Advanced Fiber Optic Sensing Technology in Aerospace: Packaging, Bonding, and Calibration Review

Abstract

With the continuous development of science and technology, aircraft structural health monitoring (SHM) has become increasingly important in the aviation field. As a key component of SHM, wing deformation monitoring is of great significance for ensuring flight safety and reducing maintenance costs. The traditional strain gauge measurement method can no longer meet the needs of modern aeronautical engineering. Fiber Bragg grating (FBG) sensors have been widely used in the engineering field due to their unique advantages, and have shown great potential in aircraft wing deformation monitoring. In the context of SHM in the aircraft field, this article provides an overview of four aspects: classification and principles of fiber optic sensors, packaging forms of FBG sensors, bonding technology, and calibration technology. The packaging forms includes tube-packaged, embedded package and surface-attached package. It then discuss the bonding technology of FBG sensors, and the principle and influencing factors of fiber optic bonding technology are analyzed. Finally, it conducts in-depth research on the calibration technology of FBG sensors. Through comprehensive analysis of these four aspects, the suggestions for optical fiber sensing technology in aircraft wing deformation measurement are summarized and put forward.

1. Introduction

With the development of current aviation technology, large aircraft SHM has become an indispensable part of aviation engineering. As an important part of SHM, wing deformation monitoring is very important to ensure flight safety and reduce maintenance costs [1,2,3]. As shown in Figure 1, in order to measure the structural health of aircraft, a traditional strain gauge is usually used for measurement, but this can form a complex sensor network [4]. The weight of the connection wire associated with the strain gauge is also very large, which will affect the strain measurement response of the structure. Thus, a convenient and high-precision wing deformation monitoring system for flight state is badly needed. FBG sensors have been widely used in the engineering field because of their low weight, anti-electromagnetic interference, low cost and easy mass production [5,6,7].

The NASA Langley Research Center used optical fiber sensors and strain gauges in the test and performed multiple flight maneuvers and the results show that the FBG sensor system performed well throughout entire flight [8]. Researchers of NASA used 2D shape sensing to measure wing deflection and the results verified the excellent characteristics of FOSS in wing shape measurement. In 2018, researchers of NASA published a research report on the use of optical fiber shape sensing [9]. As shown in Figure 2, researchers have applied optical fiber strain sensor networks to the strain and displacement monitoring of wings in a small unmanned aerial system (SUAS), and the results show that the methods and techniques of load redistribution and active shape control proposed by the researchers are effective. Martins et al. used the distributed optical fiber sensor network to measure the dynamic strain of aerospace structures, and the results show that the SHM algorithm based on the distributed optical fiber sensor network can represent the dynamic behavior of the wing model globally [10].

Hyuk et al. designed a full-scale static load detection system for measuring the temperature-compensated strain of the wing structure of a full-scale small aircraft [11]. The researchers applied the load to the tip of the wing structure, and the low-speed interrogator could detect changes in the central wavelength of the FBG sensors at 10 Hz sampling rate. The experiments verify the applicability of the low-speed FBG interrogator in small aircraft. He et al. studied the influence of optical fiber sensors with different embedded depths on shape monitoring of deformed wings, and analyzed the sensitivity of FBG sensors with different embedded depths in polyimide film [12]. The end curvature of the wing is measured by FBG sensors, and then the three-dimensional shape of the skin is reconstructed by integration. The experimental results show that the three-dimensional shape obtained by FBG sensors is in good agreement with the results obtained by visual measurement.

The research progress of many scholars above shows the outstanding effect of fiber optic sensors in detecting aircraft wing deformation. In addition, fiber optic sensors can achieve structural health monitoring of aircraft, strain measurement during flight, environmental temperature measurement during aircraft operation, among others. For various application scenarios, this paper categorizes fiber optic sensors according to their application situations, as shown in

Table 1. Classification of fiber optic sensor applications in the aviation field.

Application Area	Function	Example	Ref.
Structural health monitoring	Real-time monitoring of structural integrity and detection of damage	FBG array in composite beam/wing box	[13,14,15,16]
Load/strain/vibration monitoring	Detect changes in load and vibration modes	FBG array for hypersonic aircraft	[17,18,19]
Thermal environment monitoring	Evaluate the temperature of the thermal protection system or the surface of the body	STORT aircraft thermal protection system monitoring	[20,21,22]
Photon system support	Support high-speed optical communication and modulation systems	FBG demodulator, tunable laser, filter	[23,24,25,26]

.

As shown in the table above, fiber optic sensors have a variety of application scenarios. However, in practical engineering, the packaging and bonding of fiber optic sensors may be subject to certain interference, which will affect the measurement results. Based on this reason, this paper will introduce the classification and working principle of fiber optic sensors in aircraft application scenarios, analyze the packaging form of fiber optic sensors, and discuss the factors that affect the strain transmission efficiency of fiber optic sensors.

To ensure the rigor and credibility of the research, this paper summarizes the research results of using FBG sensing technology for aircraft wing deformation monitoring through a standardized literature search. The main data sources are Web of Science, Scopus, and IEEEXplore databases, with literature concentrated in academic research literature from the past decade. This review follows the following approach when selecting the literature: (1) the research focuses on the application of FBG sensors for aircraft wing deformation monitoring, rather than their use in civil engineering or marine detection; (2) includes data or engineering validation from wind tunnel or flight experiments, with verifiable engineering data or experimental evidence; and (3) the core technology research direction involved in the article (sensor types, packaging forms, bonding forms, calibration methods) can support the research content of this article. Therefore, the literature selected from the above literature is attached to the reference list for application in this study.

2. Classification of Optical Fiber Sensors

The classification of fiber optic sensors can be based on the characteristics of the materials. This section introduces several main types, including FBG sensors, sensors based on Rayleigh scattering and Brillouin scattering, etc., and analyzes their principles and application characteristics [27].

2.1. FBG Sensor

K.O. et al. discovered the photosensitive effect of germanium-doped fiber, and successfully developed the world’s first fiber grating by standing wave method [28]. Meltz et al. used two coherent ultraviolet beams to write a fiber Bragg grating into a hydrogen-loaded fiber. The fabrication technology makes fiber Bragg grating have potential application value [29]. Hill et al. proposed the phase mask method, which promoted the industrialization of fiber Bragg grating applications, and promoted the application of fiber Bragg grating in the sensing field, and the principle is shown in Figure 3 [30]. When broadband light enters the fiber Bragg grating (FBG) from the left side, due to the periodic refractive index modulation in the fiber core, it will selectively reflect light of a specific wavelength, forming a back reflection spectrum, and sharp reflection peaks appear at the corresponding wavelength. At the same time, the unreflected light is transmitted through, and the transmission spectrum shows troughs at these specific wavelengths. The principle behind a FBG sensor is to precisely measure the Bragg wavelength of the reflected radiation. Since this is sensitive to environmental influences like temperature or strain, any changes in the Bragg wavelength indicate a presence of the disturbance that one wants to measure.

2.2. Rayleigh Scattering

In the optical fiber production process, there is a phenomenon that the material of fiber core and the refractive index inside are not even hooked. When the light wave propagates in the fiber, it produces the scattering in all directions, which is called Rayleigh scattering. Rayleigh scattering belongs to elastic scattering, and the frequency of scattered light is the same as that of incident light wave. The intensity of Rayleigh scattering is directly proportional to the original light intensity, and the ratio of the two is called Rayleigh scattering coefficient [31,32,33,34].

Rayleigh scattering is a naturally occurring phenomenon caused by the elastic collision between incident light and micro-particles in the medium. During the fabrication of optical fibers, the refractive index distribution in the optical fiber has a slight difference and impurity. The defect of this structure causes the fluctuation of dielectric constant, which makes the amplitude and phase of Rayleigh scattering random. When the optical fiber is manufactured, its refractive index distribution is very stable, and Rayleigh scattering can be considered as a long-term stable phenomenon of the optical fiber itself.

Barnosky et al. first proposed OTDR technology based on Rayleigh scattering [35]. OTDR technology can effectively measure the attenuation and continuity of optical fibers. It is very suitable for measuring the length of optical fibers, determining the loss of optical fibers in all areas and locating the fault points in optical fibers. Daichi et al. designed the flight test bed and they used their own OFDR system to demonstrate the distributed sensing flight of the main wing of a medium-sized jet airliner [36]. OFDR-FBG sensing system can measure the strain distribution in FBG in real time with millimeter-level spatial resolution. The results show that the OFDR-FBG sensing system can successfully monitor the variation in strain distribution according to various operations, and the measured strain is in good agreement with the reference strain.

2.3. Brillouin Scattering

Ippen et al. first observed Brillouin scattering spectrum in optical fibers [37]. Since then, researchers have conducted extensive research on stimulated Brillouin scattering (SBS) technology in optical fibers, and Horiguchi et al. proposed Brillouin optical time-domain analysis (BOTDA) in 1989 [38]. Researchers have found that BOTDA has the characteristics of long measuring distance and high measuring accuracy [39,40].

According to the characteristics of the sensor, the distributed sensors also includes Distributed Phase-Encoded Brillouin Optical Time-Domain Analysis (DPP-BODTA), Brillouin Optical Time-Domain Reflectometer (BOTDR), Optical Frequency-Domain Reflectometer (OFDR), Phase Optical Time-Domain Reflectometer (OTDR) [41,42,43]. The different characteristics are listed in

Table 2. Performance chart for distributed sensors [41,42,43].

	Spatial Resolution	Measurement Time	Temperature and Strain	Strain Accuracy	Dynamic Measurement
DPP-BOTDA	2 m	2–5 min	Yes	20 $\mathsf{\mu }\mathsf{\epsilon }$	Yes
BOTDR	~1 m	1–5 min	Yes	60 $\mathsf{\mu }\mathsf{\epsilon }$	No
OFDR	~1 mm	(0.01–3) s	Yes	1 $\mathsf{\mu }\mathsf{\epsilon }$	30 Hz
Phase OTDR	~0.5 m	<1 ms	No	No	No

.

Fiber optic sensors applied in the aviation field also cover more technology routes designed for specific scenarios. These technologies are optimized around different usage requirements such as high temperature resistance, miniaturization, and multi-parameter monitoring, further expanding the application boundaries of fiber optic sensing in aircraft systems, as shown in

Table 3. Classification by sensor technology route.

Sensor Technology	Application Scenario	Optimization Focus	Ref.
Extrinsic Fabry–Perot Interferometer (EFPI)	Curing strain monitoring of aviation composite laminates and axial strain detection of wings	Expanding the dynamic range by nine times through a hemispherical polymer cap to achieve sub-nanometer-resolution displacement measurement	[44]
OTDR	Distributed strain monitoring of composite material bridges and detection of debonding damage to wing bolt lines	Improve measurement accuracy and spatial resolution, achieve continuous strain/deflection distribution measurement of the entire length of the structure	[45]
BOTDR	Long-distance strain/temperature monitoring of large-area aircraft structures	Improving spatial resolution based on iterative subdivision method, up to 0.1 m of 1 km fiber can be achieved under 10 ns pulse	[46]
FBG	Micro-strain monitoring in the manufacturing process of composite materials and multi-parameter distributed monitoring during service life	3D strain mapping achieved through ceramic coated FBG	[47]
Sapphire fiber EFPI	Strain monitoring of extreme high-temperature environment near the engine	Capable of withstanding temperatures exceeding 1000 °C and achieving synchronous measurement of high temperature and high pressure	[48]
Iterative subdivision BOTDR	Fine positioning of local strain in aircraft structure and optimization of long-distance monitoring	Can extract sub-Brillouin signals and optimize the ability to capture long-distance local damage	[49]
Artificial Neural Network BOTDR	Strain monitoring and dynamic data processing of aircraft structures under complex loads	Resolution can be increased from 21.13 MHz to 2.88 MHz under 60 ns pulse	[50]
Small diameter FBG	Strain monitoring of lightweight drone structures and damage detection of thin-walled composite materials	Reduce the impact on the mechanical properties of the substrate structure and adapt to the vacuum assisted resin infusion (VARI) process	[51]
High pressure vessel compatible with FBG (PTFE protection)	Strain monitoring during the manufacturing stage of aviation composite materials and microcrack detection during the curing process	Can resist high temperature and high pressure deformation, and can evaluate residual strain	[52]
Dynamic FBG Interrogation (FFT)	Real-time health monitoring of aircraft, rapid response to fluctuations and capture	Can improve wavelength scanning speed, with a resolution of 0.11 pm at a scanning frequency of 40 kHz	[53]
Anti torsional deformation FBG	Multi-directional load aircraft component strain monitoring and precise measurement of complex stress fields	Can reduce torsional deformation error, with a relative error of 0.86% in bending direction measurement	[54]
Distributed ROTDR	Monitoring of temperature field distribution on aircraft wings and thermal strain detection in wide-temperature environments	Can maintain constant Raman Stokes power reception and extend sensing distance	[55]
Wavelength Division Multiplexing FPI	Multi-area strain synchronous monitoring of aircraft and integration of multiple sensors	Can demodulate the length of each FPI cavity through a bandpass filter, breaking through the limitations of FPI multiplexing	[56]
Prestressed FBG	Monitoring of shrinkage strain in thick composite structures and control of deformation in laminated plates	Identifying spectral distortion caused by damage through a fixed FBG filtering decoding system	[57]

.

3. Packaging Forms of FBG Sensors

According to the different packaging technology of FBG, FBG sensors can be divided into tube-packaged, embedded packaging and surface-attached packaging. These three monitoring methods have their own characteristics and applicable conditions. In application, different monitoring methods should be selected according to the different mechanical states of the matrix.

3.1. Tube-Packaged FBG Sensors

Researchers have performed extensive research on tube-packaged FBG sensors and have applied them to different engineering fields [58,59,60,61,62]. Generally, a tube-packaged FBG sensor consists of a bare FBG placed inside a tube sleeve with a certain prestressing force applied to keep the FBG straight, and then encapsulation glue poured between the FBG and the fiber, so that the FBG is firmly embedded in the tube sleeve. The encapsulation adhesive has a certain strength and can transfer the strain of the structure to the FBG well.

Ivan et al. introduced a new method to package and install a large number of sensors on metal or composite structures based on the application background of naval ships [63]. The researchers noticed that the long FBG sensors had a high fracture sensitivity in the splicing area. They used Vacuum-Assisted Resin Transfer Molding (VARTM) technology to conduct large-scale test simulation in the laboratory. The results show that this method is suitable for large area ships which need long-term SHM. The method is simple and easy to use, and can avoid complex problems, especially in the field of long FBG sensors management. The method has been proved to be suitable for ships by experiments, but the wing of aircraft has high flexibility and works in totally different harsh environments. Only a limited number of FBG sensors can be used on aircraft wings, and Floris’s research [64] found that the advantages of new shape sensors based on long FBG make up for this limitation. Whether VARTM technology is applicable to aircraft wings can be further verified by experiments in combination with Floris’s research.

Li et al. proposed a capillary encapsulation method. In this method, zinc capillary and different binders were used to package FBG, and the temperature sensing characteristics of FBG in a constant-temperature water bath were studied [65]. The results show that this packaging method can protect the FBG well and improve the temperature sensitivity of FBG effectively. However, it should be noted that the material properties of capillary and adhesive have an effect on the temperature sensitivity of FBG. Different application backgrounds and strength requirements have different requirements for the binder materials.

Sagar et al. proposed a new metal encapsulation method for FBG sensors using stainless steel, tin and high-temperature-resistant samarium cobalt (SMCO) magnets [66]. The encapsulated sensor can be directly contacted with the matrix structure such as iron tube and other ferromagnetic elements without any adhesive, making it easy to disassemble and reuse. The feasibility of using metal-encapsulated FBG sensors for structural health monitoring is verified, which provides a good choice for SHM especially for ferrous pipeline monitoring and for welded structures.

Zhang et al. studied the characteristics of metallic-packaging FBG sensors with coatings and bare fibers, as shown in Figure 4 [67]. Researchers developed metallic-packaging FBG sensors by ultrasonic welding. The good combination of metal alloy with coating and bare optical fibers is clearly shown by cross-sectional scanning electron microscopy (SEM). Compared with bare FBG, polyimide coated FBG for metal packaging has advantages in stability, repeatability, and response.

3.2. Embedded FBG Sensor

Researchers have also performed extensive research on embedded FBG sensors, which has been applied in different engineering fields, and it has been widely used in the field of damage monitoring [68,69,70,71,72]. An embedded FBG sensor is a kind of technology that embeds FBG sensor in the matrix and fills some material to make the fiber and the matrix contact closely. In this way, the stress and strain changes in the matrix can be monitored by the optical fiber. An FBG sensor has the advantages of small size and high sensitivity, so embedded FBG in intelligent composite material not only keeps the FBG’s sensitivity, anti-electromagnetic interference, and other advantages, but also does not affect the performance of composite aeronautical structures [73].

The commonly used embedded FBG sensors are carbon-fiber-reinforced polymer (CFRP) FBG sensors, resin matrix composite FBG sensors, and fiber-reinforced polymer (FRP) FBG sensors, among others [74,75]. Researchers found that the optical fiber sensors directly embedded in the adhesive will lead to a significant reduction in the bonding strength, which directly affects the normal operation of the embedded FBG sensor in the aircraft [76]. Grundmann et al. [77] have performed extensive research on the bond strength of the optical fiber sensors embedded in the CFRP film bonding joints, and results show that the influence of embedding glass fiber with a total diameter less than 100 μm on the fatigue limit of film adhesive joints can be ignored. It is suggested that the P80 optical glass fiber is highly suitable for the application of SHM in aeronautical structures. Murayama et al. installed the embedded optical fiber sensors into the full-scale composite structure of the vehicle, and monitored the temperature and strain distribution of the fuselage during the manufacturing process [78]. The results show that the application of embedded optical fiber sensors in the health monitoring of composite structures is feasible.

Baker et al. developed FBG embedding technology for military aircraft [79]. A new and effective implementation process strategy is proposed, which overcomes the technical constraints of previous problems. The end of the FBG sensor is sealed in a Teflon envelope and then sealed with cement. A frankincense dam is built around the sample to prevent resin flow. The results show that the method is effective and easy to implement in aircraft structural health monitoring. Botsis et al. studied the application of embedded FBG sensors in internal strain measurement [80], and the results show that the reasonable use of embedded FBG sensors combined with appropriate models can be used to characterize the deformation of polymer materials at different scales. Li et al. analyzed the strain transfer characteristics of embedded FBG sensors under non-axial stress [81]. Researchers established a suitable strain transfer model and analyzed the influence of sensor positioning angle deviation on average strain transfer rate. The results show that the angle of embedded FBG sensor plays a vital role in the transfer of strain from the surrounding material to the core of the optical fiber. Takeda et al. of Japan applied embedded optical fiber sensing system to health monitoring of composite structures of smart aircraft [82]. Researchers have developed a real-time impact damage detection system embedded in small-diameter OFSS, and demonstrated it in CFRP fuselage structure with diameter of 1.5 m and length of 3 m. In order to effectively detect impact damage, researchers embedded small-diameter stainless steel in the upper panel.

Liang et al. used embedded FBG sensors to monitor the strain changes during loading, and compared the measured results with the theoretical results [83]. The validity and reliability of embedded FBG sensors for structural health monitoring are studied through experimental data. The experimental results show that the measured results are in good agreement with the theoretical results, and the straightness can reach more than 0.9998. As shown in Figure 5, Maxim et al. used embedded FBG sensors to layout on the wing frame for wind tunnel research and the layout array and method of sensors are distinguished by different colors in the figure [84]. The results show that embedded FBG sensors can provide higher spatial resolution measurement compared with accelerometers and strain gauges. They are well suited for embedded composite wing structures and do not require additional wiring for each sensor. Hyunseok et al. proposed a load monitoring system for a composite wing based on embedded FBG sensors [85]. The embedded FBG sensor is embedded in the main structure of the composite wing, and the strain response of the embedded FBG sensor is used to estimate the wing load. The experimental results show that the average error of this wing load estimation method is 4.19%, which shows that the embedded FBG sensors have certain advantages in wing load monitoring.

He et al. embedded FBG sensors in PVC-enhanced silicon substrate to monitor the three-dimensional shape of the morphing wing [86]. Due to the lack of flexibility of PVC substrate, the adaptability of FBG sensor is reduced. The soft silicone skin can protect the FBG sensor and enhance its adaptability. The researchers combined the advantages of the two materials to greatly improve the sensitivity of the FBG sensor. The researchers also used visual measurement experiments to prove the effectiveness of the FBG sensing method, and the results show that the effect of morphing-wing shape reconstruction based on embedded FBG sensor method is in good agreement with that of visual measurement, and the maximum error is less than 3%. It can be seen that the embedded FBG sensor has a good effect on the SHM of morphing-wing UAV.

3.3. Surface-Attached FBG Sensor

Researchers have performed extensive research on the surface-attached FBG sensor [87,88,89,90]. Compared with the tube-packaged FBG sensor and the embedded FBG sensor, the surface-attached FBG sensor has higher requirements for the bonding process. Some researchers attach FBG sensors directly to the surface of the measured object, and some researchers also cut a small groove in the sensor base and then fixed the bare FBG in the groove [91,92,93]. The main purpose of the groove is to increase the contact area between the substrate and the fiber, so that the strain of the substrate can be effectively transmitted to the FBG.

In order to meet the needs of structural testing and improve the performance of sensors, researchers use different materials to package bare optical fibers. Common matrix materials are copper sheet, aluminum sheet, alloy port sheet, polymer composite materials, etc. The binders used for packaging are epoxide resin, silica epoxide, ceramic binder, among others. Compared with the tube-packaged strain sensor, the surface-attached strain sensor is easy to fabricate, simple in structure and suitable for monitoring the surface strain of the carrier. Sirkis et al. used an integral equation to describe the phase change in light propagating in a single-mode optical fiber when it propagates in a single-mode optical fiber with arbitrary strain field [94]. The hypothesis that strain can be transferred by surface-attached FBG is verified by experiments, which indicates the feasibility of surface-attached FBG for strain measurement. Duncan et al. used distributed sensing systems to monitor structural fatigue of aircraft undergoing full-scale fatigue tests [95]. The researchers used surface-attached optical fiber sensors with high-density on Lockheed Martin’s P-3C Orion fatigue test. The results prove the feasibility of using OFDR technology in distributed optical fiber sensing for aircraft structure monitoring.

Li et al. tested the strain transfer performance of surface-attached FBG sensor with metal welding instead of epoxy [96]. The results show that the average strain transfer efficiency of metal welding surface-attached FBG sensor is kept at 99.2%, which can give full play to the strain measurement ability of surface-attached optical fiber sensors. Sidney et al. designed a special coated surface-attached FBG to evaluate the working quality of the surface-attached FBG sensor by exposing the installed sensor to a full set of real flight conditions [97]. The applied test conditions are based on aerospace standards, including temperature cycle, pressure cycle, exposure to humidity and hydraulic fluid, and fatigue load. The results show that the surface-attached FBG is compatible with flight conditions when it is fully attached on aerospace CFRP specimens.

Sodja et al. introduced the FBG sensing system for aircraft wing deformation reconstruction [98]. According to the position of surface-attached FBG strain sensor on the wing skin, two sensor configurations are formed along the axis and spars, respectively. The researchers used 20 strain rosettes at both aircraft wing skins and 8 FBG sensors per spar. Each skin also has 10 temperature FBG sensors for the correction of thermal strain. The design is very systematic. The compatibility, range, and coast down test of the air brake, engine, and electronic components are also introduced. This study is a good reference for the research of wing deformation sensing technology.

In order to visually compare the differences between tube-packaged, embedded, and surface-attached FBG sensors, and to select the appropriate packaging form according to their needs in practical applications, as shown in

Table 4. Comparison of the characteristics of FBG sensors in different packaging forms.

Characteristic	Tube-Packaged FBG	Embedded FBG	Surface-Attached FBG	Ref.
Core integration method	Place the bare FBG in the sleeve, apply pre tension and inject encapsulation adhesive to fix it, so that the FBG is embedded in the sleeve	Embedding reinforcing fiber layers during manufacturing, permanently embedded in composite laminates	Prepare composite laminates, and then use high-performance adhesives to stick FBG onto the surface of the cured laminates	[99,100,101]
Target sensing applications	Monitor structural strain and assist in long-term structural health monitoring	Monitor the internal strain and temperature during the curing/post curing stage of composite materials. Detect BVD and locate it	Detecting surface damage and locating it. Large-scale structural SHM (such as wind turbine blades)	[102,103,104]
Detecting physical parameters	Structural strain (dependent on strain transfer between sleeve encapsulation adhesive FBG)	Interlayer strain, interlayer temperature, impact, vibration, damage	Surface strain, surface temperature, impact, damage, vibration, pressure	[105,106]
Manufacturing process	To bare FBG casing, apply pre tightening force, and seal with adhesive. The casing material (metal, capillary, etc.) affects performance and may involve process validation such as VARTM	Select interlayer pre strain fixed FBG, operate before resin impregnation/curing. Preparation of laminated panels including manual layering, hot pressing, VARIM, etc.	Curing the surface of laminated panels, adhesive/matrix epoxy bonding pre strain FBG. Preparation of laminated panels including manual layering, hot pressing, VARTM, etc.	[107,108]
temperature sensitivity	Due to the influence of temperature characteristics of casing and encapsulation adhesive materials, attention should be paid to temperature–strain coupling	Sensitive to curing/chemical reaction temperature, affected by changes in fiber/matrix temperature	Easy to be affected by external environmental temperature, which may cause inaccurate sensing	[109,110,111]
Strain sensitivity	Dependent on encapsulation adhesive strain transfer, affected by sleeve constraints, it is necessary to optimize encapsulation parameters to ensure transfer efficiency	Axial strain disturbance detection is more sensitive	Surface measurement improves sensitivity, but the value may differ from the actual internal strain	[112,113]
Maintain characteristics	Replacement after damage requires removal of sleeves, which is relatively complex	Difficult to maintain and replace, requiring replacement of the entire intelligent composite material structure	Can be replaced and maintained for easy visual inspection	[114,115]
Defects related to laminated panels	If the compatibility between the encapsulation adhesive and the laminated board is poor, it may affect the local mechanical properties of the laminated board	Easy occurrence of resin-rich areas/bubbles, debonding, decreased mechanical properties, etc.	No obvious defects related to laminated panels	[116,117,118]
Defects of sensors	There is uneven strain transmission. Long-term use of encapsulation adhesive aging affects performance	There are angular offsets, residual strains, etc.	Poor bonding process may result in adhesive residue, weak bonding, and affect sensing	[119,120]
Long-term stability	When the packaging is good, the stability of the sleeve and packaging adhesive affects long-term performance. If the material is stable, it can work stably for a longer period of time	Due to the influence of internal stability of materials and manufacturing defects, long-term monitoring should pay attention to performance drift	Due to the influence of surface environment (humidity, external forces), long-term stability is prone to fluctuations and requires regular calibration	[121,122]
Strain transmission efficiency	Due to the influence of mechanical properties of packaging materials, optimize packaging parameters to ensure efficiency is needed	Transferred through the matrix material, close to the actual internal strain, but manufacturing defects can easily cause transmission deviation	Dependent on adhesive transfer, affected by adhesive properties and bonding processes, surface strain transfer is direct but may differ from the interior	[123,124,125]

, a systematic comparison is made between them from multiple characteristic dimensions such as core integration methods, target sensing applications, and detection of physical parameters.

Based on the review research in this part, it can be concluded that for the monitoring scenario of aircraft wing deformation, the specific selection should be determined according to the characteristics of different packaging methods. Tube-packaged FBG is suitable for measuring points that require long-term stability, such as the main wing beam, and has strong seismic resistance. Once damaged, the casing needs to be disassembled and replaced, making maintenance more complex. Embedded FBG is preferred for CFRP composite wings, which can monitor internal strain and be compatible with the curing temperature of composite materials. It has outstanding survival ability when integrated with the wing structure, but requires replacement of the overall composite material layer for maintenance. It has high strain transmission efficiency and is not affected by surface environmental interference. Surface-attached FBG is suitable for rapid deployment or modification of existing wings, can be directly replaced and maintained, and are easy to visually inspect, but are susceptible to surface damage that affects their survivability.

4. Bonding Technology of FBG Sensor

4.1. Principle of Optical Fiber Bonding Technology

As shown in Figure 6, the adhesive bonding mechanical transfer model [126,127] consists of FBG sensor, adhesive layer and measured structure. When the matrix is subjected to external forces, stress is transmitted to the grid region through an adhesive layer. During the stress transmission process, there is shear stress inside the adhesive layer, causing the stress to gradually change along the length direction (x-axis) from the matrix side’s. The fiber is connected to the substrate through an adhesive layer. When the substrate undergoes deformation under external force, the adhesive layer transmits the strain of the substrate to the fiber Bragg grating through shear action, thereby achieving sensing measurement.

Researchers have performed extensive research on the strain transfer characteristics of FBG sensors, and achieved differing results. The factors influencing strain transfer characteristics have been analyzed. The relationship between various factors and transfer rate has been provided in [128,129,130,131,132,133]. Due to the effects of the protective coating and adhesive layer, the strain between the optical fiber and the host material will be inconsistent. In studies on strain transfer efficiency, many focus on the theoretical analysis of strain transfer from the host material to the optical fiber. Shiuh-Chuan et al. established a theoretical model for the transmission of strain from the host material to the optical fiber in order to study the measurement accuracy of the optical fiber strain sensor [134]. Then, the theoretical model was used to evaluate the interaction between the host material and the coating. Through the finite element numerical analysis, the correctness of the theoretical prediction is verified by the researchers. Researchers used Mach–Zehnder interferometric optical fiber sensors to measure strain, and the strain difference between the optical fiber strain sensor and the specimen was measured in experiments. The results show that the strain measured on the optical fiber is lower than the true strain of the sample. The percentage of strain actually transferred to the optical fiber in the sample depends on the bonding length of the optical fiber and the protective coating. The general trend of strain transition obtained from experimental tests and theoretical predictions shows that the longer the bonding length is, the harder the coating is, and the more strain is transmitted to the optical fiber. Wang et al. analyzed the strain transfer of six-layer surface-attached FBG sensors under uniform axial stress [135]. A six-layer structure, which is different from the generally recognized four-layer structure, is proposed. It consists of optical fibers, protective coatings, bonding layers, base layers, external bonding layers, and host materials. The theoretical formula of strain transfer rate from the host material to the optical fiber is established, which provides a basis for accurate theoretical prediction. Through parameter analysis, the researchers determined the selection scheme of sensor parameters, which provided a reference for sensor design. The experimental results show that the measured data are in good agreement with the theoretical prediction. The simulation results align well with the theoretical predictions. Finally, the surface-attached FBG model with one substrate layer is extended to a general model with multiple substrate layers, which provides a theoretical basis for future research and design of surface-attached FBG.

4.2. Influencing Factors

4.2.1. Effect of Bonding Thickness

In the bonding process, the function of the bonding layer is to firmly adhere the bare FBG to the surface of the measured substrate, so the bonding layer needs a certain thickness. If the bonding layer is too thin, the FBG sensor is prone to fall off when the matrix is deformed. If the bonding layer is too thick, the strain of the FBG sensor is prone to slow down. Therefore, the influence of the thickness of the bonding layer on the strain transfer of the FBG sensor cannot be ignored.

Kai et al. studied the strain transfer effect of surface-attached FBG sensors [136]. A three-dimensional finite element model of a surface-attached strain sensor was established. The effects of side width, top thickness, bonding length, and bottom thickness on strain transfer rate were studied. The shear lag characteristic is modified by empirical method, which improves the calculation efficiency of strain transfer coefficient. The results show that bonding length and bottom thickness are the main factors affecting strain transfer rate. Wu et al. analyzed the strain transfer mechanism of FBG sensors [137]. Researchers put forward a polynomial shear stress distribution mathematical model, established a strain transfer function, and calculated the strain of the fiber center using ANSYS Workbench. The results show that the strain transfer function derived from polynomial shear stress is very close to the actual situation and meets the accuracy requirements. In addition, the strain values obtained in the middle part of the FBG sensors are accurate, but the strain values at both ends are lower than the actual values. Therefore, FBG sensors need to be long enough. The experimental data also show that the thickness of the viscous layer and the elastic modulus have great influence on the strain transfer coefficient. With the decrease in bonding thickness and the increase in elastic modulus, the strain transfer coefficient and average strain transfer coefficient gradually increase, and the effective length of FBG sensor also increases.

Through analysis, it can be seen the strain transfer coefficient increases with the decreases in the thickness of the bonding layer, that is, the greater the thickness of the bonding layer, the less sufficient the strain transfer. Therefore, in the process of bonding the bare FBGs and the host substrate, under the premise of ensuring that the bare grid and the substrate are firmly bonded, the thinner the thickness of the cement layer, the more sufficient the strain transmission. With the increase in shear modulus of bonding layer, the strain transfer coefficient increases and the strain transfer becomes more sufficient.

4.2.2. Effect of Bonding Width

Besides the influence of the thickness of bonding layer, the width of bonding also affects the strain transfer efficiency. The bonding width is the width perpendicular to the axial direction of the fiber when the bare FBG sensor is bonded to the substrate surface.

Zhao et al. proposed an improved surface-attached FBG sensor connection method for airship hull SHM [138]. The strain transfer relationship between the airship cladding and the core is predicted by theoretical formulas, and then the theoretical prediction is discussed by finite element analysis. On the basis of theoretical analysis and numerical verification, the geometrical and mechanical parameters of the adhesive layer affecting strain transfer rate and effective sensing length ratio were analyzed. The experimental results show that width and Young’s modulus are the key parameters, while bottom thickness and Poisson’s ratio have little effect. Guang et al. proposed a new bonding model to accurately evaluate the measurement accuracy of FBG sensors and correct the strain obtained [139]. The correctness of the bonding model was verified by numerical method, and a theoretical strain correction coefficient was obtained. The sensitivity of the strain correction coefficient to five main parameters was discussed. It was proved that the correction coefficient had significant sensitivity to the coating material and bonding length. The study provides a practical and accurate correction for surface-attached FBG sensors installed with adhesives, and helps to deduce the true strain in the measured object. Finally, the researchers suggested that in practical applications, FBG sensors with shorter length, larger area, and that are thinner improve the correction accuracy.

4.2.3. Effect of Bonding Length

In practical applications, the bonding length of FBG is also a factor to be considered, which directly affects the stress concentration of the sensor on the measured surface and the accuracy of surface measurement. Shiuh-Chuan et al. studied the effect of bonding length on strain transfer efficiency [140]. Researchers used Mach–Zehnder interferometric technology to perform experimental measurements. The experimental results show that the strain transfer increases from 56% to 82% as the bonding length of epoxide resin increases from 5 cm to 12 cm. That is to say, the actual percentage of strain transferred to optical fibers depends on the bonding length of optical fibers and adhesives. The experimental results verify the correctness of the theoretical prediction and reveal the strain difference between the optical fiber strain sensor and the specimen. The general trend of strain transfer obtained by researchers from both experimental and theoretical predictions shows that the longer the bonding length, the higher the bonding strength, and the more strains transmitted to optical fibers. Kwona et al. studied the effect of bonding length on the signal characteristics of surface-attached FBG sensors under different load types [141]. The experimental results show that the strain transfer ratio varies with the loading mode. If the FBG sensors are not equipped with enough bonding length, the strain cannot transfer adequately. Therefore, it is necessary to determine the effective combination length of different load types in order to better obtain stable signals from surface-attached FBG sensors.

However, this can cause the displacement of the measuring point and the inaccurate position of the measuring point if the sensor is too long. If the sensor is too short and the strain transmission is not enough, it is necessary to study the influence of the bonding length of the FBG sensor on the strain transmission of a specific substrate. It can be seen that the strain transfer coefficient decreases with the increase in bonding length, that is, the strain transfer becomes sufficient. Therefore, in the actual strain measurement, the bonding length of the bare grating should be increased as much as possible to meet its good strain transfer and the measurement accuracy will decrease if the bond length is too long. The strain measurement should be carried out according to the usual embedded length of the FBG sensor.

5. Calibration Technology

The calibration measurement method is a vital link in the process of SHM technology. NASA has mature technology to calibrate the wing with strain gauges [142,143,144,145]. FBG sensor has many advantages over traditional strain gauges. Therefore, in addition to the installation form of optical fiber sensors and the design of bonding technology, it is also necessary to calibrate the optical fiber sensors accurately to achieve long-term, effective, and accurate measurement.

Hönisch et al. made a thorough study on the details of strain transfer and the corresponding calibration procedures of resistance strain gauges [146]. Resistance strain gauges are mostly calibrated, verified, and applied according to VDE/VDI 2635 guidelines [147,148,149]. Inaudi et al. used double FBG sensors to measure the mechanical strain on the surface of an aluminum beam [150]. The results show that strain compensation can be quantified by introducing the strain delay factor into an FBG sensor during heating. Wolfgang R Habel proposed a verification method to evaluate the overall quality of sensor system functions to promote the use of optical fiber sensors [151].

As shown in Figure 7, according to the German VDI/VDE 2660 guideline, Roths et al. of Germany developed a strain calibration device, which can provide reference strain conditions with less uncertainty without using electrical strain gauges [152]. Through a series of calibration measurements of two different types of FBG strain sensor components and a specially developed bonding process, a calibration device based on four-point bending structure of stainless steel beams was established. The device can provide a reference strain condition within ±1000 μm/m with a relative uncertainty of about 0.3%. The results show that the relative standard deviation of calibration measurement for the same type of FBG sensor is about 0.16%. and there are significant differences in strain sensitivity between the two FBG sensors. The strain calibration device designed by researchers has a certain reference significance for improving the calibration accuracy.

Habel et al. developed a calibration test device for FBG sensors [153]. The unbiased digital image correlation (DIC) and electronic speckle interferometer (ESPI) are used for calibration measurement, and the resolution of DIC is 2 um and ESPI is 30 nm. The KALFOS (calibration of optical fiber sensors) facility designed by researchers consists of a load-bearing facility and a temperature chamber. In the experiment, the temperature and load changes are controlled by the equipment, and the strains in the materials, samples, adhesives and optical fiber sensors are analyzed objectively by a visual method. The experimental results show that the KALFOS method designed by the researchers is a non-contact method which can measure displacement and strain in all directions. Compared with traditional strain gauges, the KALFOS method can analyze their strain transfer mechanism, and can easily observe the relative motion between sensors and samples. The calibration device helps to validate the definition and quantitative description of component requirements in new or existing standards.

The International Electrotechnical Commission (IEC) adopted a new standard, IEC 61757-1-1:2016, which specifies the most relevant characteristics and characteristics of strain sensors based on FBG and its measuring procedures [154]. The fabrication of FBG sensor packages, the properties of materials and bonded joints can lead to the non-linear behavior of an FBG sensor, which can affect the accuracy of strain measurement. Lazarev et al. designed a nondestructive calibration method for an FBG strain sensor using a mechanical nano-motion sensor [155]. Researchers designed a customized calibration device, which uses a dovetail slideway mechanized by stepping motors. Through the analysis of experimental data, the performance of an FBG strain sensor is studied, the calibration curve of an FBG strain sensor is given, and the feasibility of this method is demonstrated. Święch proposed a calibration measurement system of wing load based on resistance and FBG strain gauge [156]. The data shows that the measurement error is between 12% and 40% according to the different measurement sections. A system of instruments embedded in the wing beam cover can achieve sufficient accuracy for measurement to evaluate the performance of the structure during operation. Święch proposed that because of the heaviness of the interrogator, the FBG sensor has more advantages than the electric strain gauges if a large number of FBG sensors are used. The method proves the feasibility of the calibration method based on FBG sensor, which does not simulate the test effect in a dynamic environment. Whether this method is suitable for the test of flight environment needs further verification.

NASA’s strain gauge calibration technology, the four-point bending calibration device under the German VDI/VDE standard, and the KALFOS non-contact calibration system have achieved differences in calibration principles, equipment, and accuracy for different sensors. FBG sensors based on temperature strain coupling require dual-grating compensation to achieve high calibration accuracy. Traditional strain gauges are influenced by VDI/VDE standards and are susceptible to external electromagnetic fields. Piezoresistive sensors are limited by material dispersion. There has been some research on the calibration methods for the above sensors, but there is a lack of horizontal comparative studies on different calibration techniques for sensors. In order to visually observe the calibration characteristics of various sensors, as shown in

Table 5. Comparison of calibration techniques for FBG sensors, traditional strain gauges, and piezoresistive sensors.

Calibration Technology Dimension	FBG Sensor	Traditional Strain Gauges	Piezoresistive Sensor	Ref.
Core calibration principle	Based on the relationship between Bragg wavelength and strain/temperature variation	Based on the relationship between resistance change and strain change	Based on the relationship between resistance changes and strain/damage changes	[157,158]
Mainstream calibration methods	Static calibration, dynamic calibration, temperature–strain coupling calibration	Static calibration, dynamic calibration	Static calibration, damage calibration	[159,160]
Calibration accuracy	High (static calibration error ± 1 με, dynamic calibration relative standard deviation ≈ 0.16%)	Middle (static error ± 5–10 με, affected by lead resistance)	Low (error ± 5–10%, affected by CNT dispersion and interface damage)	[161,162]
Environmental interference factors	Encapsulation stress release, fiber eccentricity. Temperature cross-sensitivity requires specialized compensation calibration	Vulnerable to electromagnetic interference and changes in lead resistance	Affected by humidity, CNT aggregation, and fiber matrix interface state	[163,164]
Calibration equipment complexity	High	Low	Middle	[165,166]
Feasibility of on-site calibration	Low	High	Middle	[167,168]
Long term calibration stability	High	Middle	Low	[169,170]

, the calibration methods and accuracy of FBG sensors, traditional strain gauges, and piezoresistive sensors (CNT/CFRP) were compared from the basic principles and methods of calibration.

Based on the review and analysis in this chapter, it can be concluded that onboard calibration is suitable for FBG sensors installed on aircraft wings, and can be completed in situ without disassembly, which is in line with the characteristics of aircraft wings that cannot be easily disassembled. However, due to the interference of flight environment vibration and wide temperature variation, the achievable measurement uncertainty is ±2–5 με. Temperature compensation requires the use of a dual grating layout, which can counteract the impact of temperature fluctuations on the wavelength of the wing surface at −40~85 °C in real time, and optimize the algorithm according to the “Dynamic Environment Compensation” clause in IEC61757-1-1 [171,172]. The laboratory bench calibration based on KALFOS equipment for four-point bending can accurately reproduce the typical force state of the wing, with good environmental controllability and low uncertainty. In addition, a constant temperature box can be used to precisely control the temperature, and temperature compensation can also be achieved through “single-point calibration + temperature coefficient correction”. The calibration process can be carried out according to the “static calibration process” of IEC61757-1-1, and the temperature compensation accuracy should be ensured to meet the requirements of mechanical property matching of materials such as CFRP.

Before calibration, it is necessary to confirm the packaging form of the FBG sensor and prepare a standard strain source according to of IEC61757-1-1. When calibrating the test bench, KALFOS equipment is used to simulate the 0–1000 με bending strain of the wing, record the wavelength strain curve, calculate the sensitivity coefficient, and synchronously test the compensation effect at temperatures of −50~150 °C. On time, the pilot selects non-critical force-bearing areas on the wing (such as wing tip auxiliary structures) as calibration reference points, uses dual grating real-time compensation, and compares the calibration data of the same period test bench to correct deviations. After calibration, according to IEC61757-1-1, “Calibration Effectiveness Determination”, ensure that the deviation of three consecutive measurements is ≤5%, and the wavelength drift after temperature compensation is 0.5 pm/°C, before it can be used for wing deformation monitoring.

6. Conclusions

On the basis of summarizing and analyzing the use of fiber optic sensing technology in aircraft wing deformation monitoring, this article provides a comprehensive introduction to the packaging form, bonding technology, and calibration methods of FBG sensors, and draws the following conclusions. In terms of packaging adaptability, tube-packaged FBGs can play a good role in vibration reduction and have excellent results in the long-term monitoring of the main load-bearing structure of the wing, but their maintenance is difficult. The integration of embedded FBG and CFRP composite wing materials makes internal strain measurement irreplaceable, with a strain transfer efficiency of 90% to 98%. Surface-attached FBG is the best choice for rapid retrofitting of existing wings due to its ability to achieve strain transfer efficiency of up to 99.2% by connecting the structure through metal welding. In terms of bonding technology, the thickness, width, and length parameters of epoxy adhesive directly affect the strain transfer efficiency. Its shear modulus is 3.2 GPa, which can effectively reduce the occurrence of strain hysteresis and has good stability. In terms of standardization of calibration technology, the four-point bending calibration device (VDI/VDE2660) based on the IEC61757-1-1 standard can achieve a static calibration error of <1 με. After calibration using the KALFOS non-contact calibration system, dynamic strain measurements with a relative standard deviation of ≤0.16% can be obtained. The use of dual-grating temperature compensation can offset the temperature fluctuations of −40–85 °C on the wing surface.

Based on the literature review and analysis mentioned above, this article suggests addressing the practical needs of aviation engineering. The CFRP composite wing should first be embedded with FBG with a diameter of 125 μm or less in the prepreg, and then embedded in the grating during the manufacturing process of the laminated board to match the thermal expansion coefficient of the grating with the composite material. It is recommended to use surface-attached FBG for aluminum alloy wings, and use ultrasonic welding technology to firmly weld the sensors on the wing skin with a thickness greater than or equal to 1 mm. The welding point diameter should be greater than or equal to 3 mm to ensure good long-term reliability. Sapphire fiber EFPI sensors can be used in the engine nacelle area, and temperature and pressure can be detected simultaneously through dual chambers. In terms of optimizing the bonding process, it is recommended to use a dispensing process to ensure the uniformity of the adhesive layer when using epoxy adhesive. In the surface treatment of bonding, the wing surface needs to be sandblasted first, and then cured under different temperature conditions to remove internal stress. During static calibration, a temperature strain coupling model can be established using a four-point bending device (VDI/VDE2660) produced in Germany. During dynamic calibration, the KALFOS system can be used to measure the sinusoidal strain excitation applied to the sensor, in order to verify its dynamic response characteristics and accuracy. During on-site calibration, single-point calibration should be conducted within 6 months, and the calibration work should be based on the laser interferometer.

The application of fiber optic sensing technology in the aviation field still needs continuous breakthroughs in the following areas. Due to the harsh flight environment of aircraft, it is recommended to develop new fiber optic materials that are resistant to high temperatures (>1500 °C) and radiation. It is recommended to combine fiber optic sensors with wireless transmission technology to achieve real-time data transmission and remote diagnosis. Sensor arrays for simultaneous measurement of multiple parameters should be researched to improve the integration of monitoring systems.