Monitoring of Epoxy-Grouted Bonding Strength Development between an Anchored Steel Bar and Concrete Using PZT-Enabled Active Sensing

Anchored steel bars have been widely used in retrofitting of existing concrete structures. The bonding strength between the anchored steel bar and the concrete is critical to the integrity of the strengthened concrete structure. This paper presents a method to monitor epoxy-grouted bonding strength development by using a piezoceramic-enabled active sensing technique. One concrete beam with an anchored steel bar was involved in the monitoring test, and two concrete beams with six anchored steel bars were used in the pull-out test. To enable the active sensing, a Lead Zirconate Titanate (PZT) patch was bonded to the surface of the exposed end, and piezoceramic smart aggregates were embedded in each concrete specimen. During the monitoring experiment, signals from PZT sensors and smart aggregates were acquired at intervals of 0, 20, 40, 60, 80, and 100 min. In addition, a pull-out test was performed on each of the remaining six anchored steel bars in the two concrete beams, while the signal was recorded in the test. Furthermore, a wavelet packet analysis was applied to analyze the received signal energies to investigate the bonding strength development between the concrete and the anchored steel bar during the epoxy solidification process. The test results demonstrate the effectiveness of the proposed method in monitoring the bonding strength development between the anchored steel bar and the concrete, using the PZT-enabled active sensing.


Introduction
The retrofit, reinforcement, and strengthening of existing structures have been widely practiced and researched [1,2], and various guidelines and standards have been developed [3][4][5][6]. Anchored steel bar technology has been widely used in retrofits and new construction since the 1970s [7]. The technology, as a post-installed connection method, plants a steel bar into a concrete base with structural adhesives [8]. Anchored steel bars in concrete are required when an additional structural member is to be linked to the existing concrete structure [9]. Integrated with the original structural members, the anchored steel bars make the entire structural system more reliable, increasing the bearing capacity of the structure [10].
Due to its wide application, the anchored steel bar technology has received much attention in the literature. In the 1990s, Cook [11] and Cook et al. [12,13] carried out a series of experimental

Piezoceramic-Based Active Sensing Approach
Due to the piezoelectricity, PZT can be either used as an actuator or a sensor [36,39,46]. Therefore, PZT transducers and PZT smart aggregates can detect the propagating stress wave when they are deployed on the wave path [48,55,60]. Figure 1 shows a three-dimensional (3D) sketch of the specimen. Two PZT transducers, namely PZT1 and PZT2, were mounted to the concrete beam surface as sensors. Two PZT smart aggregates (SAs), namely SA1 and SA2, were embedded in the concrete beam by drilling holes in the existing concrete. One PZT transducer, namely PZT3, was mounted on the steel bar as an actuator. In this research, the specimen with PZT3 attached to the exposed end of the steel bar was used to verify the feasibility of monitoring epoxy-grouted bonding strength by using the PZT-enabled active sensing technique.

Piezoceramic-Based Active Sensing Approach
Due to the piezoelectricity, PZT can be either used as an actuator or a sensor [36,39,46]. Therefore, PZT transducers and PZT smart aggregates can detect the propagating stress wave when they are deployed on the wave path [48,55,60]. Figure 1 shows a three-dimensional (3D) sketch of the specimen. Two PZT transducers, namely PZT1 and PZT2, were mounted to the concrete beam surface as sensors. Two PZT smart aggregates (SAs), namely SA1 and SA2, were embedded in the concrete beam by drilling holes in the existing concrete. One PZT transducer, namely PZT3, was mounted on the steel bar as an actuator. In this research, the specimen with PZT3 attached to the exposed end of the steel bar was used to verify the feasibility of monitoring epoxy-grouted bonding strength by using the PZT-enabled active sensing technique.  Figure 2 shows the schematic of the active sensing approach in the monitoring of the epoxy-grouted bonding strength development between the anchored steel bar and the concrete during the epoxy solidification process. The stress wave generated by PZT3 propagates from the anchored steel bar to the concrete beam through the epoxy inside the hole. PZT1 and PZT2 are on the concrete beam's surface, and SA1 and SA2 are inside of the concrete beam. Therefore, these transducers can detect the stress wave, since the concrete material is a good conduit for stress wave propagation [68,71,75]. During the curing process of the epoxy, the propagation of stress waves highly depends on the status of the transmission medium [80]. When the epoxy is in the liquid form, the compressive stress wave highly attenuates and the corresponding detected signals are very weak [59,81,82]. When the epoxy cures, the increased stiffness of the epoxy will directly influence the characteristics of the received signal from PZT3. Simultaneously, the epoxy-grouted bonding strength between the anchored steel bar and the concrete beam increases, and the stress wave energy propagated from the steel bar to the concrete will be greatly strengthened. As a result, the signal strength received by PZT1, PZT2, SA1, and SA2 will also increase, respectively. Based on these signals, the strength development of epoxy between the anchored steel bar and the concrete can be monitored.
(a) Figure 1. Sketch of the specimen with deployed SAs (lead zirconate titanate smart aggregates) and PZT (lead zirconate titanate) transducers. Figure 2 shows the schematic of the active sensing approach in the monitoring of the epoxy-grouted bonding strength development between the anchored steel bar and the concrete during the epoxy solidification process. The stress wave generated by PZT3 propagates from the anchored steel bar to the concrete beam through the epoxy inside the hole. PZT1 and PZT2 are on the concrete beam's surface, and SA1 and SA2 are inside of the concrete beam. Therefore, these transducers can detect the stress wave, since the concrete material is a good conduit for stress wave propagation [68,71,75]. During the curing process of the epoxy, the propagation of stress waves highly depends on the status of the transmission medium [80]. When the epoxy is in the liquid form, the compressive stress wave highly attenuates and the corresponding detected signals are very weak [59,81,82]. When the epoxy cures, the increased stiffness of the epoxy will directly influence the characteristics of the received signal from PZT3. Simultaneously, the epoxy-grouted bonding strength between the anchored steel bar and the concrete beam increases, and the stress wave energy propagated from the steel bar to the concrete will be greatly strengthened. As a result, the signal strength received by PZT1, PZT2, SA1, and SA2 will also increase, respectively. Based on these signals, the strength development of epoxy between the anchored steel bar and the concrete can be monitored.

Piezoceramic-Based Active Sensing Approach
Due to the piezoelectricity, PZT can be either used as an actuator or a sensor [36,39,46]. Therefore, PZT transducers and PZT smart aggregates can detect the propagating stress wave when they are deployed on the wave path [48,55,60]. Figure 1 shows a three-dimensional (3D) sketch of the specimen. Two PZT transducers, namely PZT1 and PZT2, were mounted to the concrete beam surface as sensors. Two PZT smart aggregates (SAs), namely SA1 and SA2, were embedded in the concrete beam by drilling holes in the existing concrete. One PZT transducer, namely PZT3, was mounted on the steel bar as an actuator. In this research, the specimen with PZT3 attached to the exposed end of the steel bar was used to verify the feasibility of monitoring epoxy-grouted bonding strength by using the PZT-enabled active sensing technique.  Figure 2 shows the schematic of the active sensing approach in the monitoring of the epoxy-grouted bonding strength development between the anchored steel bar and the concrete during the epoxy solidification process. The stress wave generated by PZT3 propagates from the anchored steel bar to the concrete beam through the epoxy inside the hole. PZT1 and PZT2 are on the concrete beam's surface, and SA1 and SA2 are inside of the concrete beam. Therefore, these transducers can detect the stress wave, since the concrete material is a good conduit for stress wave propagation [68,71,75]. During the curing process of the epoxy, the propagation of stress waves highly depends on the status of the transmission medium [80]. When the epoxy is in the liquid form, the compressive stress wave highly attenuates and the corresponding detected signals are very weak [59,81,82]. When the epoxy cures, the increased stiffness of the epoxy will directly influence the characteristics of the received signal from PZT3. Simultaneously, the epoxy-grouted bonding strength between the anchored steel bar and the concrete beam increases, and the stress wave energy propagated from the steel bar to the concrete will be greatly strengthened. As a result, the signal strength received by PZT1, PZT2, SA1, and SA2 will also increase, respectively. Based on these signals, the strength development of epoxy between the anchored steel bar and the concrete can be monitored. Remark: In some applications, the exposed end of the steel bar may not be available to be bonded with the PZT patch. In such cases, we will design and try re-usable PZT transducers with integrated magnets. These re-usable PZT transducers can be attached to the end surface or the side surface of the anchored steel rod by using magnetic force. Furthermore, the re-usable PZT transducers will help to broaden the potential applications of the proposed approach.

Wavelet-Packet-Based Energy Index
The wavelet-packet-based energy index (WPEI) analysis has been demonstrated to be an effective technique to quantitatively estimate the differences between the stress wave responses due to structural changes, such as damage [22,46,60,67,68,72,73], and structural material phase changes, such as concrete curing [59,80] and soil freeze-thaw [81]. In wavelet analysis, a signal can be divided into several wavelet packets and the signal energy of each packet can be computed [56,67]. Consequently, the energy of the signal can be computed by the summation of energies of all the packets [67,79]. The received signal is decomposed by n-level wavelet packet decomposition into signal sets , , … , in each decomposed signal from the original signal , where j is the frequency band , , … , . In this study, .
can be further expressed as: where is the sampling data collected from the data acquisition system. Thus, the energy of the decomposed signal can be computed as: ,

Remark:
In some applications, the exposed end of the steel bar may not be available to be bonded with the PZT patch. In such cases, we will design and try re-usable PZT transducers with integrated magnets. These re-usable PZT transducers can be attached to the end surface or the side surface of the anchored steel rod by using magnetic force. Furthermore, the re-usable PZT transducers will help to broaden the potential applications of the proposed approach.

Wavelet-Packet-Based Energy Index
The wavelet-packet-based energy index (WPEI) analysis has been demonstrated to be an effective technique to quantitatively estimate the differences between the stress wave responses due to structural changes, such as damage [22,46,60,67,68,72,73], and structural material phase changes, such as concrete curing [59,80] and soil freeze-thaw [81]. In wavelet analysis, a signal can be divided into several wavelet packets and the signal energy of each packet can be computed [56,67]. Consequently, the energy of the signal can be computed by the summation of energies of all the packets [67,79]. The received signal is decomposed by n-level wavelet packet decomposition into 2 n signal sets {X 1 , X 2 , . . . , X 2 n } in each decomposed signal from the original signal X j , where j is the frequency band j = 1, 2, . . . , 2 n . In this study, n = 5. X j can be further expressed as: where m is the sampling data collected from the data acquisition system. Thus, the energy of the decomposed signal E j can be computed as: The total energy of the signal can be computed by the summation of all the decomposed signals. The computed energy of the signal E can be expressed as: The received signal can be characterized by using the WPEI analysis. The WPEI has been used to monitor or evaluate the structural health condition in various concrete structure applications [82,83]. In this research, the WPEI was applied to monitor the received signal energy change in the active sensing approach when the bonding strength between the concrete and the steel bar was changing.

Specimen Fabrication
The dimensions of the specimen for the monitoring test and the location of PZTs and SAs are presented in Figure 3. PZT1 and PZT2 were attached to the concrete beam (CB1) surface by epoxy, while SA1 and SA2 were embedded into CB1 by drilling holes in the concrete. Additionally, PZT3 was bonded to the surface of the exposed end of the steel bar (SB1) by epoxy. The locations of the three PZT transducers and the two smart aggregates are shown in Figure 3. Figure 4 shows the experimental specimen for the monitoring test. In order to find out the connection between the energy index and bonding strength, it is necessary to perform the monitoring test and the pull-out test simultaneously. Figure 5 shows the experimental specimen for the pull-out test. There are two concrete beams (CB2 and CB3) and three steel bars (SB2-SB7) were anchored into each of them, while SB1 was anchored into CB1 with the same epoxy. The total energy of the signal can be computed by the summation of all the decomposed signals. The computed energy of the signal can be expressed as: The received signal can be characterized by using the WPEI analysis. The WPEI has been used to monitor or evaluate the structural health condition in various concrete structure applications [82,83]. In this research, the WPEI was applied to monitor the received signal energy change in the active sensing approach when the bonding strength between the concrete and the steel bar was changing.

Specimen Fabrication
The dimensions of the specimen for the monitoring test and the location of PZTs and SAs are presented in Figure 3. PZT1 and PZT2 were attached to the concrete beam (CB1) surface by epoxy, while SA1 and SA2 were embedded into CB1 by drilling holes in the concrete. Additionally, PZT3 was bonded to the surface of the exposed end of the steel bar (SB1) by epoxy. The locations of the three PZT transducers and the two smart aggregates are shown in Figure 3. Figure 4 shows the experimental specimen for the monitoring test. In order to find out the connection between the energy index and bonding strength, it is necessary to perform the monitoring test and the pull-out test simultaneously. Figure 5 shows the experimental specimen for the pull-out test. There are two concrete beams (CB2 and CB3) and three steel bars (SB2-SB7) were anchored into each of them, while SB1 was anchored into CB1 with the same epoxy.
The dimensions of concrete beams CB1-CB3 is 550 mm × 150 mm × 150 mm and each specimen was made with concrete of class C40. The size of steel bars SB1-SB7 is Φ12 mm × 50 mm and the type is HRB335. The material properties of the test specimens are shown in Table 1. The depth of all the anchored steel bars within the concrete beam is 120 mm.

Experimental Setup
The experimental setup of the monitoring test includes the concrete beam (CB1) with an anchored steel bar (SB1), PZTs and SAs, a power amplifier (Trek 2100HF) for piezoceramic load, and a data acquisition system (NI USB6363) hosted by a laptop computer, as shown in Figure 6. The sampling frequency of the data acquisition system was 2 MS/s. During the test, a sweep sine wave was generated by the data acquisition system, and then amplified by the power amplifier with a gain of 50. The amplified signal was used to drive the actuated signals of PZT3. PZT1, PZT2, SA1, and SA2 were used as sensors to receive signals from the concrete. Signals were recorded at the 0, 20, 40,

Experimental Setup
The experimental setup of the monitoring test includes the concrete beam (CB1) with an anchored steel bar (SB1), PZTs and SAs, a power amplifier (Trek 2100HF) for piezoceramic load, and a data acquisition system (NI USB6363) hosted by a laptop computer, as shown in Figure 6. The sampling frequency of the data acquisition system was 2 MS/s. During the test, a sweep sine wave was generated by the data acquisition system, and then amplified by the power amplifier with a gain of 50. The amplified signal was used to drive the actuated signals of PZT3. PZT1, PZT2, SA1, and SA2 were used as sensors to receive signals from the concrete. Signals were recorded at the 0, 20, 40, The dimensions of concrete beams CB1-CB3 is 550 mm × 150 mm × 150 mm and each specimen was made with concrete of class C40. The size of steel bars SB1-SB7 is Φ12 mm × 50 mm and the type is HRB335. The material properties of the test specimens are shown in Table 1. The depth of all the anchored steel bars within the concrete beam is 120 mm.

Experimental Setup
The experimental setup of the monitoring test includes the concrete beam (CB1) with an anchored steel bar (SB1), PZTs and SAs, a power amplifier (Trek 2100HF) for piezoceramic load, and a data acquisition system (NI USB6363) hosted by a laptop computer, as shown in Figure 6. The sampling frequency of the data acquisition system was 2 MS/s. During the test, a sweep sine wave was generated by the data acquisition system, and then amplified by the power amplifier with a gain of 50. The amplified signal was used to drive the actuated signals of PZT3. PZT1, PZT2, SA1, and SA2 were used as sensors to receive signals from the concrete. Signals were recorded at the 0, 20, 40, 60, 80, and 100 min mark during this experiment. The start frequency, stop frequency, amplitude, and time interval of the excitation signal were 1000 Hz, 150 kHz, 10 V, and 0.5 s, respectively. The pull-out test involved two concrete beams (CB2 and CB3) with anchored steel bars (SB2-SB7), as shown in Figure 7. A hydraulic pump was used to apply the load during the pull-out test. SB2, SB3, SB4, SB5, SB6, and SB7 were pulled out at the 0, 20, 40, 60, 80, and 100 min mark, respectively. The ambient temperature was approximately 15 ℃ during the experiments. Since all the tests were conducted within 2 h in the laboratory, the influence of temperature disturbance on the results was negligible.

Experimental Results and Discussions
The results of the monitoring test are shown in Figures 8-11, which show the time domain signal responses that PZT1, PZT2, SA1, and SA2 received at the 0, 20, 40, 60, 80, and 100 min mark, respectively. Each curve represents the signal response of the swept sine wave received by the sensor in 0.5 s. Since the bonding strength between the steel bar and the concrete changes with time during the initial curing period, the signals received by two PZTs and two SAs were different at different time intervals.
It can be seen from Figure 8 that the time domain signal received by PZT1 at the 0 th min and the 20 th min was close. Afterwards, the time domain signal amplitude increased significantly between 20 and 40 min, similarly, as represented from 40 to 60 min. From 60 to 80 min, the time domain signal amplitude still increased, but at a much a slower rate. In addition, the change in the time domain signal's amplitude from 80 to 100 min was small. Figures 9-11 show that the variation trend of the time domain signals of PZT2, SA1, and SA2 are almost the same as that of PZT1. The results illustrate that the piezoceramic-based active sensing approach can effectively monitor the bonding strength development in real time during the epoxy solidification process. The pull-out test involved two concrete beams (CB2 and CB3) with anchored steel bars (SB2-SB7), as shown in Figure 7. A hydraulic pump was used to apply the load during the pull-out test. SB2, SB3, SB4, SB5, SB6, and SB7 were pulled out at the 0, 20, 40, 60, 80, and 100 min mark, respectively. The ambient temperature was approximately 15°C during the experiments. Since all the tests were conducted within 2 h in the laboratory, the influence of temperature disturbance on the results was negligible. The pull-out test involved two concrete beams (CB2 and CB3) with anchored steel bars (SB2-SB7), as shown in Figure 7. A hydraulic pump was used to apply the load during the pull-out test. SB2, SB3, SB4, SB5, SB6, and SB7 were pulled out at the 0, 20, 40, 60, 80, and 100 min mark, respectively. The ambient temperature was approximately 15 ℃ during the experiments. Since all the tests were conducted within 2 h in the laboratory, the influence of temperature disturbance on the results was negligible.

Experimental Results and Discussions
The results of the monitoring test are shown in Figures 8-11, which show the time domain signal responses that PZT1, PZT2, SA1, and SA2 received at the 0, 20, 40, 60, 80, and 100 min mark, respectively. Each curve represents the signal response of the swept sine wave received by the sensor in 0.5 s. Since the bonding strength between the steel bar and the concrete changes with time during the initial curing period, the signals received by two PZTs and two SAs were different at different time intervals.
It can be seen from Figure 8 that the time domain signal received by PZT1 at the 0 th min and the 20 th min was close. Afterwards, the time domain signal amplitude increased significantly between 20 and 40 min, similarly, as represented from 40 to 60 min. From 60 to 80 min, the time domain signal amplitude still increased, but at a much a slower rate. In addition, the change in the time domain

Experimental Results and Discussions
The results of the monitoring test are shown in Figures 8-11, which show the time domain signal responses that PZT1, PZT2, SA1, and SA2 received at the 0, 20, 40, 60, 80, and 100 min mark, respectively. Each curve represents the signal response of the swept sine wave received by the sensor in 0.5 s. Since the bonding strength between the steel bar and the concrete changes with time during the initial curing period, the signals received by two PZTs and two SAs were different at different time intervals.     It can be seen from Figure 8 that the time domain signal received by PZT1 at the 0th min and the 20th min was close. Afterwards, the time domain signal amplitude increased significantly between 20 and 40 min, similarly, as represented from 40 to 60 min. From 60 to 80 min, the time domain signal amplitude still increased, but at a much a slower rate. In addition, the change in the time domain signal's amplitude from 80 to 100 min was small. Figures 9-11 show that the variation trend of the time domain signals of PZT2, SA1, and SA2 are almost the same as that of PZT1. The results illustrate that the piezoceramic-based active sensing approach can effectively monitor the bonding strength development in real time during the epoxy solidification process.
Furthermore, the WPEI was applied to quantify the energy of the received signal, as shown in Figure 12. As the energy indices show in this figure, the wave energy was low during the initial 20 min since the epoxy was liquid, and the stress wave can hardly travel through it. A rapid increase in wave energy can be directly observed from 20 to 40 min and from 40 to 60 min for all the PZT sensors and smart aggregates. It suggests that the epoxy begins to solidify and to bond to the steel bar and concrete beam. Therefore, the stress wave can easily transmit from the steel to the concrete beam through the gradually hardened epoxy. From 60 to 80 min, the wave energy increased but at a much slower rate, which suggests that the curing process of the epoxy was close to the end. In addition, the signal energy reached a steady-state value without any obvious changes from 80 to 100 min. It means that the curing process of the epoxy has been completed. The amplitudes of energy indices of PZT1, PZT2, SA1, and SA2 are different because of the different propagation paths of stress waves in the concrete. For example: concrete is not a homogeneous material, and PZT1, PZT2, SA1, and SA2 were deployed in different directions on CB1. Nevertheless, the energy index trends of PZT1, PZT2, SA1, and SA2 are almost the same as the initial 100-min period. Furthermore, the WPEI was applied to quantify the energy of the received signal, as shown in Figure 12. As the energy indices show in this figure, the wave energy was low during the initial 20 min since the epoxy was liquid, and the stress wave can hardly travel through it. A rapid increase in wave energy can be directly observed from 20 to 40 min and from 40 to 60 min for all the PZT sensors and smart aggregates. It suggests that the epoxy begins to solidify and to bond to the steel bar and concrete beam. Therefore, the stress wave can easily transmit from the steel to the concrete beam through the gradually hardened epoxy. From 60 to 80 min, the wave energy increased but at a much slower rate, which suggests that the curing process of the epoxy was close to the end. In addition, the signal energy reached a steady-state value without any obvious changes from 80 to 100 min. It means that the curing process of the epoxy has been completed. The amplitudes of energy indices of PZT1, PZT2, SA1, and SA2 are different because of the different propagation paths of stress waves in the concrete. For example: concrete is not a homogeneous material, and PZT1, PZT2, SA1, and SA2 were deployed in different directions on CB1. Nevertheless, the energy index trends of PZT1, PZT2, SA1, and SA2 are almost the same as the initial 100-min period.  Figure 13 shows the specimen after the pull-out test. All steel bars were pulled out from the concrete beams after the test. As shown in Table 2, the maximum tensile force was small and negligible during the initial 20 min, which means that the epoxy has not begun to cure and the corresponding epoxy-grouted bonding strength was very small. The epoxy began to solidify and the bonding strength greatly increased by 31 KN rapidly from 20 to 60 min. From 60 to 80 min, the maximum tensile strength was still increasing, but only increased by 4.93 KN. From 80 to 100 min, the maximum tensile strength also reached a maximum steady-state value to indicate that the solidification process of the epoxy was completed. Comparing the WPEI with the maximum tension in Figure 14, where the four dotted lines represent the energy indices of PZT1, PZT2, SA1, and SA2, and the red solid line represents the maximum tension value of the pull-out test, clearly reveals that the maximum pull-out tensile and WPEI possess a similar trend. The pull-out results also verified the effectiveness of the monitoring results.  Figure 13 shows the specimen after the pull-out test. All steel bars were pulled out from the concrete beams after the test. As shown in Table 2, the maximum tensile force was small and negligible during the initial 20 min, which means that the epoxy has not begun to cure and the corresponding epoxy-grouted bonding strength was very small. The epoxy began to solidify and the bonding strength greatly increased by 31 KN rapidly from 20 to 60 min. From 60 to 80 min, the maximum tensile strength was still increasing, but only increased by 4.93 KN. From 80 to 100 min, the maximum tensile strength also reached a maximum steady-state value to indicate that the solidification process of the epoxy was completed. Comparing the WPEI with the maximum tension in Figure 14, where the four dotted lines represent the energy indices of PZT1, PZT2, SA1, and SA2, and the red solid line represents the maximum tension value of the pull-out test, clearly reveals that the maximum pull-out tensile and WPEI possess a similar trend. The pull-out results also verified the effectiveness of the monitoring results.   Remark: In field implementations, it is important to know when the epoxy reaches its full bonding strength so that the next step of the retrofit or reinforcement, that requires the application of a programmed load through the anchored steel bar, can be carried out. The WPEI results (Figure 12) clearly show that the acquired signal energy reached the maximum value at the 80 th min and would not increase after the 100 th min. This phenomenon correlates well with the results of the pull-out tests. Basically, from the sensor responses, it is feasible to estimate when the epoxy reaches its full strength, thereby determining the moment of the realization of the planned program of loads for the upgraded structure.

Conclusions and Future Work
In this paper, a straightforward piezoceramic-based active sensing approach was utilized to monitor epoxy-grouted bonding strength development between an anchored steel bar and the concrete in real time. To verify the proposed methods, experiments were carried out on specimens with anchored steel bars. A PZT patch was bonded to the exposed end of each of the anchored steel bars, and piezoceramic smart aggregates were embedded in the concrete specimens to realize the active sensing approach. In the monitoring test, the time domain signal amplitude and WPEI demonstrate that, during the first 20 min, since the epoxy was liquid, the stress wave can hardly travel through it. During the next 40 min, the epoxy began to solidify and to bond to the steel bar    Figure 13. The specimen after the pull-out test  Remark: In field implementations, it is important to know when the epoxy reaches its full bonding strength so that the next step of the retrofit or reinforcement, that requires the application of a programmed load through the anchored steel bar, can be carried out. The WPEI results (Figure 12) clearly show that the acquired signal energy reached the maximum value at the 80 th min and would not increase after the 100 th min. This phenomenon correlates well with the results of the pull-out tests. Basically, from the sensor responses, it is feasible to estimate when the epoxy reaches its full strength, thereby determining the moment of the realization of the planned program of loads for the upgraded structure.

Conclusions and Future Work
In this paper, a straightforward piezoceramic-based active sensing approach was utilized to monitor epoxy-grouted bonding strength development between an anchored steel bar and the concrete in real time. To verify the proposed methods, experiments were carried out on specimens with anchored steel bars. A PZT patch was bonded to the exposed end of each of the anchored steel bars, and piezoceramic smart aggregates were embedded in the concrete specimens to realize the active sensing approach. In the monitoring test, the time domain signal amplitude and WPEI demonstrate that, during the first 20 min, since the epoxy was liquid, the stress wave can hardly travel through it. During the next 40 min, the epoxy began to solidify and to bond to the steel bar  Remark: In field implementations, it is important to know when the epoxy reaches its full bonding strength so that the next step of the retrofit or reinforcement, that requires the application of a programmed load through the anchored steel bar, can be carried out. The WPEI results ( Figure 12) clearly show that the acquired signal energy reached the maximum value at the 80th min and would not increase after the 100th min. This phenomenon correlates well with the results of the pull-out tests. Basically, from the sensor responses, it is feasible to estimate when the epoxy reaches its full strength, thereby determining the moment of the realization of the planned program of loads for the upgraded structure.

Conclusions and Future Work
In this paper, a straightforward piezoceramic-based active sensing approach was utilized to monitor epoxy-grouted bonding strength development between an anchored steel bar and the concrete in real time. To verify the proposed methods, experiments were carried out on specimens with anchored steel bars. A PZT patch was bonded to the exposed end of each of the anchored steel bars, and piezoceramic smart aggregates were embedded in the concrete specimens to realize the active sensing approach. In the monitoring test, the time domain signal amplitude and WPEI demonstrate that, during the first 20 min, since the epoxy was liquid, the stress wave can hardly travel through it. During the next 40 min, the epoxy began to solidify and to bond to the steel bar and the concrete beam. Therefore, the stress wave can easily transmit from the steel to the concrete beam through the gradually cured epoxy. During the 60-80 min period, the time domain signal amplitude and WPEI increased but at a much slower rate, which suggests that the curing process of the epoxy was almost completed. In addition, both the received time-domain signals and signal energy reached a steady-state value from 80 to 100 min. It directly indicated that the curing process of the epoxy was completely done. Comparing the WPEI with the maximum tensile force from the pull-out tests clearly shows that the maximum pull-out force and the energy indices are closely correlated. The results of monitoring test and pull-out tests demonstrate that the piezoceramic-based active sensing approach is capable of monitoring the epoxy-grouted bonding strength development in real time.
In further work, the different installation methods of transducers and the sensing range of the PZTs and SAs will be further studied. In addition, aspects of a real-life situation, including external loading and, especially, ambient temperature, will also be investigated. Furthermore, numerical or theoretical models will be developed to compare with the current experimental results.