With the rapid development of modern aviation and aerospace, there is an increasingly urgent demand for real-time monitoring of multiple physical parameters in extreme environments such as high temperature, high pressure, and high-speed airflow^[1–3]. Accurately measuring high-temperature force–thermal distribution and achieving high spatiotemporal resolution visualization monitoring is of great importance for improving the design, health monitoring, and fault diagnosis of major equipment. For example, in the field of aviation engines, the temperature of the post-turbine gas can reach up to 1800°C. Excessive temperatures in the post-turbine gas can reduce the strength of the turbine blade’s metal material, increase the tip clearance, cause blade creep and erosion, and in severe cases, lead to structural damage of the engine^[4–6]. Therefore, real-time and in situ monitoring of the temperature field behind the engine turbine is required. In the field of hypersonic aircraft, the working environment inside the engine combustion chamber involves high temperature, high pressure, and high-speed gas flow. Under extreme flight conditions, it is easy to exceed the flame stability limit, leading to gradual extinguishing or even complete extinction of the flame inside the engine^[7–9]. Therefore, it is necessary to grasp the distribution pattern of high-temperature force–thermal on the combustion chamber wall.

Currently, surface-contact temperature sensors mainly include thermocouples and optical fiber sensors. Technology of thermocouple is mature and widely used, with a temperature measurement range of up to 2300°C^[10]. However, thermocouple temperature sensors suffer from common-mode noise interference, poor stability, and susceptibility to oxidation and erosion. Optical fiber sensors have unique advantages, such as high sensitivity, light weight, small size, resistance to electromagnetic interference, corrosion resistance, and distributed sensing^[11–13]. However, the stability of conventional silica fibers is relatively poor in ultrahigh-temperature environments, and the doping of germanium or rare-earth ions in fibers can cause thermal diffusion, which reduces the fiber transmission performance, making it unsuitable for stable operation above 1000°C^[14]. Sapphire fibers have been demonstrated to be a better candidate for ultrahigh-temperature measurements above 1200°C, due to the advantages of high melting point (∼2050°C), excellent transparency, and chemical corrosion resistance^[15–18], but they are susceptible and suffer from multimode problems. Developing new high-temperature sensors with a higher temperature measurement range, multiparameter sensing ability, erosion resistance, and distributed or quasi-distributed sensing capability has good prospects.

Compared with surface-contact sensors, embedded sensors can monitor critical positions inside the structure in real time. Metal-embedded optical fiber sensors have been extensively studied in the past decades due to the advantages of corrosion resistance and erosion resistance^[19–22]. Metal-embedded optical fiber sensors can enhance the fatigue resistance of fibers without affecting the integrity and the thermal protection performance of structural components, making them suitable for health monitoring of structures in extreme environments. The embedding method of optical fiber sensors includes ultrasonic additive manufacturing (UAM)^[23], laser additive manufacturing, casting^[24], electroplating^[25], etc.

In this review, we present the current research status of fiber Bragg grating (FBG) and Fabry–Perot interferometer (FPI) optical fiber high-temperature sensors, and summarize the progress of the encapsulation technique for optical fiber high-temperature sensors. We also discuss the extreme environmental adaptability of different encapsulation technique structures, and explore the critical technical issues that need to be addressed for optical fiber sensors used in extreme environments.

Optical fiber high-temperature sensors can be divided into FBG-type, FPI-type and blackbody radiation-based type, according to the sensing principle. This paper mainly introduces FBG-based and FPI-based fiber sensors. The sensitive element of the blackbody radiation-based fiber sensor is located in the front of the probe blackbody cavity^[26–29]; the optical fiber only is a light-guide medium, and its measurable dynamic range is small and has large error. Therefore, this paper does not expand to the introduction of blackbody radiation-based fiber sensors.

FBG is a wavelength-specific reflector that is formed by inducing a periodic modulation of the refractive index in the core of an optical fiber with ultraviolet (UV) light or femtosecond lasers. The Bragg wavelength λB is defined by the effective refractive index neff and the grating period Λ with the following relationship: λB=2neffΛ.(1)

By monitoring the Bragg wavelength change ΔλB as a function of temperature, the temperature sensitivity of FBG KT can obtained^[30], KT=dλdT=λB(α+ζ),(2)where α is the coefficient of thermal expansion (CTE) and ζ is the thermo-optic coefficient.

It has been found that there is a nonlinear relationship between the Bragg wavelength change and temperature in a wide temperature range^[31]. The wider the temperature range is, the more obvious the nonlinearity appears. A polynomial function of temperature T can also be used to show this relationship, ΔλB(T)=∑i=0naiTi,(3)where ai are polynomial coefficients.

It should be noted that FBG is inherently sensitive both to temperature and strain, as both neff and Λ vary with those two parameters, leading to the Bragg wavelength change. Therefore, solving the cross-sensitivity issue of temperature and strain is of great significance.

Since Hill et al.^[32] inscribed the first FBG in 1978, FBGs have been widely applied for temperature sensing in the past several decades. For conventional FBGs written by UV light, which are called Type-I FBGs, the refractive index will decay or even be erased when the temperature exceeds 300°C^[33], making it difficult to meet the requirement of high-temperature sensing. Several different methods, such as CO2 laser writing^[34], high-temperature annealing^[35], and femtosecond laser writing^[36], have been proposed to improve the upper temperature resistance limit of FBGs, making it possible to monitor temperature gradients in nuclear reactors and temperature distributions in gas turbines under high-temperature environments^[31,37,38].

Huang et al.^[34] fabricated a long-period fiber grating (LPFG) high-temperature strain sensor using CO2 laser induction and encapsulated it to achieve temperature sensing in the range of 700°C with large strain measurement in the range of 15,000 µε with a strain resolution of 0.5 µε.

Regenerated fiber Bragg grating (RFBGs) are produced by subjecting pretreated seed gratings (Type-I FBGs) to a thermal annealing process at high temperatures, typically in the hundreds of degrees Celsius. During this process, the reflectivities of the seed gratings gradually decay to almost zero and then regrow to a relatively low level, which is the origin of the term “regenerated.” To date, despite extensive research, there is currently no comprehensive explanation for the mechanisms behind the regeneration effect in RFBGs. However, two alternative theories have been proposed: the chemical composition theory and the crystallization theory^[39]. The chemical composition theory was proposed by Fokine et al.^[40] in 1997, where RFBGs are regarded as oxygen–chemical composition gratings that are formed at high temperatures. In the fiber core, the concentration of oxygen is periodically modulated in the illuminated and nonilluminated areas. In 2008, Canning et al.^[35] proposed crystallization theory, and they believe that it is the high temperature that reduces tensile core–cladding interface stresses and even inverts those stresses in the illuminated areas into compressive stresses, which causes crystallization and refractive index modulation in those areas. Both of these theories have been supported by various experimental observations, but neither one of them can fully explain all of the phenomena related to regeneration.

In 2015, Yang et al.^[41] fabricated two cascaded Type-I FBGs with center wavelengths of 1304 nm and 1547 nm, respectively, and obtained RFBGs by annealing. Simultaneous high temperature and strain measurements were achieved within the temperature and strain ranges of 900°C and 1000 µε. In 2020, Gunawardena et al.^[42] reported for the first time a resurgent regenerated fiber Bragg grating (R2FBG). An FBG was inscribed in a six-hole microstructured optical fiber (SHMOF), followed by high-temperature annealing to generate RFBG. The generated RFBG was rapidly heated, resulting in its disappearance at 1363°C and the regeneration of a new grating at 1405°C. Subsequently, the R2FBG was formed by rapid cooling to room temperature. Figure 1(a) illustrates the response of the peak power and wavelength shift of the RFBGs in SHMOF. Figure 1(b) shows the reflection spectra of the RFBG and the R2FBG at various temperatures throughout the heating process. It was demonstrated that this R2FBGs could be applied for ultrahigh-temperature measurements exceeding 1400°C.

Over the past 10 years, the suitability of RFBGs as temperature sensors for high-temperature applications has been investigated.

In 2015, Rinaudo et al.^[43] demonstrated the feasibility of RFBG application for temperature monitoring in the case of building fires. The RFBG sensor was mounted on concrete and tested for 1 h according to the ISO-834 fire curve. The sensor was subjected to direct flame and high-speed temperature rise of 200°C/min, and the highest gas temperature measured was about 970°C, demonstrating the feasibility of the RFBG sensor for real fire monitoring. The research results are instructive for fire protection engineering research and applications.

In 2018, Dutz et al.^[44] reported the application of RFBG arrays for high-temperature distribution measurements in chemical reaction vessels and gas turbine exhausts. 24-point RFBG and three-point RFBG arrays were used to measure the temperature change profiles of chemical catalytic reaction processes and the temperature gradients in gas turbine exhausts, respectively, and their maximum temperatures in both cases did not exceed 550°C. This study demonstrated that RFBG array-type temperature sensors can be used for high-temperature measurements in harsh industrial environments.

In 2018, Laffont et al.^[38] applied the RFBG array high-temperature sensor encapsulated in metallic capillaries to the temperature measurement of liquid sodium in a nuclear reactor, as shown in Fig. 2. The sensor measured the temperature gradient of liquid sodium during heating from 47°C to 500°C, with a response time of 144 ms. The research results demonstrate the feasibility of RFBG application in high-temperature and high-radiation nuclear reactors, and show the capability of RFBG for structural health monitoring in nuclear power plants and nuclear reactors.

In recent years, with the development of laser processing technology, ultrafast lasers, typically femtosecond lasers, have been widely applied for the inscription of FBGs in silica-based fibers^[45]. The high intensity of femtosecond laser pulses can cause physical damage to the fiber core, resulting in permanent refractive index modulation. The grating inscribed by the high-energy laser is called a Type-II FBG, which shows a temperature stability similar to that of RFBGs^[39].

In 2004, Martinez et al.^[46] first used the femtosecond laser point-by-point inscription method to inscribe first- to third-order FBGs in single-mode fibers (SMFs) and dispersion-shifted fibers (DSFs), respectively. In 2006, Grobnic et al.^[47] showed that Type-II FBGs exhibit remarkable thermal stability up to 1000°C for 150 h. A reflectivity decay and a permanent drift of Bragg wavelength were observed at temperatures exceeding 1050°C. In 2011, Li et al.^[48] annealed Type-II FBGs at 1000°C, and after the air quenching and the pre-stress treatment, those FBGs were able to operate stably for 26 h at 1200°C. In 2019, Warren-Smith et al.^[49] studied the high-temperature stability of Type-II FBGs using femtosecond laser etching on silica suspended-core fibers, where the sensor was stable at 700°C for approximately 300 h, as well as at up to 1100°C for 48 h. The thermally annealed sensor showed higher stability at temperatures less than 1050°C, with a sensor temperature drift of less than 0.5°C/day at 1050°C. The effect of long-term annealing on the high-temperature stability of the sensor was further investigated by Grobnic et al.^[50] in 2021, who annealed Type-II FBGs at 1000°C for several hundred hours, and the temperature drift of the annealed sensors was reduced to 0.03°C/h at 900°C.

Studies have shown that Type-II FBGs have excellent high-temperature stability, and they have been used for temperature measurements in the real industrial scenarios.

In 2013, Xia et al.^[1] monitored the circumferential and radial temperature distribution at the outlet of a gas turbine engine using a Type-II FBG array. After calibrating the sensor for the temperature range of 0°C to 650°C, the sensor was installed at the gas turbine engine exhaust, and its long-term thermal stability was monitored. The results showed that the sensor exhibited comparable performance to thermocouples with approximately 9°C fluctuations over 100 min.

In 2018, Walker et al.^[51] implemented the monitoring of temperature gradients over the combustion chamber flame tube using a Type-II FBG array encapsulated in a chromium–nickel (Ni) alloy tube, as shown in Fig. 3. Combustion tests showed that the sensor exhibited good thermal repeatability, withstood multiple thermal cycles, and sustained temperatures up to 900°C. The distributed Type-II FBG sensor can significantly reduce wiring compared to thermocouple temperature sensors, showing great advantages in aerospace applications such as gas turbine monitoring.

In 2018, Zaghloul et al.^[52] inserted the Type-II FBGs into a Massachusetts Institute of Technology (MIT) research nuclear reactor (6-MW, at a temperature>600°C) for sensing temperatures. The results show that the FBG sensors can survive extreme environments of the nuclear reactor cores under intense gamma and neutron irradiation at high temperatures.

However, it should be noted that silica-based optical fibers are limited by their material softening point, and it would be challenging for those sensors to maintain long-term stability in high-temperature environments above 1000°C^[14]. Even with processes such as quenching and annealing, the maximum measurable temperature can barely be achieved at 1400°C^[42].

Compared with silica-based optical fibers, sapphire fibers exhibit some superior properties such as high melting point (2053°C), high hardness, and chemical corrosion resistance^[53]. In recent years, they have attracted extensive attention in the field of high-temperature sensing. The multiphoton absorption also applies to other materials, so Type-II FBGs can be inscribed in sapphire fibers as well, which are called sapphire fiber Bragg gratings (SFBGs). There are typically three different methods for inscribing SFBGs: the phase mask method^[16], the point-by-point method^[54], and the line-by-line scanning method^[55].

In 2004, Grobnic et al.^[16] for the first time inscribed an SFBG using a 800 nm femtosecond laser and the phase mask method. The fifth-order grating with a pitch period of 2.14 µm was on a sapphire fiber with a diameter of 150 µm; its microscope image is shown in Fig. 4(a). Figure 4(b) shows the reflectance spectrum of the SFBG at room temperature. It can be seen that, due to the multimode characteristics of sapphire fibers, the reflected lights contain various modes, leading to a broad linewidth of the reflection spectrum. To solve this problem, Grobnic et al.^[56] in 2006 used the mode filtering effect to fuse a tapered SMF to a sapphire fiber to obtain an SFBG with a 3 dB bandwidth of 0.33 nm, enabling ultrahigh temperature measurements at 1500°C.

In 2009, Busch et al.^[57] further tested the temperature response of the SFBG in the range of 20°C–1745°C. When the temperature exceeded 1400°C, the signal-to-noise ratio (SNR) of the spectrum decreased significantly due to the increased blackbody radiation, and the reflection intensity of the SFBG decreased significantly, which indicated that the SFBG could not operate at high temperatures above 1400°C for a long time. In 2015, Habisreuther et al.^[58] measured the temperature distribution in a high-temperature tube furnace using an SFBG prepared by a femtosecond laser, as shown in Fig. 5. By moving the high-temperature furnace at a speed of 5 cm/min, the sensor measured the temperature distribution in the high-temperature tube furnace at 1500°C with a measurement error of less than 2°C. The authors also raised the temperature to 1900°C by filling the furnace with argon gas, and the SFBG still survived. Their research results show that SFBG has the potential for 1900°C temperature sensing.

In 2022, He et al.^[59] encapsulated SFBGs with a sapphire tube and inert gas, solving the high transmission loss issue caused by high-temperature oxidation. The schematic diagram of the encapsulation technique structure is shown in Fig. 6. The results demonstrated that the encapsulated sensor could operate stably for 20 h at 1600°C, with a maximum operating temperature of 1800°C.

SFBGs have demonstrated excellent performance in the field of ultrahigh-temperature sensing and have the potential to meet the high-temperature sensing requirements of complex force-heat environments such as aviation engines and hypersonic aircraft. However, there are still several issues that need to be solved in the practical application of SFBG, such as the connection difficulty between sapphire and silica fibers, the multimode behavior of reflection spectra, and the cross talk in multiparameter sensing of SFBGs.

The high-temperature sensing performance of different material fibers and different types of FBGs is summarized in Table 1. The term “Type” in Table 1 and in the rest of this paper refers to the underlying photosensitivity mechanism by which grating fringes are produced in the optical fiber^[60].

The sensitive element of the fiber FPI temperature sensor is the F–P cavity. When the ambient temperature changes, the refractive index of the fiber and the length of the F–P cavity change accordingly, due to the thermo-optical effect and the thermal expansion, which leads to a change in the optical path difference and causes a change in the phase, in turn leading to a wavelength drift in the interferometric spectrum. The sensing of external temperature is achieved by demodulating the interferometric spectrum. The wavelength of the interferometric trough of FPI can be expressed as λ=2nLm,(4)where n is the refractive index of the medium, m is the number of interference order, and L is the cavity length.

The temperature sensitivity is denoted as KT=ΔλΔT=(1LΔLΔT+1nΔnΔT)λ=(α+ζ)λ,(5)where Δλ is the change in wavelength, Δn is the change in refractive index, ΔT is the change in temperature, α is the CTE, and ζ is the thermo-optic coefficient.

The FPI temperature sensors are susceptible to external disturbances. For example, the external strain acting on the F–P cavity causes compression or stretching of the cavity length L, leading to a wavelength drift in the interference spectrum. As with FBG sensors, the cross-sensitivity of temperature and strain for FPI sensors is also an important issue needing attention.

FPI can be divided into the intrinsic Fabry–Perot interferometer (IFPI) and the extrinsic Fabry–Perot interferometer (EFPI), according to the structure of the sensitive element. Table 2 summarizes the performance parameters of IFPI and EFPI high-temperature sensors in recent years. Due to the thermo-optical effect, IFPI is more sensitive to temperature than strain or pressure^[67,79,80]. From Table 2, it can be seen that IFPI is often used for temperature sensing and less often for strain sensing. For the fabrication of fiber FPI high-temperature sensors, high-temperature resistant materials such as silica, sapphire, and silicon carbide are usually used. Currently, scholars have used high-temperature-resistant materials combined with femtosecond laser micro- and nanoprocessing^[81], micro-electromechanical systems (MEMS) processing^[82], and CO2 laser welding^[83] to develop high-temperature fiber FPI sensors.

Optical fiber FPI-type sensors were first developed in 1982 by Yoshino et al.^[84] and applied for sensing temperature, mechanical vibration, acoustic wave, AC voltage and AC and DC magnetic fields. In 1991, Lee et al.^[85] fabricated an IFPI temperature sensor by fusing a section of TiO2-coated fiber to an SMF to achieve 108°C temperature measurement. In 2008, Choi et al.^[86] made a compact FPI sensor by splicing a section of hollow-core fiber (HCF) and a section of SMF to PCF, and achieved high temperature sensing from 50°C to 1000°C. Benefiting from the development of femtosecond laser micromachining technology, in 2008, Wei et al.^[87] used a femtosecond laser to micromachine a groove in an SMF to form an EFPI temperature sensor, as shown in Fig. 7, and the cavity length of the sensor was measured to be 30 µm by scanning electron microscopy (SEM), with an ablation depth of ∼72μm. Its interference spectrum had an extinction ratio greater than 14 dB, and tests showed that the sensor could achieve temperature sensing up to 1100°C, with a sensitivity of 0.074 pm/°C.

In 2014, Kaur et al.^[72] micromachined an FPI microcavity on the end face of an SMF using femtosecond laser micromachining technology to achieve temperature sensing at 800°C, and performed a strain measurement at room temperature up to 3700 µε. In 2019, Lei et al.^[88] reported an IFPI high-temperature sensor based on hollow-core photonic crystal fiber (HC-PCF), the structure of which is shown in Fig. 8. It was formed by splicing a section of HC-PCF in between an SMF and a piece of pure silica. The results showed that the sensor could achieve high-temperature sensing up to 1200°C with excellent linearity and had a temperature sensitivity of 15.68 pm/°C.

The highest temperature measurement limit of silica-based optical fiber FPI temperature sensor is 1200°C^[88,89]. In order to further improve the temperature measurement limit, many scholars have proposed the FPI temperature sensor based on sapphire fiber^[90–92].

In 1992, Wang et al.^[93] fabricated an IFPI sensor by fusing a section of sapphire fiber to an SMF and applied the sapphire FPI sensor to high-temperature sensing for the first time, achieving a resolution of 0.2°C at 310°C–976°C. To further improve the upper limit of temperature measurement, in 2006, their team^[94] made an EFPI high-temperature sensor by bonding a 45° polished sapphire fiber to the surface of a sapphire wafer, which achieved temperature sensing up to 1170°C, with a resolution of 0.4°C. In 2019, Yu et al.^[95] proposed an EFPI high-temperature sensor based on sapphire fiber and sapphire wafer, as shown in Fig. 9. The sensor used a sapphire sleeve to fix two polished sapphire optical fibers with a sapphire wafer to form an F–P cavity structure, and the other end of the sapphire optical fiber was fused to a multimode fiber to achieve long-distance transmission. The sensor used two sapphire fibers to isolate the input and output so as to achieve the separation of background light and interference signal, which significantly improves the fringe contrast and SNR of the interference signal. Test results showed that the sensor could achieve temperature sensing up to 1080°C, and the temperature resolution was better than 0.25°C.

In 2020, Yang et al.^[96] fabricated an all-sapphire microfiber FPI high-temperature sensor by machining a cylindrical microgas cavity on the end face of a sapphire fiber using a femtosecond laser and welding a sapphire wafer to the end of the sapphire fiber using the CO2 laser fusion technique. The results showed that the sensor was capable of temperature sensing up to 1455°C, with an average temperature resolution of 0.68°C.

The high-temperature sensing performance of different material fibers and different structures of FPIs is summarized in Table 3.

At present, optical fiber high-temperature sensors are widely used in high-temperature sensing. Those real harsh environments, however, involve not only high temperature but also other factors, such as high pressure, high-speed air flow erosion, and corrosion. Therefore, it is necessary to encapsulate the sensors to enhance their adaptability and robustness. The encapsulation technique for fiber high-temperature sensors in this section can be classified into three kinds: tubular encapsulation^[97–102], substrate encapsulation^[103–106], and metal-embedded encapsulation^[107–112].

Tubular encapsulation is a cost-effective approach to improving the reliability and fatigue resistance of high-temperature sensors. In this technology, high-temperature-resistant tubing is typically used to encapsulate the sensor with single or multilayer protection. The resulting encapsulated sensor has good stability, erosion and corrosion resistance, and high mechanical strength, and is suitable for high-temperature measurements in various engineering applications.

Since the mechanical strength of RFBG decreases under high-temperature annealing and is prone to fracture after annealing, RFBG needs to be encapsulated and protected to meet practical applications. In order to improve the fatigue resistance of RFBG, in 2011, Barrera et al.^[113] used an alumina ceramic tube and a Ni alloy shell to encapsulate an RFBG sensor, as shown in Fig. 10. First, the seed grating was encapsulated in a composite structure with an inner ceramic tube and an outer Ni alloy shell using high-temperature ceramic glue, and then the encapsulated sensor was annealed at 1000°C to generate the RFBG. The tests on the encapsulated sensor showed that the encapsulated sensor could achieve temperature sensing up to 1100°C without exhibiting any hysteresis, and its time response was about 9 s, consistent with the performance of the corresponding thermocouples.

To prevent the formation of microcracks and devitrification on the fiber surface at high temperatures, in 2014, Mamidi et al.^[114] used five different tubular encapsulation materials: silicon carbide, borosilicate glass, stainless steel, copper (Cu), and aluminum nitride for Type-I FBGs, respectively. The effect of different encapsulation materials on the temperature response of the sensor was investigated in the temperature range of 20°C–500°C. The experiments showed that the response of the Cu-encapsulated fiber grating exhibited a slight nonlinearity, and the aluminum nitride encapsulation showed better linearity and faster response compared to the other three materials. In 2015, Mamidi et al.^[100] encapsulated a femtosecond laser-inscribed FBG in stainless steel and ceramic tubes. The Type-II FBG was first encapsulated in an aluminum nitride (ceramic) capillary tube to achieve strain isolation and then protected using a stainless-steel tube. The sensor probe was further inserted in a steel rod by making use of a brass holder and was sealed with a high-temperature adhesive. The sensor was used as a temperature probe with a measuring range of up to 650°C and a resolution of 1°C.

In 2018 Wilson et al.^[115] found that sapphire fibers generated some form of aluminum hydroxide on the fiber surface due to high-temperature oxidation in high-temperature environments at 1400°C and above, resulting in a significant increase in the transmission loss of sapphire fibers, which seriously affects their optical transmission performance. To address this problem, their team used an inert gas to isolate the sapphire fiber from the outside air and found that any surface layers generated by high-temperature oxidation of the sapphire fiber were completely eliminated at 1400°C. Their research results showed that the inert gas could inhibit the high-temperature oxidation of sapphire fiber. In 2019, Yang et al.^[116] fabricated an SFBG array by the point-by-point method, which was then encapsulated using a sapphire tube, as shown in Fig. 11. The encapsulated sensor was annealed at 1000°C for 110 h, and the calibrated sensor was used to measure the temperature field of commercial coal-fired and gas-fired boilers, achieving long-term stable testing up to 1200°C. In 2022, He et al.^[59] further raised the temperature limit up to 1800°C using sapphire tubes and inert gas to encapsulate SFBG.

It should be noted that the tubular encapsulation technique sometimes employs adhesives that might create air gaps between the protective tube and the optical fiber, leading to a loose fit and reduced structural strength of the sensor. This can cause the sensor to break easily in high temperature and high speed air flow scour environments, posing a significant risk of failure. In addition, for sapphire fiber, which has an air-clad waveguide structure, the refractive index of the adhesive must be lower than that of the sapphire to ensure effective total internal reflection of light through the fiber.

The tubular encapsulation technique can protect the optical fiber sensor, but it cannot be applied to the strain measurement on the structure surface. For this reason, scholars have proposed the substrate encapsulation technique. Substrate encapsulation involves attaching the optical fiber sensor to a substrate by welding or glue bonding, allowing the substrate to be welded to structural components. This technique offers several advantages, including high flexibility, good fatigue resistance, and compact structure.

In 1995, Inoue et al.^[117] first proposed the use of aluminum alloy sheet to encapsulate FBG and conducted temperature sensing tests, but the sensitivity enhancement was weak. In 1996, Gupta et al.^[118] successfully enhanced the temperature sensitivity of the sensor by encapsulating FBG on aluminum and polymer substrates with high CTE. The temperature sensitivity of the sensors encapsulated in the two materials was 2 to 4 times that of bare FBG, respectively. In 2005, Wnuk et al.^[119] bonded FBG to a metal substrate and introduced a residual stress of −1900με after curing of the adhesive. The sensor exhibited a linear response over the temperature measurement range of −20°C to 120°C and 1000 µε. To improve the temperature range of the sensor, their team^[120] subsequently bonded surface relief fiber Bragg gratings (SR-FBGs) to a metal substrate in 2006 and successfully achieved a high-temperature linear response up to 800°C. In 2009, Li et al.^[121] bonded a steel tube encapsulated FBG to a semi-cylindrical metal sheet to enhance temperature sensitivity and eliminate the effect of stress. The average temperature sensitivity in the range of −15°C to 200°C reached 18.64 pm/°C. In 2014, Tu et al.^[104] developed a Ni-coated RFBG fiber sensor encapsulated in a metal substrate, as shown in Fig. 12. The sensor was subjected to a uniaxial tensile test ranging from room temperature to 400°C. The results showed good strain linearity, stability, and repeatability. However, the sensor was found to be susceptible to temperature disturbance, resulting in large strain errors. Although the encapsulation structure exhibited excellent high-temperature creep and oxidation resistance, the sensor suffered significant thermal stress at high temperatures due to the significant difference between the CTEs of the optical fiber and the metal. In addition, due to the weak mechanical strength of the RFBG, the maximum operating temperature was limited below 400°C.

In 2016, Habisreuther et al.^[103] proposed a bonding encapsulation technique method for SFBG sensors. As shown in Fig. 13, the multimode SFBG high-temperature strain sensor was first encapsulated in an Inconel tube using high-temperature ceramic adhesive. Then the protective tube was bonded with high-temperature adhesive to a base steel plate with a V-shaped groove, in which the sensor was embedded. The steel plate can be welded at metal material structures to achieve high-temperature contact strain measurement. In this research, the sensor realized a temperature measurement in the range of 600°C and the strain measurement up to 1500 µε.

In 2021, Yao et al.^[105] designed a dual Type-II FBGs cascaded sensor with a stainless-steel substrate for high temperature, vibration, and strain measurement of structural components, the schematic diagram of which is shown in Fig. 14. The substrate bonding with the optical fiber sensor was securely fastened to the object with screws through four through-holes to enable contact measurement. FBG 1 was used for strain measurement; FBG 2 was used to measure temperature and vibration, and its measurement results were used to compensate for the temperature effect of FBG 1. The sensor was capable of measuring the three parameters of temperature, strain, and vibration in the range of 1100°C, 580 µε and 14 g, respectively. This sensor provides a viable solution for the engineering application of multiparameter sensing.

In summary, the substrate encapsulation technique first attaches the optical fiber sensor to the substrate by bonding or welding, and then attaches the substrate to the structural components by welding or tapped holes, thus enabling the monitoring of physical parameters of structural components. This encapsulation technique facilitates the measurement of actual physical parameters on the surface of structural components, providing high measurement accuracy and engineering applicability, However, the encapsulation technique process is achieved by bonding or welding, which cannot completely protect the entire fiber, and the mechanical properties of the bonding or welding area will be affected under high temperature, strong shock, and other environments, which may lead to sensor failure.

In recent years, embedding optical fiber sensors directly into parts of metal materials for measurement and integration into structural components have been reported^[112]. The encapsulation of embedded optical fiber sensors helps to avoid the risks associated with fiber sensors surviving in extreme environments and increases the fatigue resistance of optical fibers without compromising the integrity and thermal protection performance of the structural component. The metal-embedded encapsulation techniques can be mainly divided into three different methods: electroplating, casting, and metal 3D printing.

Electroplating is a low-cost and straightforward way for producing metal-embedded encapsulation, which can efficiently increase the thickness of the metal layer on the optical fiber surface and improve its fatigue resistance. To achieve a better fit between fiber and metal materials, the bare fiber needs to be surface-metallized to deposit a conductive layer before electroplating. Currently, the primary methods of fiber surface metallization include magnetron sputtering^[122–124], vacuum evaporation deposition^[125], and electroless plating^[126]. In electroplating, the metallized optical fiber and the plated pure metal are used as cathode and anode, respectively, so that the metal cations in the solution are deposited on the surface of the optical fiber to form the plating layer.

As early as 1979, Pinnow et al.^[127] applied a 15–20 µm thick aluminum coating to the fiber surface to improve the strength and fatigue resistance of the fiber. In 1989, Bubel et al.^[128] used vacuum evaporation deposition to coat the fiber surface with a Ti/Pt/Au multilayer metal coating to ensure the fiber was unaffected by environmental moisture. In 2000, Watson et al.^[129] used chemical plating to coat the fiber with a Ni-P alloy layer to achieve the metallization of the fiber surface. In 2006, Sandlin et al.^[130] first proposed a combination of electroless plating and electroplating techniques to chemically plate silver and electroplate Ni onto an FBG for effective protection and temperature sensitivity enhancement. This sensor can operate in harsh environments up to 600°C. Normally, the metallized FBGs are solderable and can be embedded into metal structures using brazing, laser welding, ultrasonic welding, and other welding methods to monitor the physical parameters inside the structures. In 2007, Sandlin et al.^[131] used a vacuum brazing method to weld Ni-plated FBG to Inconel 600 alloy, with the high temperature of the welding process reaching 900°C. The FBG survived after welding and was used for temperature measurement up to 600°C. It should be noted that the residual stress after welding was large. To eliminate possible damage to the FBG caused by the high temperature of the welding process, in 2008, Müller et al.^[132] used a soft-brazing method to encapsulate the metallized FBG on the metal structure using a tin-based brazing material at a temperature of 200°C to achieve the structural health monitoring.

In 2009, Li et al.^[109] used a combination of electroless plating and electroplating to fabricate a Ni-coated FBG temperature sensor, as shown in Fig. 15. Temperature measurements of Ni coatings with different thicknesses were performed in the range of 20°C to 300°C. The results showed that, when the coating thickness was relatively thin, the temperature sensitivity of the FBG increased rapidly with the thickness of the Ni coatings, and gradually reached a relatively stable value. In addition, the mechanism behind the increase in temperature sensitivity of FBG after metallization was theoretically analyzed, and a mathematical model of the FBG after metallization was established.

In 2014, Zheng et al.^[111] used a magnetron sputtering and electroplating method to create a 22.5 µm thick iron-carbon (Fe-C) film on the surface of an FBG. The sensor was capable of monitoring corrosion for over 40 days. In 2021, Wang et al.^[133] used a combination of titanium-copper (Ti-Cu) magnetron sputtering and Cu electroplating to fabricate a Ti-Cu-coated FBG temperature sensor, as shown in Fig. 16. The authors performed multiple thermal cycles in the temperature range of 79 K–293 K and conducted fatigue tests on the sensor, which failed after 560 cycles. This sensor has potential for temperature measurement in cryogenic engineering.

In 2023, Liang et al.^[134] developed a high-temperature strain sensor composed of a Ni-coated FBG-FPI using electroless plating and electroplating, as shown in Fig. 17. The surface metallized fiber sensor equipped with an FBG cascaded, and an air bubble F–P cavity could measure temperature and strain simultaneously in the range of 0°C to 545°C and 0 to 1000 µε. The cascaded structure reduced the impact of cross sensitivity of temperature and strain. The metallized optical fiber not only improves the fatigue resistance of the sensor but also enables easy integration with structural parts, thereby expanding the application range of the optical fiber sensor in structural health monitoring.

Optical fiber sensors can also be encapsulated or embedded into metal structures by the casting process. This method involves placing optical fibers in a casting mold prior to the casting process, where the fibers will be completely surrounded by molten metals during casting and later embedded into the solidified metal structures after cooling. This method allows for the embedding of optical fiber sensors deep within the structure, which improves their environmental adaptability while maintaining the integrity and thermal protection performance. Therefore, the embedded casting method has broad application prospects in the fields such as aviation, automotive industry, and construction, which require complex metal structures.

In 1991, Lee et al.^[135] for the first time encapsulated an FPI sensor with an aluminum casting process, which could be applied for monitoring and controlling a variety of mechanical systems such as auto and aircraft, building structures, or the power plant industry. In 1998, Lin et al.^[136] encapsulated an FBG sensor with lead (solder) cladding with a casting process in an aluminum U-groove; the sensitivity of the encapsulated FBG was enhanced by ∼5 times that of a bare FBG.

Research groups from Technical University of Munich and Munich University of Applied Sciences have used FBG/RFBG sensors to monitor the high-temperature metal casting processes in the past several years. Those FBG/RFBG sensors, in direct contact with metal, could survive after the casting process with the maximum temperature of 650°C and compressive strain—10,000 µε for aluminum^[31,137] and 1100°C and 14,000 µε for Cu^[112]. The cast parts with embedded optical fibers were further manufactured into standard tensile test specimens to achieve tensile tests, and simultaneously temperature and strain measurement^[138]. Figures 18(a) and 18(b) show the physical image and the schematic diagram of the tensile test specimen embedded with FBG sensors. Figure 18(c) shows the embedding condition of the optical fiber sensor. It can be seen that the fiber was fully surrounded by the aluminum without an obvious gap or void.

While encapsulating optical fiber sensors by the casting method has apparent advantages, such as the excellent protection due to the deep burial and the potentials for mass manufacturing and fabrication of complex structures like turbines and engines, it also presents certain challenges. The high melting points of metals, impurities in the melt during casting, and the significant contraction that occurs during the cooling process are just some of the obstacles that need to be overcome.

Metal 3D printing technology can produce metal parts with complex shapes and internal structures that meet the requirements for embedded encapsulation of optical fiber sensors in various environments, including high temperature, high pressure, high-speed air scouring, and other extreme conditions. Metal 3D printing methods described in this paper mainly include laser additive manufacturing^[139–141] and UAM^[142–145].

In 2002, Li et al.^[146] first reported the use of laser-assisted shape deposition manufacturing (LASDM) method to embed electroplated Ni FBGs into stainless-steel structures and successfully performed 0 to 1500 µε strain and 0°C to 400°C temperature measurements using embedded FBG sensors. The results showed that the temperature sensitivity of the embedded FBG sensors was twice that of a bare FBG. This report demonstrates the feasibility of applying this technique for the embedded encapsulation of optical fiber sensors. Their findings have inspired many scholars to use different additive manufacturing processes to realize the embedding of optical fiber sensors. In 2011, Alemohammad et al.^[147] used laser solid freeform fabrication (LSFF) method to deposit tungsten carbide-cobalt (WC-Co) layer on a Ni-electroplated FBG to achieve embedded encapsulation, and the embedded sensor maintained a linear response in the temperature range of 32.2°C to 121.7°C. In 2015, Havermann et al.^[24] used selective laser melting (SLM) to embed the FBG with a thin Ni layer (∼350μm diameter) into stainless-steel 316 components; the embedded FBG exhibited a linear response in the range of 100°C to 400°C and was tested for 500 cycles of strain, a strain sensitivity of ∼0.92pm/με was determined. In 2017, Zou et al.^[148] used a laser-engineered net shaping (LENS) process to embed optical fiber sensors into Ti-6Al-4V assemblies. In 2017, Mathew et al.^[141] used the SLM method to fabricate a high-temperature sensor consisting of an embedded metal-coated optical fiber FPI sensor, as shown in Fig. 19. The optical fiber FPI sensor was inserted into a capillary glass tube using soldering, and then silver and Ni plating was applied to allow metallization of the protective tube. Finally, the Ni-plated capillary was embedded into the SS316 stainless-steel substrate using the SLM method to achieve embedded encapsulation of the sensor. The sensor achieved the high-temperature test at 1000°C, with an accuracy better than ±10°C.

In 2020, Lei et al.^[140] embedded an optical fiber IFPI temperature sensor into an Al2O3 substrate by CO2 laser sintering, as shown in Fig. 20. The IFPI sensor was inserted into the substrate, and the CO2 laser sintering process was then used to fill the Al2O3 slurry, resulting in the compact encapsulation of the embedded sensor. Test results showed that the embedded IFPI sensor had long-term stability and could operate continuously for 17 h at 800°C, with a deviation of less than 10°C.

In 2022, Kim et al.^[149] used a directed energy deposition (DED) process to embed a Ni-plated FBG sensor into a miniature turbine blade and monitored the structural integrity and functionality of the Ni-plated FBG sensor in situ during fabrication. The schematic diagram and the physical image of the miniature turbine blade with the Ni-FBG sensor embedded are shown in Fig. 21. The miniature turbine blade was tested for three temperature cycles in the range of 100°C to 500°C. The sensor exhibited excellent linearity, and the thermal sensitivity of the embedded FBG sensor was well-matched, with no slippage throughout the temperature monitoring. The results strongly demonstrate that the embedded optical fiber sensor is capable of high-temperature in situ monitoring of structural components and has promising applications for structural health monitoring of large structures.

The kind of laser additive manufacturing described above for embedding sensors in matrices requires operation at high temperatures, which poses a risk of transient high-temperature damage to optical fiber sensors.

UAM, a solid-state additive manufacturing technology, was invented in 1999 by Dawn White^[150]. UAM is a technique for creating solid metal objects by ultrasonically joining a series of metal foils into a three-dimensional structure^[151]. The UAM process does not require the melting of the metal, allowing the insertion of optical fiber sensors into matrices at temperatures below 25% of the melting point of the matrix^[152], thus reducing a negative impact on sensor performance due to high temperature.

In 2012, Li et al.^[153] successfully embedded Ni-plated FBGs into aluminum foil using ultrasonic welding (UW) processes, and the embedded FBGs were tested at 20°C–300°C and 0–40 N tensile load on the aluminum foil; the results showed linear responses. In 2016, Zhang et al.^[154] achieved the embedded encapsulation of FBG sensors in metals using UW technology. As shown in Fig. 22, the optical fiber sensor was first placed on an aluminum alloy substrate, and then used UW technology to embed the sensor into a molten indium alloy. The sensor can independently measure temperature and strain in the ranges of 200°C and 2000 µε, and shows good linearity and stability.

In 2017, Schomer et al.^[155] used the UAM method to embed an acrylate-coated FBG into aluminum 6061, and test results showed that the embedded FBG sensor could accurately track strain at temperatures up to 450°C. In 2019, Chilelli et al.^[23] used UAM to embed FBG into various base specimens to fabricate temperature and strain sensors, respectively, as shown in Fig. 23. The temperature sensors were repeatedly operated at 300°C, and the strain sensors demonstrated that the embedded FBG sensors were capable of crack detection.

In 2019, Petrie et al.^[156] used UAM to successfully embed Cu-coated, aluminum-coated, and Ni-coated fibers into an aluminum substrate, as shown in Fig. 24. After embedding, the three types of fibers showed good survivability at 500°C; embedding Cu-coated fibers in Cu allows for higher temperature operation.

In 2022, Hyer et al.^[157] used UAM to embed optical fiber sensors into pipes and hexagonal assemblies made of stainless steel SS304 used for nuclear reactors, which were tested for distributed temperature and strain, demonstrating the potential of embedded optical fiber sensors for spatially distributed health monitoring during microreactor operation.

UAM has the advantage of low temperature manufacturing and avoids the risk of sensor failure due to high temperatures during the metal embedding encapsulation of optical fiber sensors compared to laser additive manufacturing and casting methods. However, the temperature resistance of sensors fabricated by UAM technology needs to be further improved to meet the requirements of extreme environments.

Table 4 summarizes the development of metal-embedded optical fiber sensors and lists the temperature measurement performance of embedded sensors and their different application scenarios.

Optical fiber FBG-type and FPI-type high-temperature sensors respond to both temperature and strain; therefore, when those sensors are used to monitor the temperature of structures in high temperature environments, the presence of thermal stress might affect the temperature measurement results, causing large measurement errors. At present, there are two main solutions: the dual-parameter matrix method and the temperature compensation method. The dual-parametric matrix method uses a single or multiple structures with different temperature and strain response characteristics to differentiate temperature and strain. Examples include dual Type-I FBG cascade^[158], dual FPI cascade^[81], dual RFBG cascade^[41,159], Type-I FBG cascaded FPI^[134], RFBG cascaded FPI^[160,161], Type-II FBG cascaded FPI^[162], cascaded dual-core PCF^[163], and dual LPG cascade structures^[164]. However, normally, such a dual-parameter matrix can only estimate the performance of the measurement system in a limited temperature range. In high-temperature environments, the strain sensitivity of optical fiber sensors usually varies with the ambient temperature, and the use of the dual-parameter matrix method to demodulate temperature and strain might result in large measurement error^[73,165]. To address this problem, some scholars have proposed an iterative matrix algorithm^[166,167], which greatly improves the accuracy of temperature and strain decoupling and the systematic error caused by the temperature dependence of the strain sensitivity. The temperature compensation method^[168,169] is used to insulate one of the sensor structures from strain by encapsulation so that it only provides temperature information that can be used to compensate the temperature influence on the other structure. In addition, machine learning^[170–172] has also been used to solve the problem of cross-sensitivity, and this method can achieve the decoupling of temperature and strain with a single sensor structure, reducing the cost and complexity of the sensor system. However, its limitation is that the demodulation accuracy is not high and needs more computational power.

Due to the significant material differences in characteristics between metals and silica, there are mismatches in their combination. As mentioned in the previous section, there might be large gaps or defects between the metal and the embedded fiber, which greatly affects the encapsulating effect and the strain transfer between metals and optical fibers. Metal casting can achieve a better embedding condition. However, due to the significant temperature changes during the casting process and the huge difference in CTEs between the metal and silica, this process often introduces significant compressive strain, typically exceeding −10,000 με, under which ordinary optical fibers cannot survive after exposure to high temperatures. To solve this problem, optical fibers with larger dimensions and higher mechanical strength are used^[112,137]. Additionally, it should be noted that during some encapsulation technique processes, high-temperature adhesive is used for bonding, and the mismatch in its CTE can also lead to a deterioration in the final encapsulating effect^[173,174].

In recent years, extensive research about ultrahigh-temperature sensors based on sapphire fibers has been carried out. It is foreseeable that embedding sapphire fibers into metals can largely improve the upper temperature limit of the corresponding sensors. Despite their high-temperature tolerance, sapphire fibers suffer from intermode interference due to their air-clad and multimode structure. Excitation of higher-order modes will cause the SFBG reflection peak bandwidth to enlarge, and reduce the SNR and contrast of interference fringes, which seriously affects the measurement accuracy of the sensors. Therefore, reducing the number of modes in sapphire is crucial for the development of practical sapphire fiber high-temperature sensors. To address this issue, many scholars have done a great deal of research on sapphire-derived fiber (SDF)^[175–178] drawing, sapphire fiber waveguide Bragg grating^[179,180] fabrication, and mode field matching fusion splicing^[181,182].

In 2012, Dragic et al.^[175] first proposed the scheme of SDF, i.e., the SDF with a high concentration of alumina-doped glass core and silica cladding was fabricated by a drawing tower, which solved the problem of excessive modes in the sapphire fiber. In 2014, Elsmann et al.^[176] inscribed Bragg gratings in SDFs using femtosecond laser processing and proved their high temperature resistance up to 950°C.

In 2017, Blue et al.^[177] proposed a nearly single-mode sapphire fiber with a cladding layer and conducted high-temperature testing up to 1500°C. The sapphire fiber was placed into a ruby tube, and the gap was filled with lithium-6-rich lithium carbonate (Li2CO3) powder before subjecting it to irradiation in a nuclear reactor. The resulting sapphire fiber had a cladding layer, as shown in Fig. 25. The introduction of the cladding layer enabled the sapphire fiber to achieve quasi-single-mode transmission, and the layer remained intact at temperatures up to 1500°C.

In 2019, Zhang et al.^[74] fabricated a fiber-optic FPI sensor using a high-aluminum oxide-doped silica cladding SDF, and demonstrated high-temperature sensing up to 950°C and strain sensing up to 1000 µε, as shown in Fig. 26. The introduction of SDF enhances the fringe contrast of IFPI.

In 2019, Liu et al.^[92] used arc discharge to generate mullite particles in the SDF core region to develop SDF-FPI sensors that can withstand high temperature up to 1600°C and can operate stably at 1200°C for 6 h. In 2021, Guo et al.^[178] fabricated a near-single-mode FBG with a 3 dB bandwidth of 0.33 nm on SDF and demonstrated 1000°C high-temperature sensing and up to 1100 µε strain sensing.

Currently, research shows that SDF could achieve near-single-mode transmission, but the silica cladding material limits its upper temperature measurement. The arc discharge-based SDF-FPI sensor can withstand temperatures up to 1600°C^[92], but can only operate stably at 1200°C. Although SDF can be manufactured from sapphire fibers with a reduced refractive index of the cladding through irradiation modification technology to achieve quasi-single-mode operation, this method requires the use of nuclear facilities such as neutron sources. Additionally, some scholars have explored using high-temperature resistant materials such as magnesium aluminate (MgAl2O4)^[183], boron nitride (BN)^[184], zirconium dioxide (ZrO2)^[185], and silicon boron carbon nitride (SiBCN)^[186] as the cladding of sapphire fibers, but due to the higher refractive index of these materials compared to sapphire, the difference in CTE with optical fiber is too large, and the thermodynamic properties are not stable at high temperatures^[187].

In 2021, Guo et al.^[179] made some breakthroughs in the development of quasi-single-mode sapphire fibers. The research group inscribed a helical sapphire fiber Bragg grating (HSFBG) inside the sapphire fiber using a femtosecond laser. The HSFBG, coupled with a multimode fiber, achieved a reflectivity of 40%, a full width at half-maximum (FWHM) of 1.56 nm, and an SNR of 16 dB. The sensing capability at high temperatures below 1600°C was also demonstrated, with a temperature sensitivity of up to 35.55 pm/°C, as shown in Fig. 27. Helical gratings have structural symmetry and a large refractive index modulation region, thus reducing the excitation of higher-order modes by the increased length of the sapphire fiber and the presence of surface defects. This study indicates that the helical fiber grating can suppress the transmission of high-order modes, thereby improving the system’s SNR. However, a multimode fiber is still required to obtain stable spectra.

In 2022, Wang et al.^[180] created a depressed cladding waveguide (DCW) on a sapphire fiber and inscribed a ring-shaped FBG inside it, resulting in a single-mode SFBG with a reflection peak bandwidth of less than 0.5 nm. The FBG could survive even after annealing at 1000°C, as shown in Fig. 28. The DCW structure enabled single-mode transmission on both bulk sapphire and sapphire fibers. This study provided a technological framework for the development of single-mode sapphire fibers based on femtosecond laser processing, but its insertion loss is greater than 6.8 dB/cm, so this method can only be used to manufacture single-mode SFBG with centimeter-level length.

In 2022, Jones et al.^[181] made some progress in developing a single-mode sapphire fiber system by using direct fusion splicing technology to control the tapering of the fusion zone and achieve filtering of higher-order modes in the sapphire fiber. This has resulted in the conversion of the sapphire fiber system to a single-mode one. A physical image of the fusion splice region is shown in Fig. 29. These findings provide valuable insights into the development of single-mode sapphire fiber systems.

In cooperation with Shenzhen University, recently, our research group^[188] has proposed a fusion splicing method based on mode field matching to suppress the high-order modes in SFBGs. An overlapped double conical splicing structure between sapphire fiber and SMF was formed, as shown in Fig. 30(a). A quasi-single mode reflection spectrum from SFBG was obtained based on this method with an FWHM of 0.3 nm and an SNR of 8 dB. Further improvement has been made on this fusion splicing method by using a pretapered SMF; the SNR reached 14 dB, as shown in Fig. 30(b). In addition, the SFBG sensor was further combined with a commercial single-mode interrogator and achieved a real-time temperature monitoring up to 1160°C.

Despite considerable progress made by many research groups in the field of single-mode transmission of sapphire fibers, there is still no SFBG sensor with quasi-single-mode spectrum that can withstand testing above 1300°C. Research on SFBG sensors remains at the experimental stage, and commercial products suitable for practical use in extreme environments have yet to be reported.

This paper presents the research progress of FBG-type and FPI-type high-temperature sensors. Temperature measurement capability and defects of silica-based fiber high-temperature sensors have been reviewed, and temperature sensing performance of sensors based on sapphire fiber has been summarized. The research progress of the encapsulation technique for fiber high-temperature sensors is comprehensively discussed. Different encapsulation technique methods are compared and analyzed in terms of their ability to ensure sensor consistency and enhance extreme environment adaptability.

Metal-embedded fiber sensors are found to be capable of monitoring the structural health of high-end manufacturing equipment such as engines under extreme conditions while enhancing fiber fatigue resistance without compromising structural integrity and thermal protection. Despite the significant breakthroughs achieved in the upper temperature limit and encapsulation technique aspects of fiber sensors, there are still many technical challenges that need to be addressed, such as the cross sensitivity between temperature and strain, the material properties mismatch between metals and optical fibers that might lead to insufficient sensor reliability, and how to suppress sapphire fiber high-order modes to obtain SFBG sensors that can be applied in practical environments.