A Feasibility Study on Monitoring Pile-Soil Bonding Condition Using Piezoceramic Transducer and Horizontal Impact

The in situ evaluation of pile-soil bonding condition plays an important role for pile safety assessment in its life cycle. However, so far, there is still no fully mature tool to analyze such couplings, since the pile-soil coupling exhibits complex and time-varying relationships. This paper innovatively proposes a health monitoring approach to evaluate the bonding status of the soil and pile contact area. An impact method based on a piezoelectric ceramic sensor is proposed to monitor the bond of pile and soil. A horizontal impact was introduced near the top of the pile, and the induced stress waves were detected by the piezoceramic smart aggregate (SA) sensor embedded in the pile. Different crack damage sizes were made between the soil and the pile to investigate the change of the bonding. An energy index was developed to quantitatively evaluate the quality of the bonding as a pile-soil bonding index. The proposed approach inspired a potential way to directly judge if there is crack damage between the pile and soil and to evaluate pile safety.


Introduction
In the construction process, pile as a structural foundation is widely used in soft soil, in order to ensure the safety and stability of the project, and pile foundation detection has become an essential link in the construction process. The testing is conducted on the quality of hole formation, the bearing capacity of pile foundation, and the integrity of pile foundation, respectively. The testing of pile foundation can provide the data of the maximum bearing capacity of pile foundation structure and can control the construction process in real time to ensure high quality and efficiency. Though pile engineering is mostly based on the concept that soil is perfectly bonded to the pile due to the contact skin friction, in practice, soil can be denuded, destroyed, separated, transported, and deposited in the open air and under load. As a result, the contact area between soil and pile foundation will slip or even separate, making the combination of soil and pile rarely perfect. Pile-soil interaction exhibits complex and time-varying mechanisms, which has attracted much investigation. For example, Pak and Ji [1] studied the rational mechanics of axial soil-pile interaction. Abedzadeh and Pak [2] investigated a cylindrical pile interaction with transversely isotropic soil. Shahmohamadi et al. [3] studied the static axial soil-pile interaction in a cross-anisotropic medium. The above is the study of pile-soil interaction under static state, but in life, most pile-soil interaction is under dynamic state. Chen et al. [4] studied a pile response under transient torsional loading, and Wang et al. [5] investigated the dynamic torsional response of an end-bearing pile in transversely isotropic saturated soil under harmonic excitations. In practice, low-strain and high-strain integrity testing methods have been widely used for assessing the construction quality of piles in civil engineering [6][7][8]. The effect of the dynamic loading or excavation unloading on p-y curves for laterally loaded piles has been studied as a tool for assessing the pile property [9,10]. However, due to the very complex interaction between piles and soil, there is no tool or technology that can truly and accurately monitor the in situ response of the pile-soil contact area, read the interaction data between piles and soil directly, and monitor the performance of piles and soil in the life cycle for a long time. Thus, important issues have to be addressed properly to perform monitoring of the pile-soil coupling for the pile structural safety.
Structural health monitoring (SHM) technology has become an increasingly important research topic to economically maintain infrastructure [11][12][13][14]. Because of its special working environment, pile inspection can only be performed via nondestructive testing (NDT) methods. Additionally, pile monitoring is a challenging task and monitoring by using embedded piezoceramic transducers [15][16][17]. Due to the advantages of low cost, quick response, solid-state actuation, and embeddability, piezoceramic materials have been used as actuators and sensors for health monitoring purposes [18][19][20][21]. To protect the fragile piezoceramic transducer for applications in concrete structures, the piezoceramic smart aggregate (SA) was introduced [22]. At present, the smart aggregate sensors have been used to detect the damage of the concrete [23][24][25][26][27], metal [28,29], and composite structures [30,31]. Using piezoelectric smart aggregate, Zhou et al. [32] studied the attenuation characteristics of stress waves in cracked concrete beams from the perspective of stress wave and time inversion techniques. In addition, an active sensing approach has been used in the health monitoring of various civil infrastructures [33][34][35]; however, not enough research has been yet conducted on pile-soil bonding monitoring and assessments, to the best knowledge of the authors.
In this paper, a piezoceramic sensor-based impact approach is developed to monitor the pile-soil coupling. The lateral load was excited by an experimental hammer on the upper body of the pile. Various crack strip sizes were applied to investigate the different degrees of bonding between soil and pile during the experiment. An energy index was developed to quantitatively evaluate the bonding. The induced stress waves were detected by the SA sensor embedded in the pile. This paper innovatively proposed a health monitoring approach to evaluate the bonding status of the soil and pile. This method can nondestructively monitor the combination of pile and soil, without damaging the pile itself, and can evaluate the safety of the pile, increasing the research method of the direction of the combination of pile and soil detection and evaluation.

Monitoring Principle and Algorithm
2.1. Principle of the Horizontal Impact Response Using SA Sensor for Studying the Pile-Soil Bonding. Lead Zirconate Titanate (PZT) is one of the commonly used piezoelectric materials which can generate an electric charge when subjected to mechanical stress (direct effect) and, conversely, generate a mechanical strain in response to an applied electric field (converse effect). From previous studies, we know that the induced voltage is proportional to the dynamic stress on the PZT sensor [36,37]. The smart aggregate shown in Figure 1(a) employs a PZT patch, and the detailed construction of the smart aggregate is shown in Figure 1(b). The material type of the PZT used here is PZT-5H, which is highly sensitive to dynamic measurement. Table 1 shows the related parameters of PZT-5H. The size of the PZT patch is 10 mm × 10 mm × 0:3 mm. When the SA transducer is embedded in a concrete structure, a stable mechanical performance is ensured by this sandwiched structure. Due to their advantages of high bandwidth and strong piezoelectric effect, PZT transducers are commonly used to detect signals with high-frequency components [38,39], such as ultrasonic waves [40][41][42][43] and shock-or impact-induced stress waves [44][45][46][47]. In this research, a piezoceramic SA-based transient impact sensing approach is utilized to monitor the bonding of pile and soil.
In the proposed research, as shown in Figure 2(a), on the top of a reinforced concrete pile which is used as a testing pile, a horizontal vibrating hammer was used as an actuator by transient impact, and smart aggregate (SA) was functioned as a sensor to detect the response. To fix the SA during pouring the concrete, it was attached to the nearby steel rod next with epoxy. The cross-section of the test pile is shown in Figure 2(b). According to previous studies, when the  2 Journal of Sensors piezoelectric sensor works in high humidity or underwater environment without waterproof protection, the dielectric constant, insulation resistance, and dielectric loss of piezoelectric ceramics will be affected after hydraulics absorption. The dielectric constant goes down. Insulation resistance will drop significantly. The dielectric loss will increase, in the high humidity environment, and the response characteristics of the piezoelectric sensor will become worse, or even fail, so it is necessary to choose high insulation materials and take waterproof measures to avoid the piezoelectric ceramic damp. So we put the sensors between the concrete blocks, and we put them in epoxy resin to make them waterproof. The existence of crack damage between pile and soil will induce a quite different boundary restraint on the pile. According to PZT electric-mechanical effect relationship, the severity of the soil constraint on the SA sensor is greater which means PZT mechanical performance is more restrained, and the SA sensor response signal will be weaker. By analyzing the response signals, the bonding status of the pile by soil can be evaluated.

Basic Principles of the Time Domain Analysis.
The propagation of the test wave signal between pile and soil is sensitive to the degree of bonding between the pile and the soil. In this paper, the time domain analysis is applied to study the signal received by the SA sensor. The energy of the signal is calculated to measure the degree of the bonding. In the proposed health monitoring algorithm, a signal energy index E was defined. Use X i to represent a set of discrete data of a signal obtained by the sensor in a given sampling duration at time i, and x ij is the value of the sensor voltage of the jth sam-pling point at time i. A total of m sampling points contributes to the calculation in each sampling duration, and x i is the average value of the m sampling points' sensor voltage at time i. S i is equal to the integral area enclosed by the signal curve recorded at time i and the x axis. X i can be expressed by the following equation: To monitor the bonding status between pile and soil, an energy indicator is defined as The E i will be used as an indicator of the pile-soil bonding condition. The proposed energy index shows the wave transmission energy detected by the SA sensor and will be used to define the coupling relationship between soil and pile. In this experiment, the cracks existing between piles and soil were temporarily considered. Moreover, the microcracks of concrete had little influence on this experiment, so they could be ignored. With the existence of crack damage between pile and soil, the boundary restraint of the pile by soil is weaker, thus less wave energy will be dissipated into the soil, resulting in more energy detected by an SA sensor. From the sensor response in the pile, the pile-soil bonding conditions can be investigated. But because the SA performance is greatly influenced by environmental temperature, in the process of heating up, the low-frequency signal amplitude (below 1 kHz) with the temperature rising slowly rising first, when the  Journal of Sensors temperature exceeds 30°C after entered a stage of rapid growth, this is mainly due to the PZT piezoelectric coefficient, dielectric constant, and the change of SA packaging material properties with temperature. The mode frequency of PZT thickness direction and the compressive stress in SA also increase with the increase of temperature. Therefore, the ambient temperature has a great influence on the sensor. However, this experiment mainly studies whether there is damage between piles and soil. Therefore, the ambient temperature is kept at 20°C to reduce the influence of temperature on the sensor signal.

Experimental Setup
3.1. Experimental Setup. The testing soil was filled in the test box with sizes 500 mm ðlengthÞ × 450 mm ðwidthÞ × 800 mm (height), and a 1500 mm long-reinforced concrete pile with a cross-sectional radius of 4 mm was buried in the soil, as depicted in Figure 2(a). The soil tested here is clay, and the density of the soil measured by the ring knife method is 1436 kg/m 3 . The experiment system is shown in Figure 3(a). INV306U signal acquisition instrument is selected to record the signal. The sampling frequency and recording time are 7160 Hz and 0. 1 s, respectively. A SA functioned as a sensor to receive the real-time signal, which was embedded in the pile, and the location of the SA sensor in the pile is shown in Figure 2(a). The horizontal transient impact was applied by a hammer, and the controlled impact location was read from the ruler which was bonded on the top surface of pile as shown in Figure 3(b). When the horizontal vibration was applied on the pile, the signals received by the SA sensor were different with the existence of crack damage; thus, signal variation can be used to estimate the bonding degree between the pile and the soil. Two types of crack damage modes were conducted to investigate the bonding degree between the pile and the soil.

Experimental Procedure.
Because of the influence of the environment will make the slide pile-soil contact area of separation, even in this experiment, we studied two kinds of different damage models in group A and group B, and the assumed pile without cracks, the main research of single crack between pile and soil, for concrete cracks between soil and pile cracks will be in the future experiments for further research, as shown in Table 2. The width of the cracking damage was selected as 2 cm for all test cases in group A and group B. In group A, cracking damage depths were varied over four sizes: 0, 5, 10, and 15 cm (see Figure 4). The cracking damage range was selected as 90°for all test cases in group A. For group B, the cracking damage range was varied over five values: 0°, 90°, 180°, 270°, and 360°(see Figure 5). The cracking damage depth was selected to be 5 cm for all test cases in group B. Table 2 gives the crack damage status of all the test cases in this study. The horizontal vibration was induced by a hammer impacted on the top surface of the testing pile. For a fair comparison, the initial horizontal transient impact energy shall be maintained the same; therefore, the hammer was released from the same height. This practice is controlled by using the installed ruler, as shown in Figure 3(b). And when the hammer struck on the pile actuating the initial signal, the SA sensor began to respond to the impact, and response variations would reflect the bonding differences. SA sensor response signals were recorded by the data acquisition system, as shown in Figure 3(a).

Experimental Results and Discussion
The time domain signal responses of the SA sensors of groups A and B are given in Figures 6 and 7, respectively. The curve in the figure is the sensor signal response of a horizontal impact, and the sampling time is 1 second. Group A and group B received different SA sensor signals due to

Journal of Sensors
proportional to the deformation of the SA; thus, the amplitude of the signal recorded by the SA increased when the crack damage occurs and increases.
In order to quantitatively evaluate the attenuation of the stress wave energy, the energy indicators of all the test cases in groups A and B were calculated based on Equations (1) and (2). The index value increases with the increasing depth and range of crack damage, as shown in Figures 8 and 9. Because the pile-soil bonding was decreased by the crack damage, relatively more energy remained on the SA and less energy could be transmitted through the damaged pile-skin soil. Compared with signal response in the time domain, the damage indicators show promise for estimating the bonding between pile and soil. The data presented by realtime monitoring can not only warn damage but also evaluate the severity of pile debonding in real time.
Through some papers and the experimental data, it was shown that the active sensing method based on stress wave is able to monitor a wide range of structural damage, can use more stress wave signals, and can actively, real-time, continuously monitor the structure of the health status, therefore having great potential for development, but some problems   Journal of Sensors still exist. For example, the crack damage mode in pile-soil cannot be identified using this method. Also, the effect of humidity, epoxy resin in smart aggregate, and temperature has not been considered in this research. Cracks caused by concrete and multiple cracks between pile and soil have not been considered in this experiment. In the future, further research works should be done to figure out these issues.

Numerical Verification
In order to further study the impact of cracks on pile-soil, we used ABAQUS software to build a finite element model. Figure 10(a) is a sketch of the computational model. Like the laboratory model, the finite element model is composed of two parts, namely, clay and pile. The clay model is 60 cm long, 60 cm wide, and 70 cm high. The model pile is 150 cm in height and 8 cm in diameter. We set up four models, respectively, and set the nondamage data as the healthy state. There are three damage states, respectively, with a depth of 5 cm, 10 cm, and 15 cm. In order to facilitate the comparison of experimental data, the crack ranges of these three models are set at 90°. The density of soil is set at 1,436 kg/m 3 , the pile is made of C30 concrete, the elastic modulus is set at 30 GPa, and Poisson's ratio is set at 0. 2. Pile and soil are connected by bonding. Figure 10(b) shows the grid model of health and injury states. The software allows us to observe the stress at different depths after a horizontal instantaneous impact on the hammer. Figure 11(a) is the stress result for the healthy there is no obvious change of stress magnitude on the right side of the crack, while the stress magnitude on the left side of the crack increases with the increase of the crack depth. This is because the stress wave cannot propagate through the crack, and more stress wave energy remains inside the pile. The deeper the crack depth is, the wider the range of stress diffusion is. Figures 12(a)-12(d) show four damage states at a depth of 10 cm but with different division ranges. The splitting range was 90°, 180°, 270°, and 360°, respectively. The sensor is mounted on piles 10 cm below the surface of the soil. As can be seen from the figure, within the range from the soil surface to 10 cm down, with the increase of the splitting range, the smaller the soil-pile adhesion, the smaller the interaction between the pile and soil, and more stress wave energy remains on the pile. Therefore, under the action of the same external force, the larger the splitting range, the     Journal of Sensors deeper the stress diffusion, and the stress value in the same place will increase. The above model reflects that the stress value on the pile will increase with the increase of the crack depth and range, that is, the index value of the signal on the piezoelectric ceramic sensor will increase.

Conclusions
To monitor pile-soil bonding conditions, an innovative approach based on transient impact response using smart aggregate (SA) sensors was proposed. Two typical types of debonding damage were studied in this study. The effects of crack depth and crack range were investigated by changing their sizes. It can be observed that the signal amplitude becomes larger with an increase in crack damage depth and range. The increase of crack damage depth and range results in lower pile-soil bonding. However, the lower bonding will give rise to the larger amplitude of the piezoceramic sensor signal. In addition, the energy indicator values offered a quantitative assessment of the bonding severity. Damages on the pile-soil interface, such as slippage or separation, can be detected by the SA sensor responses through comparing with the data of no damage. The case study inspires a potential way to judge if the bonding between the pile and soil fails.
The study can be used as a monitoring of whether the pile and soil are tightly bonded during the whole construction      Journal of Sensors process. It can also be used as a structural health monitoring to determine whether the cohesion between pile and soil is reduced due to environmental factors in the future.

Data Availability
The experimental data used to support the findings of this study are included within the article.

Conflicts of Interest
The authors declare that they have no conflicts of interest.