smaller at the same slippage (after the peak bond stress), which could be explained by the fact that low-strength concrete is more susceptible to damage. It can be noted that the higher bond strength occurs in the SCC-50%-GFRP and SCC-65%-GFRP series rather than the normal concrete (NC-GFRP) series, which may be due to the strong chemical cementation between GFRP reinforcement and SCC.

4.2 Time-domain and frequency-domain analysis of the bond-slip behavior

The acquisition timing of piezoelectric signals in the pull-out test is shown in Table 5. Figure 11 shows the acquisition timing of piezoelectric signals on the force-slip curve, where the dots on the slip-force curve represent the signal acquisition position.

Table 5 The acquisition timing of piezoelectric signals in the pull-out test
Full size table
Fig. 11
figure 11

The acquisition timing of piezoceramic signals on the force-slip curve

Full size image

The received time-domain signals and the frequency spectrograms obtained by using fast Fourier transform (FFT) are presented in Fig. 12. It can be found that the signal amplitude in time-domain and the signal energy in the frequency spectrogram of SCC-50%-steel-1 are drastically reduced when the peak bond stress is reached (see Fig. 12a, b), which is attributed to the significant damage in the SCC introduced by the steel bar thread at the ultimate bond strength. As shown in Fig. 12c, e, g, and i, the received signals in time domain have slightly changed during the pull-out process between FRP reinforcement and concrete, while the energy of these signals is gradually attenuated in frequency spectrograms in the range of 120–200 kHz. This implies that the interface damage status can be assessed by the frequency spectrograms in 120–200 kHz. In addition, it provides a reference for selecting the secondary low frequency band (D3 frequency band) in wavelet energy analysis.

Fig. 12
figure 12figure 12

Time-domain/frequency-domain signals

Full size image

4.3 Wavelet energy ratio index (WERI) of different reinforcing bars

To evaluate the bond damage mechanism and predict the occurrence of the peak bond stress of the interface between different reinforcing bars and SCC in the pull-out test, the WERI values are calculated, as shown in Fig. 13.

Fig. 13
figure 13

The WERI values and bond-slip curves of different reinforcing bars

Full size image

For the SCC-50%-steel series, as shown in Fig. 13a, b, the WERI values are fluctuated slightly before the peak bond stress. In addition, the WERI values are decreased rapidly or even close to zero at the corresponding load level of the ultimate bond strength. This illustrates that the stress wave propagation is barely attenuated and the energy of received signal is scarcely varied at a small slippage due to the good mechanical interlock between the steel bar and the SCC. However, the energy of the received signal is significantly reduced when the bond stress approaches or reaches the peak value. This is due to the fact that the failure of mechanical interlock results in significant damage of the interface between SCC and reinforcing bar. Additionally, the stress wave propagation is severely hindered by the concrete crack on the interface. Therefore, the failure of mechanical interlock (ultimate bond strength) of the between steel bars and concrete can be accurately captured by using the WERI values.

As reported in the literature [59], the bond behavior of FRP reinforced concrete structures is dependent on chemical adhesion and friction but not mechanical interlock, which is different from the test results of steel reinforced SCC test specimens. As shown in Fig. 13c, d, the WERI values for the SCC-50%-BFRP series are decreased by increasing the slippage in the initial stage. It can be summarized that the smooth surface of BFRP bars (see Fig. 5) results in lower chemical adhesive capability of the interface and the smaller friction, which can be also seen in the bond-slip behavior shown in Fig. 10. Interestingly, the WERI values are stabilized at the loading level of around the peak bond stress. This is due to the complete loss of chemical adhesion and the serious surface damage of BFRP reinforcement (lower shear strength of the BFRP bars compared to SCC) in the pull-out test.

For the SCC-50%-GFRP (see Fig. 13e, f), the WERI values are also increased up to the ultimate bond strength. This implies that the chemical adhesion does not fail before the bond stress reached the peak value. Thereafter, WERI values are reduced owing to the failure of bond behavior and the falling off of sand coatings around the reinforcing materials. This indicates that the bond strength of local (chemical adhesion) damage can be determined based on the point A of WERI values. Interestingly, the development of WERI values from point A to point B is different in two test specimens (see Fig. 13e, f). This difference could be due to the effect of sand particles on the interface adherence by a mechanical anchoring phenomenon [60].

In summary, the effect of different reinforcing materials on the bond loading transferring and failure mechanism can be effectively detected based on the variation of WERI values. Further, the peak of bond stress (the failure of bond behavior) can be also identified by using the WERI values. Therefore, it is feasible to evaluate the bond behavior and damage failure of different reinforcing bars and SCC using piezoceramic transducers.

4.4 Wavelet energy ratio index (WERI) of different concrete materials

It has been well reported in the literature [23] that the concrete matrix material is the one of the factors affecting the bond behavior due to the distinction of internal structure characteristic. To study the effect of concrete matrix material on the bond damage mechanism, the WERI values in the test specimens using different concrete materials are also calculated and presented in Fig. 14.

Fig. 14
figure 14

The WERI values and bond-slip curves of different concrete materials

Full size image

In the comparison of the bond-slip behavior of the test series of SCC-50%-GFRP series (see Fig. 14a, b) and SCC-65%-GFRP series (Fig. 14c, d), it is noted that increasing the fly ash replacement level results in faster deterioration after the ultimate bond strength. The WERI values are increased significantly before the peak bond stress in the SCC-65%-GFRP series (Fig. 14c, d). This indicates that the bond behavior provided by the chemical adhesion between GFRP bars and SCC is in good condition before the bond failure. After the bond failure, the WERI values are reduced rapidly. Since the lower strength of concrete material with high-volume fly ash results in concrete to be more vulnerable to damage around the interface during the friction process. Additionally, Fig. 14b, c also illustrates that bond failure and bond strength can be captured effectively and accurately using the WERI values.

The development of WERI values in the test specimens conducted in normal concrete (coded as the NC-GFRP series) is a bit different from those in SCC test specimens. As shown in Fig. 14e, f, the curves of WERI and slip are dropped sharply as soon as the bond stress reaches the ultimate strength. This monitoring result is attributed to the poorer compactness in the normal concrete compared to that in SCC that results in smaller chemical adhesion between FRP bars and concrete. This phenomenon can also be found in the bond-slip behaviors of those test specimens. It can be summarized that the degradation degree of chemical adhesion between FRP bars and concrete has a strong effect on the degree of the stress wave attenuation between two transducers. Therefore, the damage level of chemical adhesion can be determined by using the WERI values. Additionally, it can be inferred from the study of those monitoring results that the bond behavior of FRP reinforcing bars and concrete is majorly dependent on the chemical adhesion.

As discussed above, the damage mechanism and failure modes of the bond between different concrete materials and GFRP bars can be effectively identified by using the WERI values. Simultaneously, the peak of bond stress, which indicates the bond failure, can be detected based on WERI values.

5 Conclusions

In this paper, a series of pull-out tests were carried out to investigate the bond behavior between reinforcing bars and SCC materials by using piezoceramic transducers. A piezoceramic patch transducer, which worked as an actuator, was bonded on the outer surface of reinforcing bars, and a smart aggregate (SA) transducer, which worked as a sensor, was embedded in concrete. The wavelet energy ratio index (WERI) was proposed to analyze the damage mechanism of the interface between concrete and reinforcing bars. The following conclusion can be obtained:

  1. 1.

    It is evident in the test results that geometric shape of surface and lower elastic modulus of FRP reinforcing bars result in smaller bond strength and different failure mode compared to those in steel reinforced test specimens. The high-volume fly ash, instead of cement, has little effect on the bond behavior of GFRP reinforcement and SCC before the peak bond stress. However, the bond strength of GFRP bars and SCC is higher than that of GFRP bars and normal concrete, which indicates that the combined using FRP and SCC can produce an effective engineering application.

  2. 2.

    The loss of mechanical interlock resistance, the surrounding concrete around the interface being damaged, results in a sudden and brittle bond failure between steel reinforcement and concrete. The signal amplitude in time-domain and the signal energy in the frequency spectrogram of steel bar and SCC are drastically reduced when the peak bond stress is reached, which indicates that the bond failure of steel bar and concrete can be detected accurately.

  3. 3.

    The bond strength of FRP bar and concrete mainly depends on the friction and chemical adhesion. The time-domain signal amplitude representing the stress wave propagation between the FRP reinforcement and concrete is barely varied during the pull-out process. However, the signal energy in frequency spectrograms at 120–200 kHz is gradually attenuated, which can be used to assess the interface damage status between FRP bars and concrete.

  4. 4.

    Using the WERI values, the damage mechanism, including friction, chemical adhesion and mechanical interlock damage, of different reinforcing bars and concrete materials can be effectively identified. The failure modes of interface bond behavior between different reinforcing bars and concrete materials can be predicted effectively, and the peak of bond stress can be captured accurately by using the WERI values.

  5. 5.

    It is feasible to identify the bond damage on the interface between reinforcing bars and concrete by using piezoceramic transducers enabled active sensing and the wavelet energy analysis. This feasibility established in the present paper warrants a more detailed investigation into structural health monitoring technology for bond behavior related to FRP reinforced SCC structures, particularly in large-scale structures.