Fiber optic sensors have been widely studied in academia due to their resistance to electromagnetic interference, corrosion resistance, and other characteristics. Fiber optic-based sensors include Mach-Zehnder interferometer (MZI) [[1], [2], [3]], the fiber Bragg grating (FBG) [4], and the long-period fiber grating (LPFG) [5]. Therein, LPFG boasts advantages, e.g., simple structure, easy fabrication and good repeatability. These qualities have led to its widespread application in measuring various physical parameters, e.g., temperature, refractive index, bending, strain, torsion. LPFG based sensors can generally use wavelength-modulation and intensity-modulation [6,7]. LPFG has the problem of high optical path loss when using intensity-modulation. Due to the unstable lateral distribution of optical fibers in LPFG, especially when the structure is complex, the optical path loss will be more significant. The use of wavelength-modulation can effectively solve the problem of optical path loss. Torsion sensors based on LPFG wavelength-modulation have been widely used in structural health monitoring [8,9]. Structural health is assessed by observing the amount of change in resonant wavelength caused by torsion. The amount of change in the resonant wavelength represents the torsional sensitivity, which currently needs to be further improved for LPFGs.

To improve the sensitivity of torsional sensing, many research teams have proposed fabrication methods and special structures based on LPFG torsional sensors. In 2023, Song et al. proposed a twist sensor using helical twin-core fiber (HTCF), which could simultaneously measure twist angle and direction, achieving a twist sensitivity of 0.242 nm/rad/m [10]. In 2024, Gui et al. introduced a twist sensor based on polymer optical fiber (POF), with a twist sensitivity of 3.3 p.m./rad/m [11]. These methods improve twist sensitivity by using specialty fibers but come with relatively high fabrication costs. Using single-mode fiber (SMF) to fabricate twist sensors could reduce the cost. In 2017, Dong et al. proposed to use a femtosecond laser to fabricate LPFGs by unilaterally grooving in SMFs, achieving a torsional sensing sensitivity of 181.7 p.m. (rad/m) [12]. At the same year, Sun et al. proposed segmented grooving, i.e., a method that introduces line birefringence through grooving at different datums, to fabricate segmented-long period fiber grating (S-LPFG). The S-LPFGs exhibited a torsion sensitivity of 0.3 nm/(rad/m) [13]. The underlying idea behind all the methods in this category is to enhance sensitivity by asymmetrically modulating the refractive index on one or more sides of a conventional SMF. Although the fabrication process is straightforward, this type of scheme does not effectively introduce twisting force within the optical fiber, thus the torsional sensitivity requires further improvement. In 2019, Liu et al. proposed twisting the SMF with fiber fusion splicer to prepare periodic helical ultra-long period fiber grating (U-LPFG) torsion sensor with a torsional sensitivity of 0.085 nm/(rad/m) [14]. In 2021, Bai et al. used a Fujik-ura fusion splicer for spiral processing of SMFs to create helical long-period grating (H-LPFG), increasing its torsional sensitivities reaching up to 0.138 nm/(rad/m) [15]. Liu and Bai enhanced the torsional sensitivity by using a fusion splicer discharge to cure the torsional force simultaneously, which introduces circular birefringence and thus achieves enhancing sensitivity.

Although the scheme is inexpensive, it does not utilize a laser for asymmetric refractive index modulation, so it is potential for further improving the torsional sensitivity. In 2020, Sun et al. fabricated a long-period fiber grating in helical polished structure (HPLPFG) torsion sensor showing a torsional sensitivity of 0.199 nm/(rad/m) [16]. A sensitivity of 0.15 nm/(rad/m) was achieved by intermittent-spiral long period fiber grating (IS-LPFG) torsion sensor [17]. In 2023, Lu et al. first pre-twisted the SMF by 540° via an intermittent helical structure, and then inscribing a V-groove with a CO2 laser to form a similar helical structure. Its torsion sensitivity is 0.654 nm/(rad/m) [18].

All of these approaches involve first applying a pre-twist force and then using a CO2 laser to carve the grooves, which fixes the twist and increases the linear birefringence. This method fixes the twist and increases the linear birefringence. Although this type of approach achieves the simultaneous introduction of linear birefringence and torsion forces, the fabrication process becomes complicated when using a laser to groove the sensor. Grooving the sensor in a twisted state reduces its mechanical strength and affects its stability. Moreover, all of these methods modulate the refractive index of certain regions of the SMF cladding, leading to insufficient twisting force during the solidification process. If the entire SMF is heated and twisted by using a hydroxide flame, the refractive index of the cladding is then modulated using a CO2 laser. This not only further increases the torsional force within the cured fiber, but also enhances the line birefringence and the torsional sensitivity of the sensor.

In this paper, we demonstrate a long-period fiber grating torsion sensor based on rotating SMF (R-LPFG). A hydroxide flame is first used to process a conventional SMF into a rotating SMF (R-SMF) with a significant amount of twisting force, which introduces a large amount of circular birefringence within the R-SMF. Then, a CO2 laser is used to etch notches on the surface of the R-SMF, introduce line birefringence. The combination of circular birefringence and line birefringence results in a significant number of elliptical birefringence, which significantly increases the resonance wavelength shift and further enhances the torsional sensitivity. It torsional sensitivity is improved by more than 20 times compared with direct etching of grooves on conventional SMFs [19]. In addition, the sensor fabrication process is simple, exhibiting high repeatability and high mechanical strength. It has great potential for application in engineering structural health monitoring and other fields.