Abstract
An active sensing approach using piezoceramic induced stress wave is proposed to provide monitoring and early warning for the development of interface debonding damage of precast segmental concrete beams (PSCBs). Three concrete specimens with toothed interfaces were fabricated and bonded with high-strength epoxy resin adhesive to form PSCBs. Smart aggregates (SAs) embedded in concrete specimens are used as actuators and sensors. The PSCBs are subjected to periodic loading with hydraulic jack to test the different degrees of debonding damage. The experimental results of time-domain and frequency-domain analysis clearly show that the amplitude of the signal received by the piezoceramic sensor is reduced when debonding crack occurs. The energy analysis and damage index based on wavelet packet can be used to determine the existence and severity of interface debonding damage in PSCBs. The experimental research validates the feasibility of monitoring the interface debonding damage in PSCBs using SA transducers based on active sensing technique.
1. Introduction
Precast segmental concrete beams (PSCBs) are widely applied to civil engineering [1–4]. Compared with the monolithic cast-in-place beams, the PSCBs have advantages of easy transportation, speedy construction, low environmental impact, easy quality assurance, and strong self-resetting ability [4]. Therefore, the development of PSCBs is an irresistible trend. However, the comprehensive performance of PSCBs is associated with the quality of joint interface. The interface of PSCBs is formed by bonding brittle matrix composites (concrete) [5] with high-strength epoxy resin adhesive. Because the strength of cement mortar matrix at the interface is not as good as that of the internal concrete, the interface of the PSCB is easy to become a channel invading to the PSCB by the external harmful substances. Therefore, the properties of fracture toughness [6, 7] and internal microcracks [8, 9] exist at the interface, which becomes the weakest link of PSCBs to resist adverse environmental effects. Shear effect occurs frequently along the interface between the two bonded PSCBs under various factors, such as dynamic loads caused by moving vehicles [10], corrosions [8], or impact loading during extreme events [11], which may lead to debonding cracks at the interface and can seriously weaken the integrity and bearing capacity of the connected PSCBs. Due to the pulling effect of the epoxy resin adhesive on the concrete, it is easy to cause vertical splitting cracks at joint, which results in interface damages. Therefore, the detection of interface damage of PSCBs is a very complex problem, and its safety issues have been widely concerned.
The conventional detection techniques can be classified into destructive and nondestructive evaluation techniques [12]. The destructive evaluation technique usually refers to the core test method and the loading test method, which leads to the artificial and unexpected permanent damage of the structure [13]. The nondestructive evaluation techniques include the ultrasonic method, percussion method, infrared thermal imaging method, electromagnetic pulse method, laser shear imaging method, impact-echo method, and ground penetrating radar method [12–14]. Ultrasonic detection results are relatively discrete and time-consuming. The judgement of the first wave requires that the inspectors have rich engineering experience, and the location of the measuring points will have an impact on the detection results. The percussion method is mainly used to judge whether there is damage according to experience. There is no objective data as the object of analysis, and it is generally only used as an auxiliary detection method. The infrared thermal imaging method is often inaccurate for hidden interfaces because of its fast infrared attenuation in concrete [15]. The electromagnetic pulse method cannot be used for large structures because of the fast energy attenuation and high frequency [16]. The main disadvantage of laser shear imaging method is that it cannot locate the damage accurately and only recognizes surface damage [16]. The conventional nondestructive technologies also have some disadvantages, such as being time-consuming, requiring huge equipment, having high cost, being labor-intensive, exposing inspectors to dangerous environments, and not being suitable for inaccessible but critical locations [13, 17].
The emergence of smart materials, such as piezoceramic materials, especially, lead zirconate titanate (PZT), makes structural health monitoring (SHM) possible through fully integrated transducers, offering real-time monitoring of structural health status and adding a new dimension to traditional structural detection technology [14, 18]. Piezoceramic materials have many advantages, such as availability in different dimensions and shapes, fast response, wide frequency bandwidth, low cost, and capacities as both actuator and sensor for stress wave generation and detection [19]. These properties prepare piezoceramic materials for their wide applications in the field of structural health monitoring and damage detection [20].
Active sensing technology can identify the damage and defects of the structure by analyzing the difference between the received signal of the sensor before and after the structural damage [21]. This technology uses at least a pair of piezoceramic actuator and sensor bonded to the surface of a structure or embedded in a structure when possible. Song et al. conducted a study on surface wave propagation in concrete structures using a PZT actuator/sensor system [22]. Lim et al. developed a semianalytical model using surface-bonded piezoceramic transducers to evaluate the strength of mortars of different contents during the entire curing process [13]. Yan et al. used PZT as transducers to experiment on the interface debonding slip and separation of steel-concrete composite beams, aiming to propose an interface damage identification model based on PZT wave technology and innovative damage detection methods [16].
At present, due to the limitations of the structure working environment and the difficulty of traditional monitoring instruments in practical applications, the damage identification or monitoring method of the PSCBs interface has not been established in practical engineering. Therefore, it is necessary to establish a real-time monitoring and damage assessment method.
Although many methods have been successfully proposed for condition monitoring of different structures, there has been no report on the interface condition monitoring of PSCBs. In this paper, three-array piezoceramic smart aggregates (SAs) are embedded in concrete to enable the active sensing approach, which is used to identify the interface damage of segmented bonded concrete specimens during the loading period. The occurrence and severity of cracks attenuate the propagation of waves which can be reflected by the received signal in time/frequency domain, meaning that wave energy dissipates when it passes through the interface. Wavelet packet analysis is used to analyze the change of received signal energy, and a periodic damage index is proposed. Experimental results demonstrate that piezoceramic SAs can timely monitor the development process of cracks in segmented concrete specimens, which lays a foundation for the identification of related interface states in the future.
2. Piezoceramic Smart Aggregates
2.1. Piezoelectric Effect
The most typical characteristic of piezoelectric materials is the piezoelectric effect [23]. When the external mechanical deformation of the piezoelectric element occurs, the internal positive and negative charges will move between the polarization of the electric field. The two charges with different symbols are distributed on the surfaces of the two electrodes of the element. The external force directly determines the charge distribution density, which is called the positive piezoelectric effect [24]. The ability of piezoelectric materials to convert mechanical energy into electrical energy can be determined by this effect. The inverse piezoelectric effect refers to the relative movement of the positive and negative charge centers of the piezoelectric element under the action of voltage, which makes the element deform [25]. The inverse piezoelectric effect is mainly used to explain the ability of piezoelectric materials to convert electrical energy into mechanical energy.
Based on the piezoelectric effect, transducers with dual functions of stress wave transmission and detection can be fabricated [18]. One example is the piezoceramic smart aggregate that can be embedded in the concrete structure to enable the stress wave-based active sensing approach [26]. The structural health status can be monitored based on various algorithms by analyzing the data collected by the sensors [27].
2.2. Smart Aggregates
The selected PZT is specially processed and packaged in drum-shaped cement mortar or fine stone concrete block with a volume of approximately 8–10 cm3, which could be placed into the structure after curing. For the connection, a Bayonet Neill–Concelman (BNC) connector is welded at the end of the wire, as shown in Figure 1 [28]. The mortar and fine stone play the role of support and protection and are able to withstand a certain degree of load. In addition, they can be effectively integrated with concrete structures, consistent with the role played by real aggregates. We usually call the mortar and fine stone wrapped in the outer layer as smart aggregate (SA) because of the intelligent function of the PZT wrapped in the interior. The parameters of the SAs in this study are shown in Table 1.
3. Principle
3.1. The PZT-Enabled Active Sensor Approach
The PZT-enabled active sensing approach requires at least a pair of piezoelectric actuator and sensor placed on the surface or inside of a concrete structure [29]. The actuators are mainly driven by alternating current (AC) signals. Stress waves are generated and propagated along the structure under the inverse piezoelectric effect. After the stress wave being detected by the sensor, it can realize the transformation to electric signal and finally output electric signal under the positive piezoelectric effect.
In this paper, the active sensing technology with piezoceramic SAs is used to detect debonding damage at the interface of PSCBs. The principle diagram of the active sensing approach in monitoring the debonding damage at the interface of PSCBs is shown in Figure 2. Three SAs are embedded in the concrete, in which two SA1 were used as actuators to generate stress waves and one SA2 was used as a sensor to detect the wave response. The actuator generates stress waves, which propagate from one concrete member to another through interface joints. In order to identify the interface damage state of PSCBs, the debonding crack damage of PSCB interface joints was tested by a shear experiment. When the PSCBs are in a healthy state without debonding damage before loading, the sensor can receive strong stress signals as the basic signal. As the shear experiment progresses, the received signal will attenuate when the debonding cracks occur. The signals received will decrease continuously with the increase of the number and severity of debonding cracks.
3.2. Wavelet Packet-Based Damage Index
In order to quantitatively describe the debonding crack damage of PSCBs during the loading process, the signal can be specially characterized based on wavelet packet analysis [30]. The wavelet packet analysis has not only the advantages of the Fourier transform and wavelet analysis but also the performance of the time domain and space domain as well as the more accurate decomposition of low and high frequencies. Therefore, this research uses wavelet packet principle to analyze the received energy values under different loads, forming damage index to identify the occurrence and severity of PSCB interface debonding damage.
S is assumed to be the original monitoring signal. The principle of wavelet packet analysis is used to decompose S into multiple component signals of equal width band by N-layer wavelet packets.
S is decomposed and reconstructed by the N-layer wavelet packet to obtain subsignal Si with 2N different frequency bands in the final layer [17]. S can be expressed as
The energy vectors of subsignals in each frequency band of the final signal are defined after the signal is decomposed by a wavelet packet, as shown in the following equation:where ei is the energy of each frequency band subsignal in the final layer [14], andwhere n is the number of samples points of the original signal and represents the data points in the signals of each frequency band in the final layer obtained after the decomposition and reconstruction of S.
Then, the sum of the energy vectors obtained from the reconstruction and decomposition of S by the wavelet packet is as follows [17, 31]:
The damage index is used to determine the health status of the concrete structure using the following formula to evaluate the damage area and extent of the specimens [32]:where E1,i represents the signal energy in the structural health state and Ek,i represents the signal energy of the structural damage state (k loading phase). DIk indicates the energy loss caused by the debonding crack damage of the PSCB interface in the K loading stage. Before the shear test, when there is no debonding damage at the interface, the damage index is zero, indicating that the interface is in a healthy state. When the damage index is greater than the critical value, it means that debonding cracks will occur between the interfaces in the shear experiment. A larger value of the damage index indicates that the cracking of the structure is more severe and the degree of expansion is higher.
4. Experimental Study
4.1. Specimen Fabrication
The specimens consist of three single-tooth concrete components, which are assembled with epoxy resin and embedded in the SAs in the concrete beam. The cement used for casting the concrete beam is type 32.5 Portland cement. The mixture ratio of the concrete is shown in Table 2. The average compressive strength of the concrete for 28 days is 55 MPa. Each specimen with single-tooth joint has longitudinal distribution reinforcements with a diameter of 12 mm and steel stirrups with a diameter of 10 mm. The protective layer thickness of longitudinal distribution reinforcement and steel stirrup is 30 mm. The yield strength of the longitudinal distribution reinforcement and steel stirrup is 335 MPa, and elastic modulus of the longitudinal distribution reinforcement and steel stirrup is 200 GPa. The size of the concrete beam is maintained at 600 mm × 300 mm × 100 mm. Figure 3 shows the 3D model of the specimen. The specimen details are shown in Figure 4. In the experiment, steel wires were used to fix SAs on steel bars. The two aggregates SA1 and SA2 are, respectively, located in the middle of the segment beam.
4.2. Instrumental Setup
The experiment equipment for the PSCB interface damage monitoring system used in this experiment mainly includes concrete specimens with epoxy resin joints, a data acquisition system (NI-USB 6366), a laptop with supporting software, and a hydraulic jack loading device, as shown in Figures 5 and 6. NI-USB 6366 integrates the signal generator and receiver, which can make actuator SAs produce sinusoidal wave signals and the sensor SAs collect response signals. The monitoring signal is selected to scan the sinusoidal wave. The NI LABVIEW software was used to write programs supporting the data acquisition system NI-USB 6366 to determine input signal parameters. Table 3 shows the swept sine wave signal parameters. The most prominent advantage of this method is that it can accurately simulate the changeable damage situation and meet the randomization requirements.
(a)
(b)
(c)
4.3. Experimental Procedures
During the experiment, S1-1 and S1-2 were installed on the left and right sides of the beam, respectively. This type of aggregate can be regarded as a monitoring signal transmitter. S2 installed in the middle of the beam acts as a receiver. Figure 6 depicts the condition of the specimen loading device and how it is loaded.
The hydraulic jack is used to load the specimen, and the pressure transducer is used to control the load in the experiment, which is then converted into shear force to carry out a damage test on the concrete beam. The purpose of this experiment is to study the trend of monitoring signal damage index changing with the development of cracks at the interface joint when the damage degree is strengthened. The selected parameters are shown in Table 4.
The damage monitoring system consists of signal generation, signal acquisition, and signal analysis, as shown in Figure 7. In the process of the experiment monitoring, the output of the excitation signal and the input of the acquisition signal are completed according to the control acquisition system given in Figure 7. Since the frequency value of excitation signal of the piezoceramic actuator is high, the amount of continuous data acquisition is very large. The synchronous acquisition is carried out at intervals according to the load value given by the pressure transducer, meaning that the sensor is excited once at intervals to scan the monitoring area. The loading process consists of eight operating conditions (OC1–OC8), which correspond to the loading forces of 0, 20, 30, 40, 45, 50, 55, and 60 kN, respectively.
The concrete reached the standard strength after 28 days of curing and the loading experiment was started. During the experiment, the loading method shown in Figure 6 was adopted. The load is controlled by the pressure transducer, and the experiment monitoring system is used to monitor the signal change with the sine sweep frequency wave. Finally, the signal is collected and input to the computer terminal. The feedback monitoring signal is used to process and analyze the experimental data, and the experimental results are obtained.
5. Experimental Results
5.1. Time-Domain and Frequency-Domain Analysis
In the shear experiment, SA1-1 and SA1-2 are used as actuators, respectively, and the received signals of sensor SA2 are shown in Figures 8 and 9. Each figure reflects the sensor signal response of a period from the swept sine wave signal. The results show that the amplitude of the signal decreases when the shear vertical splitting microcracks occur at the interface of the specimen. In addition, the signal attenuation intensity is more obvious with the increase of crack severity. However, the amplitude does not show a significant change from OC7 to OC8, which indicates that the shear vertical splitting crack reaches the limit level of epoxy resin thickness whose structural crack exceeds 2 mm. It is preliminarily judged from Figures 8 and 9 that OC4 is a mild damage and OC7 is a severe damage. Figure 10 shows the damage status at different stages of the PSCB joints. In the stage of experiment monitoring, the three operating points (OC1, OC4, and OC7) in the process of specimen failure are regarded as the important collection points of experimental data. In the process of experiment, every state of the structure shows the following: (1) when the specimen is not loaded, it is in a completely healthy state; (2) when the specimen is loaded, the mild damage with clear cracks begins to appear at the bottom of the specimen; and (3) the specimen is loaded until the crack develops into serious damage. The damage state of concrete beams in each stage is shown in Figure 10. Fourier transform is also used for further frequency analysis of received signals. Figures 11 and 12 show the frequency domain signals corresponding to the time-domain signals at the important operation points (OC1, OC4, and OC7). Compared with time-domain signals, the downward trend of power spectral density (PSD) energy can be seen more easily in the frequency domain.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(a)
(b)
(c)
(a)
(b)
(c)
(a)
(b)
(c)
5.2. Wavelet Packet Energy Analysis and Damage Index
In order to quantify the signal energy detected in the loading process, the wavelet packet energy analysis method is used to calculate the signal energy. The energy levels of SA1-1 and SA1-2 sensors during loading are shown in Figure 13. When debonding cracks occur, the energy value decreases sharply. Wavelet packet energy has the ability to detect debonding crack initiation. After the occurrence of debonding cracks, the value of energy decreasing continuously monitors the development of debonding cracks.
(a)
(b)
Similarly, in order to quantitatively analyze the interface damage degree of PSCBs, the wavelet packet damage indices of SA1-1 and SA1-2 are calculated, as shown in Figure 14. According to the definition of the damage index, OC1 represents the healthy state of the specimen. It can be seen that the value of the damage index increases with the increase of load for SA1-1 and SA1-2. A significantly increased value for each SA can be found in Figure 14 (OC4). According to the preliminary analysis, the debonding damage occurred in the 40 kN stage (OC4) of shear experiment, resulting in a sharp loss of stress wave energy between SAs. Therefore, the two values of OC4 suddenly increase as shown in Figure 14, indicating that the interface is damaged. When the microcracks appear, the loading force increases continuously until the crack width exceeds 2 mm and the thickness of epoxy resin adhesive is exceeded, which is called severe damage. This phenomenon occurs in OC7, and the value of OC7 in DI is close to 1. This means that the interface of the PSCBs is in a state of complete debonding. The DI value after OC7 is kept near 1 since the structure is completely debonded. The results of this analysis are consistent with the results of Figures 8, 9, 11, and 12, which explains the significant decrease in the amplitude of the SA1-1 and SA1-2 signals in the time domain at OC4 when debonding occurs between the PSCB interfaces.
(a)
(b)
6. Conclusions
This paper presents an active sensing approach for detecting the interface debonding damage of PSCBs. Actuators and sensors for interface damage detection can be formed by embedding SAs in PSCBs. The relationship between interfacial debonding damage and energy consumption is studied by the shear test using wavelet packet theory. Based on the experimental results, the following conclusions can be drawn:(1)The method given in this study can well reflect the development trend of damage and timely monitor the development of cracks in the joint of epoxy resin adhesive. The results show that the PZT-enabled active sensing approach based on the damage index of wavelet packet can effectively monitor the debonding state of the PSCB interfaces.(2)Experiments have demonstrated that embedded SA-induced stress waves are sensitive to the debonding conditions of the interface. The occurrence of debonding damage leads to cracks, which greatly reduces the energy of stress waves propagating at the interface. It can be seen that the amplitude of the signal received from SA sensor decreases with the increase of debonding damage from the time-domain and frequency-domain analysis.(3)The energy analysis and damage index based on wavelet packets can quantitatively evaluate the bonding state of the PSCB interface. With the increase of severity of debonding, the energy value of wavelet packet decreases and the value of damage index increases correspondingly. The initial and complete debonding stages can be successfully reflected in the damage index.
In addition, the damage index defined by wavelet packet theory is highly sensitive to damage. The predicted structural failure occurred earlier than the real failure. The proposed active sensing approach based on wavelet packet damage index has great potential to be applied in practice for inaccessible damage detection of PSCB interfaces.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
The authors are grateful for the partial financial support received from the Major State Basic Research Development Program of China (973 Program, grant no. 2015CB057704), the National Natural Science Foundation of China (grant no. 51378081), the Natural Science Foundation of Hunan Province (grant no. 2019JJ40313), and the Hunan Provincial Innovation Foundation for Postgraduates (CX20190651).