An electromechanical impedance-instrumented corrosion-measuring probe

Abstract

Corrosion of metallic structures widely existed in multiple industries, such as oil and gas, civil infrastructure, aerospace, mechanical, mining, and processing. Current available corrosion-monitoring methods are based on different sensing principles, which have their own advantages, and some drawbacks that may limit their application on some aspects. This article presents an electromechanical impedance-instrumented corrosion-measuring probe for corrosion monitoring. The proposed probe is fabricated by attaching a circular lead zirconate titanate patch onto a metal rod. Compared to other electromechanical impedance-based corrosion-monitoring methods, the probe is capable of isolating the influence of structural complexity, variations in loading and boundary conditions. Five probes were fabricated in the experimental study and three of them were subjected to accelerated corrosion tests to mimic the corrosion-induced mass loss damage. Results showed that the peak magnitude of the conductance signatures was reduced with the increase in corrosion amount. The variations in the conductance signatures were quantified by three statistical quantifying metrics, that is, root-mean-square deviation, mean absolute percentage deviation, and correlation coefficient deviation. All these metrics increase with the increase in corrosion amount, which can be used as an indicator of the corrosion process. This study proves that the proposed corrosion-measuring probe is effective in monitoring corrosion and shows promising application potential. This research also serves as a proof-of-concept study to demonstrate the capability of the electromechanical impedance technique in monitoring mass loss due to corrosion.

Keywords

electromechanical impedance corrosion-measuring probe corrosion monitoring lead zirconate titanate metallic structures

1. Introduction

Corrosion of metallic structures widely existed in multiple industries, including oil and gas, civil infrastructure, aerospace, mechanical, mining, and processing. Corrosion-related issues can pose great threat to the safety and serviceability of the structures during operation and could possibly induce economy loss and even life loss in severe situations. According to a report by World Corrosion Organization, the cost of corrosion is estimated to be US$2.5 trillion worldwide (Koch et al., 2016). However, the corrosion loss can be greatly reduced if corrosion-monitoring and corrosion-measuring methods were adopted and proper countermeasures were undertaken. It is thus very important to develop effective corrosion-monitoring methods, assess the corrosion severity accurately, and evaluate the safety performance of the metallic structures.

Various corrosion-measuring methods have been developed over the last few decades. Since corrosion is an electrochemical process, the conventional ways of measuring corrosion are the electrochemical methods (Luo et al., 2019), including half-cell potential measurement, linear polarization resistance measurement, electrochemical impedance spectroscopy, electrochemical noise, and harmonic analysis. The results from these methods can suggest the likelihood of corrosion, which are qualitative. The major drawback of these methods is that they have limitation in localized corrosion monitoring. Recently, the advancement in the field of smart materials and structural health monitoring has opened the door for new generation of corrosion-measuring or corrosion-monitoring methods, to name a few, the fiber optic sensing (Li et al., 2016a; Zhao et al., 2011), acoustic emission monitoring (Kawasaki et al., 2013; Li et al., 2017b), and ultrasonic testing (Sharma and Mukherjee, 2015). Although these new methods have been proven to be promising, they are still in their early stage of development and further investigation is required. Also, the instrumentation for these methods is expensive, which may limit their widespread application.

Piezoelectric-based structural health monitoring involves the use of piezoelectric sensors and/or actuators to detect damages in the structure by evaluating the performance parameters. To date, extensive research endeavors have been conducted on the piezoelectric-based structural health monitoring, for example, the bolted structure monitoring (Song et al., 2017), debonding detection in fiber-reinforced polymer rebar-reinforced concrete (Li et al., 2017a), concrete infill monitoring (Luo et al., 2016), damage and retrofitting monitoring in reinforced concrete structures (Kaur et al., 2019), curing process monitoring of structural adhesives (Lim et al., 2018), hardening process monitoring of ultra-high performance concrete (Lee et al., 2018), and novel applications of smart aggregate for health monitoring of concrete structures (Du et al., 2018; Kong et al., 2013; Li et al., 2016b; Liu et al., 2013; Wu et al., 2017). Among them, the electromechanical impedance (EMI)-based damage detection method using a lead zirconate titanate (PZT) is becoming a powerful tool for local damage detection and evaluation. Different types of damages, such as pre-stress loss (Fan et al., 2018a; Huynh et al., 2018; Ryu et al., 2019; Wang et al., 2017, 2018), debonding (Li et al., 2018; Liang et al., 2016; Wu et al., 2018), stiffness change (Fan et al., 2018b; Lu et al., 2018a, 2018b; Tinoco et al., 2019), and mass gain (Shi et al., 2018), can be effectively detected by such techniques. On the contrary, novel signal processing approaches were also proposed for extracting useful damage-related information (Fan et al., 2016, 2018c, 2018d). In the EMI method, the PZT patch acts as a collocated sensor and an actuator. The PZT patch is either surface-bonded onto or embedded into the host structure, and the motions between them are then coupled. According to the electromechanical coupling property, the electrical impedance or admittance (inverse of impedance) of the PZT patch is a function of the mechanical impedance of the host structure (Park et al., 2001, 2008). Structural damages, which in the forms of stiffness change, mass loss, and geometry change, will induce a variation in the mechanical impedance of the host structure and will be reflected in the electrical impedance of the PZT patch. Therefore, the variation in the electrical impedance is indicative of structural damages. The high-frequency excitation of the PZT patch enables the EMI method to be very sensitive to local and small damages. The PZT is of low cost, nonintrusive, and linear, which makes the EMI method efficient, accurate, and capable of providing real-time, remote, and autonomous damage detection.

There are plenty reports on corrosion monitoring using piezoelectric-based health monitoring techniques (Dai et al., 2019; Moustafa et al., 2014; Nagy et al., 2014). However, the available literature studies on the topic of corrosion monitoring using the EMI method are very limited. Talakokula et al. (2014) adopted the EMI method to evaluate the corrosion process of concrete rebar. The PZT patches were bonded on the surface of the rebar. The experimental results showed that the equivalent parameters extracted from the admittance signatures were indicative of the corrosion amount. Zhu et al. (2016) proposed structural mechanical impedance for corrosion detection of steel structures based on the EMI method. It was found that the structural mechanical impedance is sensitive to corrosion damage but the detecting range is limited. Na (2017) demonstrated the possibility of detecting wall thickness loss of metal-based pipeline facilities using the EMI method. The results showed that the resonance peaks in the impedance signatures were shifted with the reduction in the wall thickness. The common problem with these studies is that the PZT patch is attached to the host structures with complicated geometry and boundary conditions. Such configuration cannot eliminate other influencing factors (e.g. variation in loading conditions or boundary conditions) or damages other than corrosion.

To solve the aforementioned problems, a new type of corrosion sensor using the EMI method was proposed in our previous studies (Li et al., 2019a, 2019b). Results showed that the peak frequencies in the conductance signatures corresponding to bending modes are linearly reduced with the increase in corrosion-induced thickness loss. The sensor is best suited for uniform corrosion monitoring. However, in the case of non-uniform corrosion, the linearity may be lost and the accuracy was compromised. Under such scenario, we explore a new corrosion-measuring probe for general usage, for either uniform or non-uniform corrosion monitoring. The corrosion-measuring probe is based on the EMI technique. The probe is a metal rod with a PZT patch attached to it. The probe is thus separated from disturbances of the main structure, eliminating the influences from the variation in loading conditions and boundary conditions. When subjected to corrosive environment, the metal rod experiences mass loss and its mechanical impedance is altered, which is then measured by the electrical impedance or admittance of the PZT patch. Five probes were prepared in the laboratory and three of them were subjected to accelerated corrosion tests to simulate corrosion-induced mass loss damage. In order to quantify the variations in the impedance signatures under different corrosion amounts, the statistical metrics were adopted. These metrics are the root-mean-square deviation (RMSD), mean absolute percentage deviation (MAPD), and correlation coefficient deviation (CCD). This study serves as a proof-of-concept to show the capability of the EMI technique in monitoring mass loss due to corrosion.

2. Theoretical background

2.1. Principle of the EMI method

A PZT is a type of piezoelectric material and operates on the principle of piezoelectricity. In the direct piezoelectricity effect, the PZT serves as a sensor in such a way that electric charges are generated when deformed by external forces. In the converse piezoelectricity effect, the PZT acts as an actuator in such a way that it deforms when an electric field is applied. The PZT patches are coupled to the host structure either through the surface bonding method or the embedding method and electrically excited by an impedance analyzer with high-frequency band, and the electrical impedance signatures are measured simultaneously. The interaction between the PZT patch and the host structure (the metal rod, specifically in this case) can be idealized as an electromechanical system, as shown in Figure 1. The analytical model of this setup was first proposed by Liang et al. (1994) and subsequently implemented by many other researchers (Annamdas and Soh, 2007; Bhalla and Soh, 2004; Giurgiutiu, 2007; Park et al., 2000). The electrical admittance, $Y (ω)$ , of the PZT is associated with the structural impedance, $Z_{s} (ω)$ , through the following equation

Y (ω) = i ω C (1 - κ_{31}^{2} \frac{Z_{s} (ω)}{Z_{s} (ω) + Z_{a} (ω)})

(1)

where $C$ is the electrical capacitance of the PZT, $κ_{31}$ is the electromechanical coupling factor, and $Z_{a} (ω)$ is the mechanical impedance of the PZT. Damages to the host structure inflict changes in the structural stiffness and/or mass and thus alter the mechanical impedance. As indicated by equation (1), any changes in the mechanical impedance of a structure induce the changes in the electrical admittance measured by the PZT patch.

2.2. Statistical quantifying metrics

In the EMI method, damage detection is usually made by observing the changes in the electrical impedance signatures as compared to a baseline measurement. The results are qualitative. Therefore, scalar metrics need to be defined to quantify the difference in the impedance signatures between the damaged conditions and the baseline healthy condition. For quantifying the variations in the impedance signatures for different damaged conditions, the commonly used statistical metrics, RMSD, MAPD, and CCD, were adopted. These metrics were found to be reliable indicators for quantifying structural damage development (Park et al., 2011; Tawie and Lee, 2010; Wang and Zhu, 2011). The mathematical formulas for these metrics in terms of the real part of electrical admittance, $Re (Y)$ , are given as follows

RMSD (%) = \sqrt{\frac{\sum_{i = 1}^{N} {[Re (Y_{i, 1}) - Re (Y_{i, 0})]}^{2}}{\sum_{i = 1}^{N} {[Re (Y_{i, 0})]}^{2}}}

(2)

MAPD (%) = \frac{1}{N} \sum_{i = 1}^{N} | \frac{[Re (Y_{i, 1}) - Re (Y_{i, 0})]}{Re (Y_{i, 0})} |

(3)

CCD (%) = 1 - \frac{1}{N σ_{Y_{1}} σ_{Y_{0}}} \sum_{i = 1}^{N} [Re (Y_{i, 1}) - Re ({\bar{Y}}_{1})] \cdot [Re (Y_{i, 0}) - Re ({\bar{Y}}_{0})]

(4)

where $N$ is the number of sampling points in the signatures; the subscripts $0$ and $1$ represent the baseline measurement and concurrent measurement, respectively; the symbols $\bar{Y}$ and $σ_{Y}$ represent the mean value and standard deviation of the signatures, respectively. The larger the number of metrics, the greater the difference between the baseline measurement and the concurrent measurement, which represents that more severe damage occurred to the structure.

3. Experimental investigation

3.1. Fabrication of the corrosion-measuring probe

The EMI-instrumented corrosion-measuring probes were fabricated in the laboratory. The probe is a metal rod with a PZT patch attached to it, as shown in Figure 2. The metal rod has a diameter of 20 mm and a length of 100 mm. The material for the metal rod is Q235 steel in Chinese standard, which is equivalent to A36 mild steel in US ASTM standard. The PZT patch is a circular one, which has the same diameter as the metal rod of 20 mm and has a thickness of 1 mm. The PZT patch was bonded onto the metal rod using epoxy. Then, the PZT patch and the connection wires were coated with epoxy to provide protection from corrosive environment and external disturbance. Apart from the two electric wires for the PZT patch, an additional wire was added to the rod for impressing electric current during the accelerated corrosion test. Therefore, copper tape was wrapped around the metal rod, and the electric wire was soldered onto the copper tape. All the electric connections were protected using epoxy. Five probes were fabricated. Three of them were used in accelerated corrosion tests and two of them were used as reference probes and for observing the stability and the temperature influence.

3.2. Accelerated corrosion of corrosion-measuring probe

To simulate the material loss or mass loss due to corrosion, the fabricated probes were subjected to accelerated corrosion tests. Prior to the corrosion tests, the initial weights of the metal rods were measured, whose values are about 250 g. The setup for the accelerated corrosion tests is shown in Figure 3. Three probes were subjected to corrosion tests. Each of them was placed in a separate beaker filled with 3.5% NaCl solution. In the accelerated corrosion tests, the anode was connected to the probe, and the cathode was connected to a copper plate. Both the probe and the copper plate were immersed in the solution. The DC power supply was used to impress current, whose value was set at 100 mA. Corrosion on the probe was initiated upon the application of electric current.

3.3. EMI measurement

The EMI measurement was taken every 4 days. The corroded probes were taken out of the water and weighted after being cleaned and dried, so as to obtain the mass loss and corrosion amount. Also, it was ensured that the EMI measurement was performed under the same condition. The setup for the EMI signature measurement is shown in Figure 4. The EMI signatures of the probes under different corrosion amounts were measured using an impedance analyzer (PV520A; Beijing Band Era Co., Beijing, China). The EMI signatures were then transferred to a computer via an RS232 cable for further processing and analysis. For the EMI method to work properly, selecting the suitable scanning frequency range is very important. Therefore, a wide frequency range is first scanned. By observing the peaks and valleys in the signatures, a refined frequency range containing multiple peaks or peaks with large magnitude is chosen because more structural dynamics were presented in this range and thus it is more sensitive to damage. First, a broad frequency range between 10 and 500 kHz was scanned for the EMI signatures. Then, the refined frequency range 80–150 kHz was selected since this range presents a peak with large magnitude in the conductance signatures. The peak frequency in the conductance signatures corresponds to certain resonant frequency of the probe. The range with large peak usually contains more dynamic interaction, and it is therefore more sensitive to damage (Lim and Soh, 2014; Yang et al., 2008). The measurements during the whole corrosion process were performed at room temperature varying between 24°C and 27°C.

According to Faraday’s law, for each 4-day period, the mass loss in the metal rod is predicted by the following equation (Li et al., 2016a)

Δ m = \frac{MIt}{zF}

(5)

where $Δ m$ (g) is the mass loss of the steel consumed, M (g) is the atomic weight of metal (56 g for Fe), I (A) is the applied current, $z = 2$ is the ionic charge, $F = 96500$ (A/s) is the Faraday’s constant, and t is the duration. According to equation (5), the predicted mass loss for each 4-day period is 10 g, which corresponds to a corrosion amount of about 4% (10 g/250 g). Table 1 summarizes the measured mass loss of the three corroded probes and the predicted mass loss. The measured mass loss is in good agreement with the predicted mass loss. In the following analysis, the mass loss is used to indicate the corrosion amount.

Table 1.

Measured mass loss of the three corroded probes and the predicted mass loss.

Corrosion day(s)	Mass loss (g)
Corrosion day(s)	Probe 1	Probe 2	Probe 3	Prediction
0	0	0	0	0
4	11.17	10.91	10.88	10
8	20.01	19.18	19.99	20
12	30.62	29.08	30.23	30
16	40.07	39.91	39.82	40
20	49.4	49.71	49.56	50
24	57.91	58.78	58.52	60
28	66.93	68.07	67.51	70

4. Results and discussion

The initial scan of conductance signatures in the frequency range from 10 to 500 kHz is shown in Figure 5. The conductance, which is the real part of admittance, was chosen among other EMI parameters since it is proven to be more sensitive to damage. As can be seen, a peak with large magnitude is presented in the range between 80 and 150 kHz for all the five corrosion-measuring probes. Therefore, this refined frequency range is chosen for any further analysis. Note that the conductance signatures of these corrosion probes may not be exactly the same due to several factors. First, the EMI characteristics of the PZT patch present slight deviation in its manufacturing process. Second, in the fabrication process of the corrosion probes, the thickness of bonding epoxy and the protective epoxy coating may not be uniform among the probes, which may induce some differences in the conductance signatures. In our method, the corrosion assessment is made by comparing the signatures of present corrosion state with its baseline measurement. Therefore, the measurement is not affected by the differences among different probes.

The conductance signatures for the three corroded probes under different corrosion amounts are shown in Figures 6 to 8. The conductance signatures for all the three corroded probes share the same tendency due to corrosion damage. It can be seen that with the increase in corrosion amount, the peak magnitude of the conductance signatures is reduced. The overall curve of the conductance signatures is flattened and smoothed. Note that the peak magnitude reduced all of a sudden at the very beginning of the corrosion process, indicating the high sensitivity of the EMI-instrumented corrosion-measuring probe. The changes in the conductance signature can be explained by the fact that corrosion induces mass loss and/or stiffness loss to the corrosion-measuring probe, which are responsible to the variations in mechanical impedance that will be measured by the EMI signatures of the PZT patch. Therefore, observing the changes in the conductance signatures, it is possible to identify the occurrence of corrosion.

The tendency of the conductance signatures of these three probes under different corrosion amounts is very similar, which proves the consistency of the method. However, the amount of changes in the signatures can be different for different probes. One major reason is that the corrosion morphology for each probe can be different during the accelerated corrosion tests. In an ideal situation, under the impressed current, the material loss of the metal rod is uniform for the exposed surfaces. In practical situation, such uniformity is usually not the case due to imperfection of the materials, as illustrated in Figure 9, which shows the morphology of the corroded probes.

The photos of the corroded probes for 4% and 28% corrosion amounts are shown in Figure 9. The corrosion of the probe reduces its cross-sectional area, which changes the mechanical properties of the probe, notably the mass loss and stiffness loss. Therefore, the variations in mechanical properties are measured by the EMI signatures of the PZT patch.

Peak frequency movement could serve as the indicator of structural damage related to stiffness change, and it was well-proven in concrete strength gain monitoring (Negi et al., 2018; Shin et al., 2008) and our previously proposed corrosion sensor (Li et al., 2019a, 2019b). However, for the probe proposed in this study, the peak frequency movement at different amounts is not consistent, as shown in Figures 6 to 8. Only the peak frequency of Probe 3 is reduced with the increasing corrosion amount monotonically. Results from other two probes are not conclusive. There are several physical parameters accounted for the changes in the conductance signatures, including mass, stiffness, and damping. Conceptually, increasing damping results in the reduction in sharpness of the peak and the curve is smoothed. For a simple mechanical system, the resonant frequency is expressed as $f = 1 / 2 π \sqrt{k / m}$ , which is dependent on both stiffness and mass. For the proposed corrosion-measuring probe, both the stiffness loss and mass loss occur. It is interesting to note that if these two parameters are changing at the same scale or close to the same scale, no obvious change in the resonant frequency would be observed since they were canceled out. This phenomenon could potentially be responsible for the inconsistency in peak frequency movement. Also, as shown in the pictures of the corroded probes, different probes may present different corrosion morphologies, which increase such inconsistency. Fortunately, as the stiffness and mass are dependent on the geometry of the probe, tuning the length and diameter of the probe could solve such an issue, which will be a topic of our further research. Furthermore, in case the peak frequency fails, the statistical metrics can serve as a reliable damage indicator.

The corrosion assessment made from the conductance signatures is qualitative. To quantify the variations in the conductance signatures under different corrosion amounts, statistical metrics, RMSD, MAPD, and CCD were adopted. Figures 10 to 12 present the changes in these metrics of the three corroded probes as the corrosion amount increases. It can be observed that all these metrics increase with an increase in corrosion amount. The values of RMSD are varying between 30% and slightly above 100%, those of MAPD are varying between 30% and 700%, and those of CCD are varying between 10% and 80%. Particular attention should be paid to the metrics of Probe 2, where the RMSD and CCD experience large fluctuation for corrosion amount of 16% and MAPD possess relatively accurate tendency. The overall tendency of RMSD and CCD is increasing. All these three metrics can be served as indicators of the corrosion process. Among the three metrics, the MAPD shows the highest sensitivity and accuracy. These results have proved the feasibility of the proposed corrosion-measuring method.

It can also be observed from these figures that each probe presents different metric values for the same corrosion amount. The reason for this phenomenon is the same as the aforementioned analysis, that is, the corrosion morphology for each probe can be different during the accelerated corrosion tests. As stated earlier, the corrosion assessment is made by comparing the signatures of the present corrosion state with its healthy baseline measurement. If an increasing trend in the values of the metrics is observed, the corrosion amount becomes more severe. The metric values for the same corrosion amount are not comparable among the probes. Thanks to such limitation, the proposed corrosion-measuring probe cannot be used to quantify corrosion amount. Instead, it can be used to evaluate corrosion qualitatively with the remote and online monitoring capability.

The conductance signatures of one of the reference probe for different measurement durations are shown in Figure 13. Also, the embedded plot shows the details of one peak frequency in the range of 95–100 kHz. It can be seen that the signatures are very consistent over the 28-day duration, showing that the performance of the probe is very stable, the repeatability is good, and the requirement for temperature compensation is minimal in this case. The quantifying metrics for the reference probe are shown in Figure 14. Their values are relatively stable over the 28-day period. The largest variations in RMSD and MAPD are both slightly above 15%, and that in CCD is below 2%, which further testifies the stability of the corrosion-measuring probe.

When it comes to real-world application of the proposed EMI-instrumented corrosion-measuring probe, the environmental conditions, that is, wet and dry conditions, may vary. It may influence the performance of the probe. To investigate the influence of wet and dry conditions, the comparison of the conductance signatures for these two conditions was made. The dry condition means that the measurement was made in the cleaned and dry condition, as with all the aforementioned measurements. The wet condition means that the measurement was made when the probe was fully immersed in water. Figure 15 compares the conductance signatures in dry and wet conditions for the cases of corrosion amount of 4% and 28%. For both cases, the signatures are almost overlapped for the dry and wet conditions, only with very minimal discrepancy. These results indicate that the probe shows very stable performance in the dry and wet conditions.

The influence of temperature on the performance of the corrosion-measuring probe was also investigated. The two reference probes were placed in the temperature control chamber. The temperature was changed from 10°C to 40°C with a step of 10°C. The conductance signatures of the reference probe under each temperature were acquired. Figure 16 shows the conductance signatures under different temperatures. The zoom-in view of signatures in the range 95–100 kHz is also presented. Figure 17 shows the quantifying metrics for reference probe under different temperatures. It can be observed that temperature has significant effect on the conductance signature. With the increase in temperature, the magnitude of the large peak is reduced and its peak frequency is shifted leftward. The large values in the quantifying metrics further testify the significant influence of temperature. Therefore, temperature compensation strategies need to be employed when the probe is used under varying temperature conditions.

5. Conclusion

In this study, an EMI-instrumented corrosion-measuring probe is proposed. The probe consists of a metal rod and a PZT patch. Such configuration isolates the influences of the complexity of host structure, the variation in the loading conditions and boundary conditions. Five probes were fabricated in laboratory and three of them were subjected to accelerated corrosion tests to simulate corrosion-induced mass loss damage. Results showed that with the increase in corrosion amount, the peak magnitude of the conductance signatures is reduced. The overall curve of the conductance signatures is flattened and smoothed. To quantify the variations in the conductance signatures under different corrosion amounts, statistical quantifying metrics, RMSD, MAPD, and CCD were adopted. All these metrics increase with the increase in corrosion amount, which can be used as an indicator of the corrosion process. Among the three metrics, the MAPD shows the highest sensitivity. The proposed corrosion-measuring probe is of low cost, high sensitivity, and capable of online and remote corrosion monitoring, demonstrating great application potential. This study serves as a proof-of-concept to demonstrate the capability of the EMI technique in monitoring mass loss due to corrosion. Future works will focus on the encapsulation and parametric optimization. The real-world performance of the probe will also be studied.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The research reported in this article was partially supported by the National Natural Science Foundation of China (nos: 51808170, 51678200, and 51678205) and China Postdoctoral Science Foundation (no.: 2018M630362).

ORCID iDs

Tiejun Liu

Jianjun Wang

References

Annamdas

VGM

Soh

(2007) Three-dimensional electromechanical impedance model. I: formulation of directional sum impedance. Journal of Aerospace Engineering 20: 53–62.

Bhalla

Soh

(2004) Structural health monitoring by piezo-impedance transducers. I: modeling. Journal of Aerospace Engineering 17: 154–165.

Dai

Wang

Zhang

et al. (2019) Corrosion monitoring method of porous aluminum alloy plate hole edges based on piezoelectric sensors. Sensors 19: 1106.

Zou

Liu

et al. (2018) A study on the influence of stage load on health monitoring of axial concrete members using piezoelectric based smart aggregate. Applied Sciences 8: 423.

Fan

Kong

et al. (2018a) Monitoring of pin connection loosening using eletromechanical impedance: numerical simulation with experimental verification. Journal of Intelligent Material Systems and Structures 29: 1964–1973.

Fan

Zhao

et al. (2018b) Damage evaluation of concrete column under impact load using a piezoelectric-based EMI technique. Sensors 18: 1591.

Fan

Hao

(2016) Piezoelectric impedance based damage detection in truss bridges based on time frequency ARMA model. Smart Structures and Systems 18: 501–523.

Fan

Hao

(2018c) Impedance resonant frequency sensitivity based structural damage identification with sparse regularization: experimental studies. Smart Materials and Structures 28: 015003.

Fan

Hao

et al. (2018d) Identification of minor structural damage based on electromechanical impedance sensitivity and sparse regularization. Journal of Aerospace Engineering 31: 04018061.

10.

Giurgiutiu

(2007) Structural Health Monitoring: With Piezoelectric Wafer Active Sensors. Cambridge, MA: Academic Press.

11.

Huynh

T-C

Dang

N-L

Kim

J-T

(2018) Preload monitoring in bolted connection using piezoelectric-based smart interface. Sensors 18: 2766.

12.

Kaur

Bhalla

Maddu

(2019) Damage and retrofitting monitoring in reinforced concrete structures along with long-term strength and fatigue monitoring using embedded lead zirconate titanate patches. Journal of Intelligent Material Systems and Structures 30: 100–115.

13.

Kawasaki

Wakuda

Kobarai

et al. (2013) Corrosion mechanisms in reinforced concrete by acoustic emission. Construction and Building Materials 48: 1240–1247.

14.

Koch

Varney

Thompson

et al. (2016) International Measures of Prevention, Application, and Economics of Corrosion Technologies Study. NACE International. Available at: http://impact.nace.org/documents/Nace-International-Report.pdf

15.

Kong

Hou

et al. (2013) Very early age concrete hydration characterization monitoring using piezoceramic based smart aggregates. Smart Materials and Structures 22: 085025.

16.

Lee

Park

Bolander

et al. (2018) Monitoring the hardening process of ultra high performance concrete using decomposed modes of guided waves. Construction and Building Materials 163: 267–276.

17.

Fan

SCM

et al. (2018) Interfacial debonding detection in fiber-reinforced polymer rebar–reinforced concrete using electromechanical impedance technique. Structural Health Monitoring 17: 461–471.

18.

SCM

Song

(2016a) Corrosion detection of steel reinforced concrete using combined carbon fiber and fiber Bragg grating active thermal probe. Smart Materials and Structures 25: 045017.

19.

SCM

Patil

et al. (2017a) Acoustic emission monitoring and finite element analysis of debonding in fiber-reinforced polymer rebar reinforced concrete. Structural Health Monitoring 16: 674–681.

20.

Kong

SCM

et al. (2016b) Feasibility study of using smart aggregates as embedded acoustic emission sensors for health monitoring of concrete structures. Smart Materials and Structures 25: 115031.

21.

Liu

Wang

et al. (2019a) Finite-element analysis of an electromechanical impedance–based corrosion sensor with experimental verification. Journal of Aerospace Engineering 32: 04019012.

22.

Liu

Zou

et al. (2019b) PZT based smart corrosion coupon using electromechanical impedance. Mechanical Systems and Signal Processing 129: 455–469.

23.

SCM

et al. (2017b) Monitoring concrete deterioration due to reinforcement corrosion by integrating acoustic emission and FBG strain measurements. Sensors 17: 657.

24.

Liang

Sun

Rogers

(1994) An impedance method for dynamic analysis of active material systems. Journal of Vibration and Acoustics 116: 120–128.

25.

Liang

Parvasi

et al. (2016) Bond-slip detection of concrete-encased composite structure using electromechanical impedance technique. Smart Materials and Structures 25: 095003.

26.

Lim

Soh

(2014) Electromechanical impedance (EMI)-based incipient crack monitoring and critical crack identification of beam structures. Research in Nondestructive Evaluation 25: 82–98.

27.

Lim

Tang

Smith

(2018) Piezoelectric-based monitoring of the curing of structural adhesives: a novel experimental study. Smart Materials and Structures 28: 015016.

28.

Liu

Huang

Zou

et al. (2013) Exploratory study on water seepage monitoring of concrete structures using piezoceramic based smart aggregates. Smart Materials and Structures 22: 065002.

29.

Lim

Soh

(2018a) Investigating the performance of “Smart Probe” based indirect EMI technique for strength development monitoring of cementitious materials–modelling and parametric study. Construction and Building Materials 172: 134–152.

30.

Lim

Soh

(2018b) A novel electromechanical impedance–based model for strength development monitoring of cementitious materials. Structural Health Monitoring 17: 902–918.

31.

Luo

et al. (2019) A recent progress of steel bar corrosion diagnostic techniques in RC structures. Sensors 19: 34.

32.

Luo

Hei

et al. (2016) Concrete infill monitoring in concrete-filled FRP tubes using a PZT-based ultrasonic Time-of-Flight method. Sensors 16: 2083.

33.

Moustafa

Niri

Farhidzadeh

et al. (2014) Corrosion monitoring of post-tensioned concrete structures using fractal analysis of guided ultrasonic waves. Structural Control and Health Monitoring 21: 438–448.

34.

(2017) Possibility of detecting wall thickness loss using a PZT based structural health monitoring method for metal based pipeline facilities. NDT & E International 88: 42–50.

35.

Nagy

Simonetti

Instanes

(2014) Corrosion and erosion monitoring in plates and pipes using constant group velocity Lamb wave inspection. Ultrasonics 54: 1832–1841.

36.

Negi

Chakraborty

Kaur

et al. (2018) Investigations on effectiveness of embedded PZT patches at varying orientations for monitoring concrete hydration using EMI technique. Construction and Building Materials 169: 489–498.

37.

Park

Cudney

Inman

(2000) An integrated health monitoring technique using structural impedance sensors. Journal of Intelligent Material Systems and Structures 11: 448–455.

38.

Park

Cudney

Inman

(2001) Feasibility of using impedance-based damage assessment for pipeline structures. Earthquake Engineering & Structural Dynamics 30: 1463–1474.

39.

Park

Kim

J-W

Lee

et al. (2011) Impedance-based wireless debonding condition monitoring of CFRP laminated concrete structures. NDT & E International 44: 232–238.

40.

Park

Lee

J-J

Yun

C-B

et al. (2008) Electromechanical impedance-based wireless structural health monitoring using PCA-data compression and k-means clustering algorithms. Journal of Intelligent Material Systems and Structures 19: 509–520.

41.

Ryu

J-Y

Huynh

T-C

Kim

J-T

(2019) Tension force estimation in axially loaded members using wearable piezoelectric interface technique. Sensors 19: 47.

42.

Sharma

Mukherjee

(2015) Ultrasonic guided waves for monitoring corrosion in submerged plates. Structural Control and Health Monitoring 22: 19–35.

43.

Shi

Luo

et al. (2018) Grout compactness monitoring of concrete-filled fiber-reinforced polymer tube using electromechanical impedance. Smart Materials and Structures 27: 055008.

44.

Shin

Qureshi

Lee

J-Y

et al. (2008) Piezoelectric sensor based nondestructive active monitoring of strength gain in concrete. Smart Materials and Structures 17: 055002.

45.

Song

Wang

et al. (2017) A review of rock bolt monitoring using smart sensors. Sensors 17: 776.

46.

Talakokula

Bhalla

Gupta

(2014) Corrosion assessment of reinforced concrete structures based on equivalent structural parameters using electromechanical impedance technique. Journal of Intelligent Material Systems and Structures 25: 484–500.

47.

Tawie

Lee

(2010) Monitoring the strength development in concrete by EMI sensing technique. Construction and Building Materials 24: 1746–1753.

48.

Tinoco

Cardona

Peña

et al. (2019) Evaluation of a piezo-actuated sensor for monitoring elastic variations of its support with impedance-based measurements. Sensors 19: 184.

49.

Wang

Huo

Chen

et al. (2017) Impedance-based pre-stress monitoring of rock bolts using a piezoceramic-based smart washer—a feasibility study. Sensors 17: 250.

50.

Wang

Zhu

(2011) Monitoring of the strength gain of concrete using embedded PZT impedance transducer. Construction and Building Materials 25: 3703–3708.

51.

Wang

Wei

Shao

et al. (2018) Structural stress monitoring based on piezoelectric impedance frequency shift. Journal of Aerospace Engineering 31: 04018092.

52.

Kong

et al. (2017) Interlayer slide detection using piezoceramic smart aggregates based on active sensing approach. IEEE Sensors Journal 17: 6160–6166.

53.

Feng

(2018) Electromechanical impedance (EMI) based interlayer slide detection using piezoceramic smart aggregates—a feasibility study. Sensors 18: 3524.

54.

Yang

Lim

Soh

(2008) Practical issues related to the application of the electromechanical impedance technique in the structural health monitoring of civil structures: I. Experiment. Smart Materials and Structures 17: 035008.

55.

Zhao

Gong

Qiao

et al. (2011) Brillouin corrosion expansion sensors for steel reinforced concrete structures using a fiber optic coil winding method. Sensors 11: 10798–10819.

56.

Zhu

Luo

et al. (2016) Mechanical impedance-based technique for steel structural corrosion damage detection. Measurement 88: 353–359.