Advanced Structural Monitoring Technologies in Assessing the Performance of Retrofitted Reinforced Concrete Elements

Abstract

Climate change induces extreme effects with lower-than-designed restoration periods, imposing the necessity of strengthening the structural integrity of existing and mainly older RC structures, which are often demonstrated to be under-reinforced in terms of the shear capacity, mainly due to outdated and old design codes/standards. Thus, finding cost-effective and feasible methods to strengthen RC elements is becoming increasingly important. Thin RC layers for jacketing represent a modern advancement in repairing and retrofitting RC members. In this context, U-shaped mortar jackets were employed to strengthen three shear-critical beams. In addition, a critical aspect in the success of any jacketing method is the degree of bonding and interaction between the original member and the new jacket. Additionally, the performance of these U-shaped jackets was assessed using an Electro-Mechanical-Impedance-based (EMI-based) method using a Piezoelectric-Transducer-enabled (PZT-enabled) technique. The integration of advanced monitoring technologies in retrofitting applications offers valuable insights into the performance and longevity of the retrofit system. Therefore, this study aims to experimentally investigate the cohesion between construction materials and assess the effectiveness of U-shaped jackets. Through the proposed Structural Health Monitoring (SHM) technique, any degradation at the interface or slippage of the retrofitting jacket can be promptly detected, restraining further damage development and potential failure of the structure.

1. Introduction

Various factors, such as seismic excitations, corrosion of the steel reinforcement, fire exposure, and fatigue, have been identified as the leading causes of severe structural degradation. Moreover, deterioration and damage to critical Reinforced Concrete (RC) members can adversely affect the structures’ ability to withstand future operating conditions, escalating the risk of catastrophic collapses, casualties, and potential economic loss if not mitigated effectively [1]. In addition, RC members in existing structures are often demonstrated to be under-reinforced in terms of shear capacity, mainly due to outdated and old design codes/standards [2]. Furthermore, climate change induces extreme effects with lower-than-designed restoration periods, imposing the necessity of strengthening the structural integrity of existing and mainly older RC structures [3].

In addition, as urban populations continue to grow, there is an increasing demand for renovating and strengthening existing buildings. One of the significant challenges in urban environments is effectively utilizing the existing infrastructure by combining structural and energy retrofitting upgrades, particularly in regions prone to seismic activity or coastal areas that suffer from corrosion [4]. Thus, the approach mentioned above protects previous investments in energy efficiency and contributes to sustainable urban development aligned with the concept of smart, resilient, and sustainable cities [5]. Reinforcing RC elements can be gleaned to the below primary reasons: achieve higher load demands, changes in the structural system, design flaws, construction errors, accidental damage, or adjustments due to reduced steel reinforcement stress and deflection limits [6].

Nowadays, it is becoming increasingly important to find cost-effective and feasible methods to strengthen RC elements [7]. In particular, RC beams are critical components of the load-bearing capacity. Hence, efficient strengthening techniques are needed to ensure the performance level of the structures [8]. RC jacketing is a well-established method for reinforcing deficient RC elements [9,10]. Numerous studies have demonstrated that traditional RC jacketing significantly improves the load-bearing capacity and stiffness, enhancing overall structural performance.

Ghalehnovi et al. used a steel-fiber-reinforced concrete jacket to achieve enhanced flexural strength in recycled aggregate concrete beams [11]. Hassan et al. retrofitted shear-damaged RC T-Beams using a U-Shaped Jacket [12], and Hung et al. implemented Ultra-High-Performance Concrete (UHPC) and high-strength steel mesh reinforcement to rehabilitate seismically damaged RC beam-column joints [13]. Further, Alhadid et al. achieved an analysis of the improved properties of RC beams strengthened using concrete jackets [14], and Mahmoud et al. comparatively investigated the behaviors of different corroded RC members strengthened by different types of concrete jackets [15].

However, traditional RC jacketing has some drawbacks, including altering the structural dynamic properties due to the increased mass and stiffness and requiring labor-intensive installation processes. The latter has prompted researchers to explore alternative jacketing materials. Thus, Liu et al. presented a review regarding the use of bolted side plates for the shear strengthening of RC beams [16]. Further, Christidis et al. used different steel configurations to strengthen non-conforming RC shear walls [17]. Moreover Mossalam et al. strengthened of two-way concrete slabs with FRP composite laminates [18], Biskinis et al. investigated the shear resistance of RC members with a closed FRP jacket [19], and Lu et al. examined the shear transferring mechanism of the FPR-to-concrete bonded joint with end U-Jacketing [20]. In addition, Tsonos investigated the effectiveness of Carbon-FRP (C-FRP) jackets and RC jackets in the post-earthquake and pre-earthquake retrofitting of Beam–Column Subassemblages [21]. Furthermore, Qureshi et al. examined the shear behavior of concrete beams retrofitted with anchored C-FRP wraps at elevated temperatures [22]. Further, Chalioris et al. introduced a new strengthening technique for RC deep beams using C-FRP ropes as transverse reinforcements [23].

Additionally, alternative jacketing materials consist of various forms of high-performance [24,25,26,27,28] or Self-Compacting Concrete (SCC) [29,30]. Innovative jacketing techniques have also gained attention, utilizing textile-reinforced mortars (TRMs), where Triantafillou and Papanicolaou strengthened RC members using TRM jackets [31]. Moreover, Shen et al. examined the residual flexural behavior of RC beams strengthened with TRC [32]. Further, Alhusban and Parvin used TRM jackets under gravity and cyclic loadings for the local strengthening of RC frames [33]. Furthermore, Guo et al. experimentally investigated the implementation of TRM in shear-strengthening pre-damaged RC beams [34].

In addition, steel fibrous is also among the proposed materials. Hence, Martinola et al. employed fiber RC to strengthen and repair RC beams [35]. Further, Achillopoulou and Karabinis used fiber-reinforced mortar to repair RC columns [36]. Moreover, Katakalos et al. used steel fiber polymers for seismic retrofitting of RC T-Beams subjected to cyclic loading conditions [37]. Additionally, Khalifa et al. used eco-friendly steel-fiber-reinforced geopolymer concrete as a flexural repair technique of pre-damaged T-Beams [38]. Moreover, Jabr et al. investigated the effect of the fiber type and axial stiffness of the fiber-/fabric-reinforced cementitious matrix on the flexural strengthening of RC beams [39].

Despite the wealth of research on retrofitting methods, the focus has predominantly been on slightly damaged RC members, with fewer studies addressing the rehabilitation of severely damaged, shear-critical beams [2,34,40]. Additionally, in seismic-prone regions, structures must possess a high ductility and deformation capacity, leading to the search for bonding solutions more compatible with masonry and concrete substrates [41]. For this purpose, applying U-shaped mortar jackets has proven beneficial, as they enhance tensile and shear strength, improve deformability, and increase ductility, resulting in greater load-bearing capacity and improved performance in the post-failure deformation stage, enhancing the overall performance and resilience of the structures [42].

The jacketing of beams is recommended for several purposes as it gives continuity to the beam–column sub-assemblages and increases the strength and stiffness of the structure. While jacketing a beam, its flexural resistance must be carefully computed to avoid creating a strong beam-weak column system [43]. In retrofitted structures, due to the jacketing of a column, there is a strong possibility of changing the failure mode and the redistribution of forces, which may consequently cause beam hinging. Further, the location of the beam-critical section and the existing reinforcement’s participation should be considered. The jacketing of beams may be carried out in different ways; the most common are one-sided jackets or three- and four-sided jackets. On several occasions, the slab has been perforated to allow the ties to go through and enable concrete casting [44].

The use of thin RC layers for jacketing represents a modern advancement in repairing and retrofitting RC members [45,46]. This technique has been made viable thanks to the development of high-strength SCC, mortars, and grouts. Thin RC jackets combine the advantages of conventional RC jackets, such as enhanced stiffness and the protection of steel reinforcement [47], with those offered by FRP jackets, including a negligible effect on the member’s dimensions [48]. Although FRP jackets are thinner, the dimensional differences between thin RC and FRP jackets are relatively minor compared to conventional thick RC jackets, which significantly increase the member size and alter dynamic characteristics [49,50].

Various modifications of thin RC jacketing have been examined in the literature. The jacket usually comprises a very thin layer of fine-aggregate concrete, generally not exceeding 25 mm in thickness. This jacket is applied to the external surfaces of existing RC members and encases small-diameter longitudinal and transverse steel reinforcement. Using SCC, mortar, or grout simplifies the casting process. This approach contrasts with conventional RC jacketing, which often employs shotcrete application, resulting in much thicker shell layers of 50–70 mm, significantly changing the geometry of the retrofitted element [51]. A critical aspect in the success of any jacketing method is the degree of bonding and interaction between the original member and the new jacket. This interaction largely depends on frictional forces between the old and new concrete, which can be enhanced through dowels [2,52] or steel connectors [53]. Recent techniques also involve shear connectors formed by welding corner bars and bending down reinforcing bars to link the jacket and the core [54].

The efficacy of any jacketing scheme can be assessed based on its ability to alter the failure mode and improve the monolithic performance of the retrofitted element. Thus, the shear transfer mechanism between the initial RC core and the applied jacket at the interface is a key aspect [55]. Recently, several researchers have studied the bond strength between old and new concrete to enhance the connection at the interface zone, which depends on factors, such as the surface roughness, jacket design, material strength, cleanliness of the surface, deterioration of the substrate, moisture content, curing conditions, interface orientation, and the type of load applied. Concrete substrates can suffer damage from various factors, which are generally classified into two types: sudden damage (from events, like natural disasters, wars, or accidents) and progressive damage (due to misuse, neglect, or harmful environmental factors, like carbonation, sulfate and chloride attacks, or alkali-silica reactions). Some of these factors may even improve the bond strength between the old and new concrete [56].

In practice, the direction in which the new concrete is poured is crucial for strengthening beams, as the interface orientation highly influences bond strength between SCC and conventional concrete. Research has shown that top bonding surfaces offer the highest bond strength, followed by side and bottom surfaces. The success of the jacketing process largely depends on achieving a solid bond between the two materials. SCC is particularly effective for this purpose due to its excellent workability, flowability, and ability to fill spaces without additional vibration, preventing issues like agglomerates or voids in the RC element or at the interface between old and new concrete. The latter makes SCC a highly reliable material for strengthening concrete members, especially RC beams.

In this context, U-shaped mortar jackets were employed to strengthen three shear-critical beams that had primarily failed due to shear cracking under monotonic loading. These beams were retrofitted with three-sided jacketing and then reloaded to failure. The experimental results demonstrated U-shaped jackets’ significant potential to improve the shear strength and deformation capacity. Although shear failure persisted due to the beams’ predisposed design and span-to-depth ratio, the U-shaped jackets made of cement-based grout, small-diameter mild steel bars, and open stirrups provided considerable strengthening effects. Additionally, the performance of these U-shaped jackets was assessed through the implementation of Structural Health Monitoring (SHM).

In the emerging field of advancing SHM technology, considerable research has focused on damage identification, localization, and assessment of RC structures [57,58,59]. Acoustic emission testing, a well-established and widely applied non-destructive technique, is used to monitor, detect, and pinpoint flaws and damage in mechanically stressed materials and structural components [60,61,62]. Alternative localized inspection techniques, such as X-ray testing, ultrasonic waves [63,64,65], laser scanning [66,67,68], infrared thermography [69,70,71], digital imaging, and image correlation techniques [72,73,74], have been explored to assess the condition of RC members.

Additionally, strain-based approaches using fiber optic Bragg grating sensors effectively monitor changes in tensile forces within RC structures [75,76]. Despite this, vibration-based and global methods, such as geodetic methods [77] and ground-penetrating radar (GPR) [78,79,80], are often inadequate for detecting localized structural damage due to their reliance on lower-order modal parameters [81,82]. Some of the methods above offer unique advantages for monitoring the performance of retrofitted structures, and many employ a combination of sensors and algorithms for improved accuracy [1,83,84,85]. However, conventional SHM methods are expensive and often insufficient, especially for challenging-to-access structural members, highlighting the importance of cutting-edge monitoring technologies.

Among the various SHM techniques, the Electro-Mechanical Impedance (EMI) method is a localized SHM technique that has shown significant potential for detecting damage in RC structural components early, as highlighted in numerous studies [3,86,87,88,89]. Recent research has demonstrated its effectiveness in identifying structural issues, such as concrete cracking [1,90], steel reinforcement yielding [91], and force quantification [92,93]. Advanced SHM techniques using cutting-edge transducers and smart materials have demonstrated their ability to monitor structural damage in RC elements continuously [94].

Recent studies have shown that embedding a network of PZT patches in regions of an element prone to damage significantly enhances the effectiveness and accuracy of SHM methods for assessing damage levels and providing reliable diagnoses [95,96,97]. Notably, Perera et al. [98] and Sevillano et al. [99] expanded PZT-based SHM by introducing an advanced clustering technique that effectively evaluates multiple measurements from the PZT transducer network, enabling the diagnosis of damage in RC elements reinforced with FRP materials. Liu et al. also investigated debonding failures in FRP-strengthened beams using PZT patches, demonstrating the technique’s applicability for detecting such issues [100].

The EMI method primarily depends on piezoelectric materials, particularly piezoelectric lead zirconate titanate (PZT) patches, which offer several advantages for SHM applications. These include low energy consumption, cost efficiency, ease of installation, compact size, portability, and the ability to function as both sensors and actuators [63]. By leveraging the electro-mechanical coupling properties of PZTs, the EMI method detects irregularities in the structural integrity of the monitored structure [101]. The impedance signatures reflect the structural characteristics of the monitored region, so changes in the structure’s properties lead to corresponding variations in impedance responses. Local deterioration can be identified effectively by comparing impedance signals recorded before and after a damaging event.

A significant portion of the existing research has focused on evaluating the effectiveness of EMI-based monitoring for detecting various forms of concrete degradation. These degradation factors include, but are not limited to, rebar corrosion in RC members [102,103,104,105], the effects of the heating duration [106,107], and concrete mass loss [108,109]. The EMI method has also been applied to assess the compactness of tendon duct grouting [96] and the setting time of fiber-reinforced concrete [110]. It has also been extensively used to evaluate damage caused by loading in concrete specimens, with particular emphasis on flexural damage in RC and fiber-reinforced concrete beams [92,111,112,113,114], as well as compression-induced damage in standard concrete specimens [90,93,115,116,117,118]. Research has further extended to examining the behavior of full-scale RC structural members, such as beams [3,98,119] and joints [120,121,122]. Moreover, the EMI method has proven effective for the early detection of FRP debonding in RC members [98,123].

In this study, the EMI-based PZT-enabled technique is used. The integration of advanced monitoring technologies in retrofitting applications offers valuable insights into the performance and longevity of the retrofit system. Despite extensive research into bond-slip behavior between concrete and various reinforcement materials [124,125], no similar studies have evaluated the bond performance of U-shaped jackets by monitoring the interfacial zone between the original RC core and the jacket.

Therefore, this study aims to experimentally investigate the cohesion between construction materials and assess the effectiveness of U-shaped jackets composed of cement-based mortar, mild steel bars, and open stirrups for retrofitting shear-damaged RC beams. Moreover, the structural integrity of these retrofitted beams is evaluated through data analysis from the embedded and the externally epoxy-bonded PZT transducers.

Through the proposed SHM implementation, any degradation at the interface or slippage of the retrofitting jacket can be promptly detected, restraining further damage development and potential failure of the structure. This way, severe and catastrophic causalities can be prevented and avoided, leading to targeted and sufficient recovery actions in real-life structures. Moreover, the proposed technique can provide useful insight into the structural integrity of the monitored structure or infrastructure, permitting the stakeholders to make critical decisions during an event.

2. Materials and Methods

2.1. Beams’ Characteristics

The initial stage of the experimental study involved three shear-critical RC beams, each 1.2 m long, forming a rectangular cross-section. These beams were subjected to a four-point monotonic loading setup and had the following dimensions: a width of 100 mm, a height of 200 mm, and an effective depth of 175 mm, with a shear span of 500 mm, and a shear span-to-depth ratio (α/d) of 2.86, respectively. Figure 1 provides a schematic representation of the geometrical details of the initial beams. Henceforth, the initial beams are notated as “B” in subsequent references for consistency.

Further, the “B” beams were reinforced with identical deformed steel longitudinal tension and compression bars, featuring 14 mm-diameter tension bars (2Ø14 at the bottom) and 10 mm-diameter compression bars (2Ø10 at the top). The tensile yield strength of these bars was f_yl = 550 MPa. The reference beam “B500” had no transverse shear reinforcement in the shear spans, while the other two beams were equipped with mild steel closed stirrups of 5.5 mm in diameter with a yield tensile strength of f_yv = 225 MPa. The stirrups were spaced at 200 mm, measuring from the beam’s mid-point for “B200a”, and at 200 mm, measuring from the support in beam “B200b”, respectively, as shown in Figure 1. Both beams numbered two stirrups in each shear span [126]; according to the two-branch crack development of shear propagation theory, “B200a” focused on limiting the growth of the inclined cracking to the upper mid of the shear span, while “B200b” attempted to intercept the propagation on the lower part of the shear span to avoid extensive cracking development and concrete spalling near the supports.

The geometric and mechanical reinforcement ratios of the tension longitudinal bars ρ_l and ρ_l f_yl/f_c and the vertical stirrups ρ_v and ρ_v f_yv/f_c for each beam are provided in

Table 1. Geometrical and mechanical reinforcement ratios for initial beam.

Beam Codified Name	ρl	ρl fyl/fc	ρv	ρv fyv/fc	fc (MPa)	fct,spl (MPa)
B500	1.76%	0.340	-	-	28.5 (0.53)	2.60 (0.26)
B200a	1.76%	0.334	0.24%	0.018	29.0 (0.98)	2.32 (0.14)
B200b	1.76%	0.328	0.24%	0.018	29.5 (0.70)	2.40 (0.20)

. As a part of the experimental design, the beams are targetally equipped with insufficient shear reinforcement relative to their provided bending reinforcement, leading to higher flexural strength than shear. Different configurations (number) of stirrups were used in the “B” elements to ensure that a range of shear strengths and deformation performances and varying degrees of cracking could be exhibited [2].

In progress to the “B” beams’ failure, the current experimental project is focused on retrofitting and retesting the three heavily damaged beams. The retrofitting method for the shear-damaged RC beams induces thin U-shaped jackets designed to restore and potentially enhance the original shear strength and deformation capacity while improving the overall performance. The retrofit includes small-diameter (5.5 mm) mild steel reinforcement and a 25 mm-thick layer of cement-based mortar.

The retrofitting process involves applying jacket reinforcement to the surface of the existing RC beams. A few dowels were utilized primarily to enhance the connection and facilitate force transfer between the old concrete and the new cementitious grout. Before implementing the reinforcement and constructing the jacket, careful surface preparation was conducted to establish a robust bond between the original concrete and the new retrofitting layer. This preparation included the utilization of a chisel and a hammer to remove damaged and spalled concrete sections. In continuation, a sharpener and an electric concrete breaker were used to clean and rough the concrete surface. Moreover, abrasive sandblasting was utilized to remove rubble, yielding a concrete substrate that exhibited appropriate smoothness, roughness, and uniformity, as depicted in Figure 2.

Furthermore, the main purpose of the jacket was to cover the added reinforcement, which consists of straight longitudinal bars and U-shaped stirrups, each with a diameter of 5.5 mm (Ø5.5) and a yield tensile strength of 225 MPa. The three retrofitted RC beams, distinguishable from the original ones, henceforward will be referred to as “B-J”. Each of them is still 1.2 m long and has the following dimensions: width (b) = 150 mm, height (h) = 225 mm, effective depth (d) = 203 mm, and shear span (α) = 500 mm. This results in a shear span-to-depth ratio of α/d = 2.46. For further clarification, please refer to Figure 1, which illustrates the mentioned geometric and reinforcement details.

“B-J” beams were reinforced with two Ø5.5 mm-diameter bottom bars, two Ø5.5 mm side bars, and two Ø5.5 mm top bars. Further, L-shaped dowels were epoxy-bonded into 7 mm drilled holes in the vertical sides of the beams. The drilled holes were cleaned and filled with epoxy resin for successful anchorage. Additionally, U-shaped transverse reinforcement with varying spacings (60 mm, 85 and 85 mm) was used for beams labeled “B500-J”, “B200a-J”, and “B200b-J”, respectively. Finally, the jacket was cast in steel molds with cementitious self-compacting mortar (Figure 3). More details on the mechanical reinforcement and the retrofitting process can be found in the research conducted by Chalioris et al. [2].

2.2. Material Properties

A concrete matrix of cement, sand, coarse aggregate, and water in a ratio 1:1.67:3.05:0.51 was employed for the “B” beams casting process. The concrete mixture included high-fineness modulus sand, 16 mm crushed stone, and Portland-type cement. On the experiment day, standard specimens were subjected to compressive and tensile splitting strength testing using a universal testing machine with a maximum capacity of 3000 kN. The results, including mean and standard deviation values of the compressive (f_c) and tensile splitting strength (f_ct,split), are mentioned in Table 1.

Moreover, a flowable cement-based grout with enhanced characteristics was used to fill the mold of the U-shaped jacket. According to the manufacturer’s technical datasheet SikaGrout-212 distributed by Sika Hellas ABEE in Kryoneri, Greece, is chloride-free and formulated to compensate for shrinkage, helping to reduce durability issues and corrosion in non-aggressive environmental conditions. Further, the high strength and the outstanding resistance to freeze/thaw cycles, oil, and water are some of the main merits of the applied grout. In addition, the mean compressive strength of the material was 41.3 MPa, while the flexural strength was 7.80 MPa.

2.3. Test Setup and Instrumentation

The beams were subjected to progressively increasing loading until total failure under a four-point bending experimental setup, as outlined in Figure 4. The “B” beams were supported at their edges on a rigid laboratory frame using two steel rollers. A 200 kN hydraulic pinned-end actuator applied the load at the mid-span of the beams’ top surface through two hardened steel rollers, affixed in a rigid metallic slab.

A load cell was fixed to the actuator to accurately measure the load with a precision of 0.05 kN. The load was incrementally subjected in a force control method of a 5 kN rate until diagonal cracking of the concrete occurred. Thereafter, the subjection was transitioned to a displacement control method. High-precision linear variable differential transducers (LVDTs) were utilized to measure the average net mid-span deflections of the beams, demonstrating an accuracy of 0.01 mm. Particularly, two LVDTs were positioned at the mid-span of the beams, while the other two were located at the supports to enhance the precision of the deflection measurements. Throughout the testing, both load values and corresponding deflections were continuously recorded.

2.4. PZTs’ Installation

Small and thin PZT patches, with dimensions 10 × 10 × 2 mm, in different configurations were placed in the retrofitted beams. Further, the PZTs were strategically positioned in the following locations: (a) embedded within the core of the initial beam, (b) externally epoxy-bonded to the jacket’s outer surface, and (c) on the core-outer surface shell IZ. Notably, two PZTs were referenced at the same point, with one placed on the front surface and the other at the back.

Thus, the total sensors’ resources for retrofitted beam “B500-J” consisted of four pre-installed PZTs from the initial beam “B500”, configurated as “Smart Aggregates”. The orientation of the polarization direction of the embedded transducers is considered a crucial factor in their monitoring effectiveness and is depicted with their embedded angle in the relevant figures. Henceforth, this configuration is denoted as “SA”. Additionally, six PZTs were epoxy-bonded on the IZ between the initial concrete surface and the inner mortar layer of the U-shaped jacket, henceforward referred to as “J”. Furthermore, two epoxy-bonded PZTs have been attached to the jacket’s external surface, further denoted as “X”. The configuration, positions, and types of the employed PZTs are illustrated schematically in Figure 5.

In addition, Figure 6 depicts the configuration, the positions and the notations of the attached PZTs of beam “B200a-J”. Thus, three pre-installed “Smart Aggregates” of the initial beam “B200a” were “survived”. One PZT of the left span has been destroyed due to the extensive shear cracking development in the left span of the initial beam. Furthermore, six PZTs are epoxy-bonded on the interfacial between the concrete surface and the mortar layer of the U-shaped jacket. Moreover, two epoxy-bonded PZTs have been attached to the jacket’s outer surface.

Further, Figure 7 presents the positions, the notations, and the configuration type of the employed PZTs. Thus, three pre-installed “Smart Aggregates” PZTs were saved from the initial beam “B200b”. One PZT of the left span was destroyed due to the extensive cracking formation developed in its “neighborhood”, during testing of the initial beam. In addition, six PZTs were epoxy-bonded on the interfacial surface of the U-shaped jacket. Moreover, two PZTs have been attached with epoxy to the jacket’s outer surface.

At this point, it is noteworthy to be mentioned that all the embedded PZTs have been coated with epoxy-resin before their installation to enhance their water-proofing resistance and to develop a protection layer, primarily for the tasks of the casting process and later for the potential cracking development. Moreover, the initially embedded PZTs had been further coated with cement-based mortar, forming the configuration of a “Smart Aggregate”, initially to improve their cohesion with the casting concrete, avoid forming voids, and increase their resistance under an aggressive environment due to subjected load and cracking propagation or/and the penetration of chemical agents.

As depicted in Figure 2 and Figure 3, the destroyed embedded PZTs of the initial beams B200a and B200b were positioned directly to the defected span area, where extensive shear cracking formation had developed. Moreover, the repair works for removing the defective concrete section triggered further damages to the “infrastructure” of the patches, cutting the connected wires.

Thus, the PZT’s functionality can be assured in a long-term period under normal operational conditions for the monitored element. Whenever severe circumstances occur near its monitoring area, it directly reflects the transducer’s functionality, potentially affecting the installation parameters (coating materials, epoxy-resin, wires, PZT patch). The latter is depicted either with massive deviations to the curves of the previously acquired EMI responses or total connection loss, constituting a significant indirect warning of damage development, raising the possibility of a false alarm.

On this scope, in real-case applications, the integrity assessment should be based on an array network of proximity PZTs’ monitoring a common area. Hence, the evaluation should also be reckoned in multiple inserted signs to extract a warning, following appropriate sensory function and communication protocols.

2.5. Monitoring Process

This experimental work implemented a wireless SHM system based on the fundamental principles of the EMI method employing piezoelectric sensors. The PZTs were selected due to their unique properties of generating a surface electric charge under mechanical stress and, conversely, undergoing mechanical deformation when exposed to an electric field. Consequently, any changes observed in the electrical impedance of the implemented PZTs are extracted from variations in the mechanical impedance of the monitored area. In addition, PZT’s voltage responses are recorded at multiple loading levels to assess the component’s structural integrity (Figure 8). Thus, fluctuations in these measurements within a specific excitation frequency range indicate potential structural degradation of the monitored element.

In the proposed approach, PZT patches were affixed to the interfacial between the existing concrete of the initial beams and the inner layer of the applied mortar using epoxy resin. Thus, the efficacy of the retrofitting scheme was investigated. The incremental imposed loading leads to beam deformations, which, in turn, develop stress on the interfacial part of the jacket.

With the development of stress due to the load subjection, the shear deformation leads to crack formation within the specimen. As a result, different effects are induced as the jacket’s structural mechanisms resist overcoming the imposed incremental loading. Hence, the development of stress in the interfacial surface and possible slippage between the U-shaped jacket and the initial concrete beam are expected due to the loss of bonding conditions.

Further, a PZT patch bonded on the interfacial surface could detect the development of cracks in the concrete and any potential jacket’s slippage. For this reason, six PZTs were epoxy-bonded to the interfacial surface of the “B-Js”, as shown in the close-up of Figure 3.

In addition, custom-made and designed impedance analyzers, referred to as “WiAMS” (Wireless Impedance/Admittance Monitoring System), are implemented to acquire the EMI responses from the PZTs (Figure 8). A WiAMS device is assigned to capture the EMI responses of a PZT patch within a specified frequency range of 10 to 250 kHz, featuring a frequency resolution of 1 kHz. The acquired EMI responses are subsequently uploaded to an online database for further data analysis (Figure 8). Consequently, the EMI response of each PZT patch in the specimens’ pristine (undamaged) state is recorded to establish a baseline. Furthermore, to assess the efficacy of the retrofitting method, the EMI responses of the attached PZT patches are reassessed after completing each loading level, corresponding to the relative damage state/condition.

As mentioned above, the EMI measurements could indicate the structural integrity degradation reflected by the monitored area’s mechanical impedance changes and demonstrated in voltage or/and frequency response curve shifting. The interaction between the PZT transducers and the host structure generates a voltage response composed of a real component (conductance) and an imaginary component (susceptance). This response reflects essential structural properties and can be represented by the complex admittance of the PZT patch, as follows in Equation (1).

$$\overline{Y}={\displaystyle \frac{\overline{I}}{\overline{V}}}=G+Bj=4\omega j{\displaystyle \frac{L^2}{h}}\left[\overline{{\epsilon }_{33}^T}-{\displaystyle \frac{2d_{31}^2\overline{Y^E}}{\left(1-\nu \right)}}+{\displaystyle \frac{2d_{31}^2\overline{Y^E}}{\left(1-\nu \right)}}\left({\displaystyle \frac{Z_{a,eff}}{Z_{s,eff}+Z_{a,eff}}}\right)\left({\displaystyle \frac{\mathit{tan}kL}{kL}}\right)\right]$$

(1)

where, $\overline{V}$, $\overline{I}$, G, B, j, L, h, d₃₁, Z_a,eff, Z_s,eff, n, v, k, $\overline{Y^E}$, and ${\epsilon }_{33}^T$ represent the Harmonic alternating voltage supplied to the circuit, the current passing through the PZT, the real part of the admittance (conductance), the imaginary part of the admittance (susceptance), the imaginary unit, the angular frequency, the half-length of the PZT patch, thickness of the PZT patch, piezoelectric strain coefficient of the PZT, short-circuited effective mechanical impedance, effective structural impedance, Poisson’s ratio, wave number related to the angular frequency, complex Young’s modulus of elasticity under a constant electric field, and complex electric permittivity of the PZT patch along the axis at constant stress, respectively.

3. Results

3.1. Mechanical Responses

In Figure 9, the mechanical performances of the “B” and “B-J” beams (initial and retrofitted) are illustrated in terms of applied stress and mid-span deflection. Additionally, the experimental results indicate that adding stirrups in the shear span improves the beams’ shear strength capacity and overall behavior, which is in alignment with the experimental design expectations. A detailed comparative analysis of the mechanical responses is presented in the relatively published experimental work of Chalioris et al. [2].

3.2. EMI-Based Monitoring

The following sub-sections present the curves of the voltage responses of the retrofitted beams. Strong indications and prompt conclusions can arise from the observations of the curves. Critical information about the structural integrity of the beams can be extracted from the comparison between the opposite and symmetrically placed PZTs. Thus, variations in the curves of a PZT placed in one side of the beam, while the relatively symmetrical one does not exhibit similar performance, strongly indicate forthcoming damage (crack) near its monitoring area. The more symmetrical PZT pairs are examined, the more accurate extracted inferences can be raised.

3.2.1. Beam “B500-J”

Figure 10 shows the shear stress response of the “B500-J” in terms of mid-span deflection and the crack width acquired by the crack meter transducers. The beam demonstrated a brittle shear failure aligning with its intended design, featuring a critical diagonal crack on the left span. One crack meter is positioned in the middle of the beam, monitoring the propagation of flexural cracking, while the other two are attached to monitor the rate of widening of the right and left shear diagonal cracks, respectively. Moreover, Figure 11 presents the crack meters’ positions and the cracking pattern of the beam at each loading level (damage state). Figure 10 also depicts the experimental voltage measurements through the EMI-based method in a red circle.

At first, for “B500-J”, eight EMI measurements were conducted. The primary measurement was acquired in the healthy state (pristine condition), notated as the “Healthy” state, while the other seven were acquired during the subsequent loading and cracking development increments, notated with the prefix Dam following by the increasing serial number of each state. Hence, Figure 7 mentions these states as “Dam1-Dam7”, respectively. According to the illustrated cracking pattern, obviously, the beam failed in the left span. Thus, according to their positions, the PZTs with the highest involvement are SA1, SA2, J1, and J4. Figure 12 presents comparative typical voltage response curves in terms of the set frequency range for PZTs J1 and J3.

From the observation of Figure 12, it is apparent that the PZT positioned on the failure side shows significant alterations in the voltage response curves with remarkable peak shifting. Further, the curves’ variations commence at early damage states (Dam3), which is an essential indicator of the method’s effectiveness in detecting the structural degradation of the retrofitting scheme (PZT J1). As presented also in Figure 10, the development of the fatal shear crack starts at Dam3, while in the next states (Dam4, Dam5, Dam6), the acquired values exhibit double widening from state to state, from approximately 2 mm to 4 mm, 4 mm to 8 mm, and 8 mm to 15 mm, respectively.

3.2.2. Beam “B200a-J”

Figure 13 depicts the shear stress of the “B200a-J” in terms of mid-span deflection and the crack width acquired by the crack meter transducers. As it was designed, the beam exposed brittle shear failure, forming a feature critical diagonal crack on the right span. One crack meter transducer is placed in the middle of the beam, monitoring the propagation of flexural cracking, while the other two are attached to monitor the widening rate of the right and left shear diagonal cracks. Figure 14 presents the crack meters’ locations and the cracking pattern of the beam at each loading level. Figure 13 also depicts in a red circle the values of shear stress and mid-span deflections, at the time of the experimental voltage measurements, through the EMI-based method (damage states).

Further, for “B200a-J”, seven EMI measurements were conducted. The initial measurement was conducted in the healthy state (pristine condition), while the six subsequent were acquired during the loading augmentation and cracking development. Figure 14 mentions these states as “Dam1-Dam6”, respectively. According to the illustrated cracking pattern, obviously, the beam failed in the right span. Thus, according to their positions, the PZTs with the highest involvement are SA3, SA4, J3, and J6. Figure 15 presents comparative typical voltage response curves in terms of the set frequency range for PZTs J3 and J6.

From the observation of Figure 15, apparently, the PZT positioned on the failure side shows significant alterations in the voltage response curves with significant peak and curve shifting. Further, the curves’ variations commence at early damage states (Dam1), which is an essential indicator of the method’s effectiveness in detecting the possibility of slippage, the occurred damages, and the structural degradation of the retrofitting scheme (PZT J3).

As presented also in Figure 13, the development of the fatal shear right crack starts at Dam2, while Dam3 starts the recording of the widening. Between the 4th and 5th damage states there is an abrupt increase from approximately 1.5 mm to 6 mm, which is reflected also by the cracking pattern pictures of Figure 14. In addition, right after the completion of Dam5, a second abrupt increment of the crack width was observed (from 6 mm to ~12 mm). Thereafter, concrete spalling of the jacket’s right shear span occurred and the transducer had been protectively removed.

3.2.3. Beam “B200b-J”

Figure 16 illustrates the shear stress of the “B200b-J” in terms of mid-span deflection. Further, the crack width acquired by the crack meter transducers is also depicted in the same figure. As designed, the beam showed brittle shear failure, forming a feature critical diagonal crack on the right span. One crack meter transducer is affixed in the middle of the beam, monitoring the propagation of flexural cracking, while the other two are attached to monitor the widening rate of the right and left shear diagonal cracks. Figure 17 presents the crack meters’ locations and the cracking pattern of “B200b-J” at each loading state. Figure 16 also illustrates, in a red circle, the values of shear stress and mid-span deflections, at the time of the experimental voltage measurements, acquired through the EMI-based method (damage states).

Further, for “B200b-J”, seven EMI measurements were acquired. The primary measurement was performed in the healthy state (pristine condition), while the six subsequent were conducted during the loading augmentation and cracking propagation. Figure 16 illustrates these states as “Dam1-Dam6”, respectively.

According to the presented cracking pattern, apparently, the beam failed in the right span. Hence, according to their positions, the PZTs with the highest involvement are SA3, SA4, J3, and J6. Figure 18 presents comparative typical voltage response curves in terms of the set frequency range for PZTs J1 and J3.

From the observation of Figure 18, apparently, the PZT positioned on the failure side shows significant alterations in the voltage response curves with significant peak and curve shifting. Further, the curves’ variations commence at early damage states (Dam1), which is an essential indicator of the method’s effectiveness in detecting deformations in the interfacial surface, and thus potential slippage and structural degradation of the retrofitting scheme (PZT J3). Further, PZT J3 could not perform the measurement at loading state Dam6 due to malfunction, which occurred through the extensive cracking in its region. This case also confirms the process described in Section 2.4 regarding the functionality of the PZTs under the development of severe cracking propagation.

As illustrated in Figure 16, the fatal shear right crack development starts at Dam2, while Dam3 starts the recording of the widening values. Up to the 4th damage state, the crack width gradually increases (from 1 mm to 2 mm). Thereafter, up to the 5th damage state, the width reached a 5 mm range, and between the 5th and 6th states, the crack width doubled, reaching approximately 14 mm.

4. Data Analysis

The extant literature review proposes the application of statistical indices to evaluate and quantify variations in EMI responses caused by modifications in the mechanical impedance of the monitored region. A commonly employed index is the Root Mean Square Deviation (RMSD), which transforms the variations in the EMI responses of a PZT patch between its pristine and examined states. The RMSD expression is provided in Equation (2):

$$RMSD=\sqrt{{\displaystyle \frac{\sum _{r=1}^M{\left({\left|Z_{\left(f\right)}\right|}_D-{\left|Z_{\left(f\right)}\right|}_0\right)}^2}{\sum _{r=1}^M{\left({\left|Z_{\left(f\right)}\right|}_0\right)}^2}}}$$

(2)

In this equation, ${\left|Z_{\left(f\right)}\right|}_D$ displays the absolute impedance value of the PZT in the examined state (D), while ${\left|{\mathrm{Z}}_{\left(\mathrm{f}\right)}\right|}_0$ represents the absolute impedance value of the PZT in its initial state, and M is the considered excitation frequency number. This comparative analysis is crucial for detecting and evaluating structural variations, providing valuable insights into the beam’s integrity under different loading conditions. It is important to note that all voltage measurements were carried out under controlled laboratory conditions to minimize the effects of temperature and humidity fluctuations on the acquired data.

4.1. Statistical Analysis

4.1.1. B500-J Evaluation

In this section, the values of the RMSD metric index for all the PZTs of “B500-J” are presented. Thus, in Figure 19b, the RMSD values of the epoxy-bonded on the interfacial zone are depicted. As shown, the values of “J1” and “J4” align with the formed cracking in the left shear span. Further, a significant observation is that both PZTs show an early warning sign about the imminent damage in their area, indicating also the loss of cohesion in the interfacial surface, as at the Dam2 state, the cracking development was limited to a small distinctive shear one with no significant width and range (Figure 11). On the other hand, the PZTs located in the right span show negligible changes, especially if these are compared with the relative of the left span, which is in alignment with the span’s structural integrity in the failure state. Further, the PZTs located in the middle of the beam demonstrate a similar performance and exhibit low values of RMSD compared to those of the left shear span, which is proven by the integrity of the lower mid-part of the beam.

Moreover, the “SA” RMSD values (Figure 20b) follow the formation of the developed cracking pattern (Figure 10). Thus, SA2 demonstrates significantly higher RMSD values than all the relative embedded PZTs. The latter comes in agreement with the beam’s structural integrity with respect to the PZTs’ locations. Further, even though the externally epoxy-bonded PZTs are considered not as sensitive as the embedded ones, “X2” shows higher RMSD values than the relative of “X3”, as depicted in Figure 21b.

From all the above observations and the cracking pattern formation presented for each set of PZTs, it could be extracted that all the different PZT patches’ configurations achieved successfully in identifying the critical effect that was assigned to investigate. Hence, the formation and the development of the damage, which finally led to a fatal shear critical crack in the left span of the beam for both the jacket and retrofitted beam, was successfully detected.

4.1.2. B200a-J Evaluation

Figure 22, Figure 23 and Figure 24 present the RMSD values for all the PZTs of beam “B200a-J”. For the PZTs located in the interfacial zone, “J3” exhibits an incremental trend, either from the first loading level until the failure, while “J6” shows a significant increment in the RMSD values, reaching the ultimate ones at the 5th loading levels (Figure 22b). Both PZTs are located in the right span, where the fatal shear failure crack occurred. All the other “Js” demonstrate insignificant variations, following the structural integrity condition of their efficient monitoring area, in terms of their location, except for “J2” at Dam, and Dam3 loading states, where the flexural crack at the middle of the beam occurred. In these states, the RMSD values of the PZT exhibit significant augmentation, while after that, they return to lower insignificant values.

Moreover, the “SA” RMSD values (Figure 23b) follow the formation of the developed cracking pattern (Figure 10). Thus, SA3 and SA4 demonstrate significantly higher values of RMSD, compared to all the relative embedded PZTs. The latter comes in agreement with the beam’s structural integrity with respect to the PZTs’ locations. Further, even though the externally epoxy-bonded PZTs are not as sensitive as the embedded ones, “X2” shows higher RMSD values than the relative of “X3”, as depicted in Figure 24b.

From all the above observations and the cracking pattern formation presented for each set of PZTs, it could be extracted that all the different PZT patches’ configurations achieved successfully to identify the critical effect that was assigned to investigate. Hence, the formation and the development of the damage, which finally led to a fatal shear critical crack in the right span of the beam for both the jacket and retrofitted beam, has been successfully detected.

4.1.3. B200b-J Evaluation

Figure 25, Figure 26 and Figure 27 present the RMSD values for all the PZTs of beam “B200b-J”. For the PZTs located in the interfacial surface, “J3” exhibits increased RMSD values, without performing a trend, but showing some fluctuations from state to state, indicating that there were no constant stress impacts over the surface. Further, “J6” presents an ascending trend from the 2nd loading state to the 5th. Both PZTs are located in the right span, where the fatal shear failure crack occurred. All the other “Js” demonstrate insignificant variations, following the structural integrity condition of their efficient monitoring area, in terms of their location.

Moreover, the “SA” RMSD values (Figure 26b) follow the formation of the developed cracking pattern (Figure 13). Thus, SA3 demonstrates significantly higher RMSD values than all the relative embedded PZTs. The latter comes in agreement with the beam’s structural integrity with respect to the PZTs’ locations. In addition, higher values of RMSD are expected for SA4; however, this was achieved only at the later loading states. Further, despite the externally epoxy-bonded PZTs being less sensitive than the embedded ones, “X3” exhibits slightly higher RMSD values than the relative of “X2”, as depicted in Figure 27b.

From all of the above observations and the cracking pattern formation presented for each set of PZTs, it could be extracted that all the different PZT patches’ configurations achieved successfully to identify the critical effect that was assigned to investigate. Hence, the formation and the development of the damage, which finally led to a fatal shear critical crack in the right span of the beam for both jacket and retrofitted beam, could be assessed.

5. Discussion

This study investigates the efficacy of an SHM scheme in identifying the failure mechanisms of U-shaped jacketing and detecting damages in retrofitted RC beams. The assessment entails the implementation of advanced structural technologies based on the EMI method facilitated by PZT sensors. Additionally, the statistical metric RMSD was utilized to analyze the collected data comprehensively, quantify the extent of damage, and assess the efficacy of the SHM method.

Furthermore, the study seeks to evaluate the effectiveness of the PZT-enabled SHM system in detecting variations in the structural integrity of an existing Reinforced Concrete (RC) element. This was achieved by strategically installing PZT transducers in various configurations across the examined RC elements. By analyzing a large set of measurements collected from the multiple deployed PZTs, the study aims to ensure the reliability and accuracy of the proposed monitoring method. The diversity in sensor placements and configurations provides a robust dataset, thereby enhancing the system’s capacity to identify and assess any structural alterations. To the authors’ best knowledge, there is no similar experimental project investigating the effectiveness of retrofitting schemes by implementing advanced structural monitoring technologies.

6. Conclusions

The PZTs installed on the beams effectively detected the initiation of cracks within their respective monitoring areas, with nearly all of them even anticipating the imminent development of these cracks.

• The PZTs “J”, positioned in the interfacial surface/zone, exhibit remarkable performance as in almost all the examined cases, they achieved to identify, even at early states, the changes in the structural integrity of the jacket.
• The RMSD index values proved to be effective indicators of damage severity and progression, allowing for a more precise assessment of the structural integrity of the beams.
• By comparing the symmetrical PZTs and their RMSD values, the severity of the jacket’s degradation and the area of the forthcoming failure were also achieved.
• The voltage response curves obtained from the PZTs exhibited significant alterations at their resonant frequencies. These alterations signify a strong correlation between the voltage measurements and the structural integrity, affirming that voltage signals could serve as reliable indicators for early damage detection and monitoring.
• An in-depth examination is necessary to evaluate the qualitative information provided by the acquired signals, such as the shifting of voltage peaks and frequency resonant frequencies.

The utilization of the proposed Structural Health Monitoring (SHM) method for real-time monitoring, aimed at detecting damages in retrofitted RC elements with U-shaped mortar jackets and predicting the jackets’ slippage, represents a noteworthy contribution of this experimental work to RC structural applications, thereby optimizing their efficiency.