**1. Introduction**

### *1.1. Nondestructive Methods of Testing Concrete Structures*

For over a century, reinforced concrete (RC) has been a dominant construction material for structures of every type and size. Usually, buildings of this kind are designed for 50–100 years of operating time. However, the remaining lifetime of a specific structure is challenging to estimate because many different factors have influences. Many structures built at the beginning of the twentieth century are still in service [1,2]. Therefore, in most countries, periodic inspections of old structures are required by a building code (usually once per five years). Even new construction acceptance tests are conducted to determine if the requirements of a specification or contract are met. The requirements may involve verification of the class, diameter, and arrangemen<sup>t</sup> of the rebars in the concrete.

Reinforced concrete could be tested in many different ways. The methods range from destructive, through semi-destructive (where the concrete is partially damaged), to utterly nondestructive testing (NDT). The NDT methods are usually cheaper and faster than methods of other groups. Unlike the destructive and semi-destructive, they can also be easily used in many points of the tested object. Therefore, they better reflect the actual state of the facility.

A full review of NDT methods used in construction diagnostics, along with their advantages and disadvantages, is given in [3]. The properties of a reinforced concrete structure which can be examined with NDT methods are presented in Figure 1.

**Citation:** Frankowski, P.K.; Chady, T. Impact of Magnetization on the Evaluation of Reinforced Concrete Structures Using DC Magnetic Methods. *Materials* **2022**, *15*, 857. https://doi.org/10.3390/ ma15030857

Academic Editor: Giovanni Bruno

Received: 3 December 2021 Accepted: 18 January 2022 Published: 23 January 2022

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

**Figure 1.** Properties of reinforced concrete structures that can be examined by NDT methods.

As described in [3] and presented in Figure 1, most of the NDT methods used in civil engineering are designed to evaluate concrete. Only methods that use an electromagnetic and mechanical wave can be effectively used for direct reinforcement assessment. The following methods can be distinguished in the mechanical group: high-frequency, active ultrasonic testing methods [3–5]; low-frequency-active mechanical methods [3,6], and passive-acoustic emission (AE) [7].

Electromagnetic methods are not universal, but on the other hand, they have many advantages over mechanical methods. The most crucial difference is that the results of the mechanical methods are affected by many factors because various phenomena may disturb the propagation of mechanical waves in complex structures. Therefore, electromagnetic and magnetic methods are preferred to assess reinforcement elements in concrete structures.

The electromagnetic methods may be used to localize rebars in the structure, precisely estimate basic structure parameters (such as the thickness of the concrete cover, the rebar's diameter, the rebars class [8–10], and detect corrosion or other flaws [11–13]). The most significant advantages of the methods from this group are the direct impact on reinforcement, the low damping of electromagnetic waves by concrete and the high spectrum of frequencies that can be used.

NDT electromagnetic methods can be categorized by the utilized excitation frequency (Figure 2). This frequency is crucial for all methods that use mechanical or electromagnetic waves. It affects resolution and an effective range. The same method may have good resolution and limited range (high frequency) or good effective range and low resolution (low frequency). In simplification, it can be assumed that the smallest size of the defect that can be detected is approximately comparable to the excitation frequency wavelength [14]. The penetration range depends on the frequency of excitation and magnetic permeability of concrete and steel. The fundamental division of NDT methods due to the frequency of excitation is shown in Figure 2.

**Figure 2.** Classification of electromagnetic NDT methods according to the excitation frequency.

The most important AC magnetic field NDT method used in civil engineering is the eddy current (EC) method. In this method, the typical excitation frequency range is from 0.5 to 10 kHz (for reinforced concrete structures). The eddy current method can be used not only to detect the presence of rebars but also to determine the thickness of the concrete cover, the rebar's diameter, the alloy of reinforcing bars (due to different electrical properties), or even to detect corrosion of rebars [8–13]. The effective range of the eddy current method is from 0 to about 100 mm. Results can be really accurate and relatively easy to interpret. Lower excitation frequencies may be used in some versions of the magnetic flux leakage (MFL) and the magnetic force induced vibration evaluation MFIVE method [3] or in the method similar to MFIVE described in [6]. Both of these methods use low-frequency magnetic waves to induce rebar vibration. Natural frequencies of the reinforcement can be used to detect structure debonding, which is usually caused by corrosion.

Another important electromagnetic method is ground-penetrating radar (GPR). The standard operating frequency ranges from 100 MHz to 3 GHz. Rebars can be detected from the distance of several centimeters up to ten meters or more (when other electromagnetic methods have the maximum detection range not bigger than 200 mm). However, results are difficult to interpret and not very accurate [15,16]. The terahertz technique is rarely used due to the limited penetration in concrete, which is usually characterized by high water content, strongly damping electromagnetic waves at these frequencies. Higher frequencies are used in radiography, which can be very effective but, on the other hand, possess many limitations. The source and detector usually must be placed on both sides of the object. Moreover, this method generates risks for human health [3].

Inspection methods utilizing DC magnetic field can be divided into two categories: continuous magnetization techniques (CMT), also called active magnetic inspection (AMI) and residual magnetization techniques (RMT), called passive magnetic inspection (PMI). In the case of CMT, not only receiving devices but also excitation is required.

The leading representative of CMT is the magnetic flux leakage method (MFL). The method is commonly used in the inspection of ferromagnetic parts and components. However, currently, the adaptations of this method for civil engineering are also popular.

In the MFL method, the detector is usually placed between the poles of the magnets or electromagnet) to detect the leakage field. The relative permeability of concrete, stones, water, and the air is close to 1. Therefore they have practically no influence on the magnetic field distribution. The reinforced bars (rebars) made of steel as ferromagnetic materials concentrate the magnetic flux. In this way, the magnetic field is influenced by rebars and can be used to localize them in the concrete structure. The magnetic flux can be disturbed by discontinuities in the material, such as breaks or cracks [12]. The magnetic flux leakage caused by rebar inhomogeneity can be detected at a distance in the range of the typical concrete cover [17,18].

In some cases, the MFL method can be used to determine the material loss caused by corrosion [19–22]. Magnetic methods also allow to identify rebar diameter [23]. The magnetic flux leakage method can also be used for structural health monitoring [22]. Other active magnetic methods, such as Barkhausen emission (MBE), magnetoacoustic emission (MAE), stress-induced magnetic anisotropy (SMA), or magnetic powder method, usually are not used for the evaluation of reinforced concrete structures. The magnetic field is higher in the case of the active magnetic methods (CMT). However, the CMT methods also have disadvantages like longer measurement time, equipment deployment, and power consumption [3].

Residual magnetization methods are more economical and straightforward. The basic RMT is the magnetic memory method (MMM). The method can be used to detect abnormal conditions arising from changes in crystalline structures resulting from stress concentration, corrosion, or cracks. One of many versions of MMM is iCAMM (infrastructure corrosion assessment magnetic method). This method works through passive magnetic inspection under the effect of the Earth's magnetic field.

### *1.2. Novelty and Significance of the Research*

Periodic evaluation of reinforced concrete structures is required by national law in most countries. However, in many cases, such inspection can be problematic. Standard 'in point' tests can be misleading (most of the structure is not checked). The point-to-point scans also cannot be used in large areas because tests of this kind are usually very timeconsuming. The obvious solution is to use area tests. In such a way, the investigation time is significantly reduced and received results are reliable. However, currently, there is not even one method that can be used in that way on a large scale. Area tests potentially can also be used as a pilot or preliminary evaluation before applying other more precise methods. There are only a few methods that theoretically can be used for such evaluation. This group includes primarily visual testing, radiography, and thermography. Unfortunately, these methods have many limitations (e.g., thermography can be used only if the concrete cover is low [8,24]; radiography requires specialized equipment, generates risks for human health, and elements of the system must be placed on both sides of the object) and they are often insufficient. The full summary of the area testing methods is shown in [3]. The magnetic methods are not always considered to be good for area tests. However, this method possesses many advantages over others tests mentioned before. Tests executed with the magnetic method are cheap, the principle of operation is easy to understand and use, the used magnetic wave can avoid damping caused by concrete cover. The test showed that the magnetization method is crucial for the effectiveness of this method. The potential of the active and passive magnetic methods is presented in further sections of the paper.

### *1.3. The Article Outline*

In the introduction of this paper, first, the importance of nondestructive testing (NDT) in periodic tests of reinforced concrete structures has been described. A brief overview of the NDT methods used in the construction sector is also presented. Next, the significance of the conducted research was indicated.

The Section 2 (Materials and methods) presents the tested samples and measurement systems. The section has much attention to magneto-optical (MO) sensors. The MO elements, one of the few magnetic field detectors, are designed for area testing. The evaluation of ferromagnetic objects remote from the sensor as much as in the case of reinforced concrete is an unusual issue for this type of sensor, which is intended and designed for surface testing. Therefore, before the tests, their accuracy in the case of reinforced concrete structure was doubtful. For more detailed investigations, AMR sensors connected in one matrix were proposed. In this section, examples of received results and algorithms of data processing are discussed.

The results of the measurements were placed in the Section 3. First, the entire section is briefly described. Next, results received for the MO sensor are presented. The experiments with the MO sensor show both the influence of magnetization on increasing the ability to detect rebars and the application potential of the MO-sensors.

In the other subsection, results received for three different samples and three different magnetization variants are presented. All experiments were conducted with the AMR sensor. The main point of the subsection is to show how significant the impact of the magnetization method on received results can be. The impact is even stronger for more complex samples. This part also presents the disadvantages of the passive method, which also becomes more significant during the tests on more complex samples.

The obtained results are summarized in the Section 4 In particular, the magnetization aspect is discussed in this part. The section 'Conclusions' discussed whether the magnetic method is finally suited for area testing and how the tests of this kind fall on the background of other methods. The two tested sensors are also compared in this part. The advantages and disadvantages of both systems are presented, and applications of the sensors have been proposed. In the section also plans for further research on the magnetic method for area testing are presented.

### **2. Materials and Methods**

### *2.1. Measuring Systems and Samples*

2.1.1. Test Samples

The main aim of the article is to investigate the influence of magnetization on the effectiveness of magnetic nondestructive testing methods in the evaluation of reinforced concrete structures. For this purpose, the three different samples are examined: S1—the sample with single rebar (Figure 3a); S2—the sample with two rebars, one placed 85 mm under the other (Figure 3b), and S3—the sample contains three rebars, all rebars placed one next to each other (Figure 3c). In the third sample, distances between rebars are 55 and 50 mm. The magnetic sensor was moved above the sample in a line perpendicular to the reinforcement bars. The distance between the rebar and the sensor (thickness of the concrete cover) is marked as *h* (Figure 3a). The results were obtained using an integrated AMR transducer that allows measuring three field components.

Configurations of the samples are presented in Figure 3. The magnetic transducer was moved along the *x*-axis, while rebars were positioned along the *y*-axis.

### 2.1.2. Systems for Active Magnetic Inspection

The measuring system consisted of four subsystems: excitation subsystem, positioning subsystem, magnetic field transducer, and data acquisition subsystem. The general block scheme of the system is presented in Figure 4. All subsystems are described in the following sections.

The simplest solution to magnetize reinforcement bars (rebars) can be achieved using permanent magnets. In the presented systems, two neodymium magnets in two different configurations were used for this purpose. The reference configuration was without any magnets, as shown in Figure 5a. In the second configuration, magnets have opposite poles facing the sample (Figure 5b). In the third configuration, the magnets were directed to the sample with the same poles (Figure 5c). The magnets were placed on both sides of the sensor at a distance of 500 mm.

**Figure 3.** The samples used in the experiments; (**a**) sample S1 (with single rebar); (**b**) sample S2 (with two rebars one under the other); (**c**) sample S3 (with three rebars next to each other).

**Figure 4.** Block scheme of the measuring system.

In the experiments, a two-dimensional area over the sample surface was scanned. The area directly above the rebars is tested with a positioning system. The example of positioning subsystem is shown in Figure 6.

**Figure 5.** System configurations used in the experiments; (**a**) reference configuration; (**b**) configuration of opposite poles magnetization (OPM); (**c**) configuration of same poles magnetization (SPM).

**Figure 6.** Photo of the MO transducer attached to the XYZ scanner: 1—camera; 2—the source of monochromatic light; 3 and 4—linear polarizers; 5—XYZ scanner.

The magnetic field sensor is an essential part of the system. Magnetoresistive (MR) and Hall effect sensors are of the greatest industrial importance among the magnetic field sensors. The Hall effect components account for approximately 85% of the world's production of magnetic sensors for DC and low-frequency applications. The MR sensors account for around 10%, and their market share grows [25].

The most used MR sensors are anisotropic magnetoresistors (AMR) and giant magnetoresistive effects (GMR) elements. The AMR and GMR sensors have high sensitivity and field resolution. Elements of this kind can operate even in the pT range. However, they can be permanently affected by strong magnetic fields and GMR sensors have a high hysteresis.

The Hall effect sensors have several advantages over MR elements. They show no saturation effects and can measure strong magnetic fields. For these reasons, the Hall effect sensors are preferably used at magnetic fields higher than 1 mT. They are the first choice in many industrial applications. However, large offset and relatively low sensitivity limit both the accuracy of the measurements and the minimum value of the magnetic field that can be measured. One of the issues examined in this research is testing non-magnetized reinforced concrete structures using magnetic methods. The MR sensors seem to be much better suitable for this purpose.

Most magnetic field sensors can measure the magnetic field at one point. The exception is magneto-optical (MO) sensors, which are well suited to constructing an area

testing system. Therefore, magneto-optical (MO) sensors are preferable for testing largescale reinforced concrete structures. The Faraday magneto-optical effect is used in MO sensors [26,27]. The main advantage of this solution is the immediate obtaining of the 2D field distribution over the sample surface.

### 2.1.3. Measuring System with Magneto-Optical Sensor

The Faraday magneto-optical effect is used in MO sensors. This effect describes an interaction between light and a magnetic field in a medium. The plane of polarization of linearly polarized light rotates parallel to the propagation direction of light waves passing through the magneto-optical medium. The mechanism of the Faraday effect is explained in Figure 7 [26,27].

**Figure 7.** Operating principle of the MOIF. MOIF—Magneto-Optical Indicator Film (sensor); MV— Magnetic Vector; IL—Polarization Plane of Incident Light; RL—Polarization Plane of Reflected Light; *Φ*—Angle of Faraday rotation.

The MO sensor presented in Figure 8a consists of four layers, as shown in Figure 8b. Additional layers are necessary to improve the quality of the measurements. The mirror layer (for visible spectral range) is used to improve the sensor reflectivity. For mirror protection, the resistant material layer is used. The sensor also contains anti-reflection coated glass [26,27].

The most important advantage of the MO-sensor over other magnetic field sensors is the large area of observation of the magnetic field and the relatively high resolution. The most significant advantage of MO-sensor over other magnetic field sensors is the large area of magnetic field observation and relatively high resolution. The manufacturers offer sensors with diameters up to 3 inches. A few different types of MO transducers are used in many different applications [26,28]. Parameters and characteristics of the type A sensor used in the experiment are provided in Figure 9. The sensitivity of this MO-sensor is comparable to the Hall effect elements.

The type A sensor used in the experiments is an out-of-plane (OOP). The MO-sensors of this kind are generally more sensitive but have a smaller range and nonlinear characteristics. The hysteresis (Figure 9a) can also cause difficulties during measurements (in the case of less sensitive MO-sensors, there are no such problems). The A-type is chosen because of the lowest dynamic range (significant visible changes with minor magnetic field changes). An alternative to the A-type transducer in this kind of application is a D-type transducer. Sensors of this kind are more sensitive than A-type; field range is from 0.03 to 5 kA/m and can be used to test printed magnetic inks or steels alloys. The sensors are sensitive, but it also depends on the quality of the camera and other elements. There is another valuable property of the D-type element.

The D-type element can be working in two modes:

• Faraday: for applications without external excitation; • Bias: for work in the environment of an external magnetic field. In this mode, performance is weaker, but other types of sensors would lose their performance entirely. This mode is used mainly with magnetically very soft materials, like inks.

The MO-transducers require a relatively complex setup. The block diagram of the system with MO-transducer is shown in Figure 10, and the setup photo is presented in Figure 6.

**Figure 8.** Magneto-optical sensor; (**a**) the photo of the A-type MO sensor in the protective packaging. (**b**) schematic showing the functional layers of the Magneto-Optical Indicator Film (sensor).

### 2.1.4. Measuring System with a Magnetoresistive Sensor

Systems based on MR sensors are less complex than these based on MO sensors. Moreover, the AMR sensors with three sensitivity axes are better suited for more accurate investigations of reinforced concrete structures.

AMR (anisotropic magneto resistance) elements belong to the MR group of sensors. The resistance of these elements decreases when a magnetic field is applied. This function is dependent on the direction of the magnetic force lines applied to the element (anisotropic). The material of the AMR element is an alloy of nickel, iron, and other metals (ferromagnetic). In these experiments, integrated transducer HMC5883L was used. The sensor has few advantages over GMR. The sensitivity is high, much higher than in the case of the MO sensor. Nevertheless, lower than it could be in the case of GMR [29].

**Figure 9.** *Cont*.

**Figure 9.** Parameters and approximate curves of characteristics-utilized MO-sensor; (**a**) A-type sensor: plot of magnetic field vs. Faraday rotation *Φ* (λ = 590 nm), and selected parameters of the sensor; (**b**) D-type Faraday version of the sensor: plot of magnetic field vs. Faraday rotation *Φ* (λ = 590 nm); (**c**) D-type bias version of the sensor: plot of magnetic field vs. Faraday rotation *Φ* (λ = 590 nm). (Based on materials received from the manufacturer Matesy).

**Figure 10.** Block diagram of the system with the MO-sensor.

On the other hand, the sensitivity of GMR would be too high for this application. With the use of 'reset strap drive' the internal offset of the sensor and its temperature dependence is corrected for all measurements. This option could be helpful in the vicinity of large magnetic fields. In opposite to GMR, AMR sensors clearly indicate the results of the magnetic field direction. Because the positive and negative sides have symmetric characteristics, the same operation is performed even if the north and south poles of the magne<sup>t</sup> are reversed. This characteristic is used to improve the reliability and accuracy of the data. The sensor also has high linearity and low hysteresis.

### *2.2. Methods of Processing the Results*

2.2.1. Measurement Results Processing in the System with MO Sensor

The results obtained from MO systems usually do not require complicated processing and are available in real-time. Nevertheless, in some cases, such as a high thickness of concrete cover *h*, even minor image changes have to be detected. Therefore the following algorithm of the image enhancement was implemented. First, the algorithm extracts the active area of the MO sensor from the image obtained from the camera. Then, since the axis of the camera lens was not perpendicular to the sensor surface, it was necessary to correct the perspective. The next step is to reduce the geometric distortions caused by the lens. Images processed in this way are saved in the system memory. Due to the relatively small size of the sensor area, the final image [A] consisted of several (5 to 7) images [An] taken at subsequent positions above the sample. The sensitivity of the MO transducer is not the same at different places on the sensor surface. Therefore, the images [Ai] are corrected using a coefficients matrix calculated from a uniform DC magnetic field measurement. In the cases of small (0–20 mm) or big (80–100 mm) thickness of concrete cover (*h*), it is also necessary to correct the non-linearity of the characteristic and hysteresis presented in Figure 8a. In order to remove noises, a 2D-median filter with a 5 × 5 mask is applied to the image [A]. In the last step, contrast and brightness were corrected. Effects of the processing are shown in Figure 11.

**Figure 11.** The processing of images obtained with the MO sensor; same pole magnetization (SPM); Sample S1; *h* = 0.5 mm.

The MO sensors enable testing areas of objects under investigation without timeconsuming point-by-point scans. Unfortunately, sensitivity, linearity, and repeatability are limited. Moreover, the images are noisy. The problems only to some degree, can be caused by hardware limitations (polarizers or video cameras). The MO sensors could be a solution for a preliminary evaluation.

2.2.2. Measurement Results Processing in the System with MR Sensor

MR systems are much more sensitive than MO systems. Moreover, systems of this kind can deliver information about three components of the magnetic field. In further investigation, measurements were taken by moving the transducer with a 1 mm step in the *x*-axis and 10 mm in the *y*-axis direction. The measurements were very time-consuming.

Examples of results received using opposite poles polarization for inspection of the sample S1 are presented in Figure 12.

**Figure 12.** Magnetic field *Bx*, *By*, and *Bz* components measured in case of the same pole polarization and sample S1 with single rebar. Measurements were carried out with the AMR sensor; concrete cover thickness *h* = 20 mm.
