为了解决随机线性双折射对光纤磁光特性测量的影响，采用旋转光纤（SF）结合法拉第旋转镜（FRM）的测量方法，研究了FRM对旋转光纤磁光特性测量的影响。首先，从理论方面研究旋转光纤与FRM的引入如何减小光纤中的随机线性双折射对磁光特性测量的影响，并搭建基于FRM的旋转光纤磁光特性测试系统。当光源波长为1310 nm时，FRM作用前的旋转光纤费尔德常数都比未旋转光纤的大，且旋转光纤的节距越短，费尔德常数越大。特别是旋转光纤的节距为1.0 mm时，其费尔德常数为0.8304 rad/（T·m），比未旋转光纤的费尔德常数［0.8029 rad/（T·m）］增大了约3.43%。当测试系统加入FRM后，不同光纤的费尔德常数测量值相较于未使用FRM的光纤费尔德常数测量值都有一定幅度的增大，尤其相比于节距为1.0 mm时的旋转光纤更进一步提高了7.50%，并且在FRM作用前后不同光纤费尔德常数测量值的均方差分别为0.99%和0.61%，说明FRM的引入提高了掺杂光纤费尔德常数的测量精度与稳定性。

Results and Discussions In the case of no FRM, the extinction ratio of the spun fibers with different helical pitches is greater than that of the unspun fibers (Fig. 3). It is shown that the circular birefringence introduced by the spun fiber can suppress the partial random linear birefringence in the spun fiber, and its extinction ratio increases as the pitch decreases. When the FRM is applied, the measured extinction ratios of the spun fibers with different pitches are relatively stable. The introduction of FRM can significantly offset the random linear birefringence effect in the spun fibers, thereby improving the stability of the extinction ratio test. When the light source is 1310 nm, the Verdet constant of the spun fibers without FRM is larger than that of the unspun fibers, and a shorter pitch of the spun fiber is accompanied by a larger Verdet constant. In particular, when the pitch of the spun fiber is 1.0 mm, the Verdet constant is 0.8304 rad·(T?m)^-1, which is about 3.43% higher than that of the unspun fiber [0.8029 rad·(T?m)^-1]. When the FRM is added to the testing system, the measured Verdet constants of different fibers are improved to a certain extent compared to those of the fibers without FRM (Fig. 5). For the spun fiber with a pitch of 1.0 mm, it is improved by 7.50%. In addition, the mean square deviations of the measured Verdet constants of different fibers without and with FRM are 0.99% and 0.61%, respectively. This indicates that the introduction of FRM improves the measurement accuracy and stability of the Verdet constants of spun fibers.

Due to the advantages of high sensitivity and electromagnetic interference resistance, the fiber-optic current sensor based on the Faraday effect has received extensive attention and is one of the research focuses in current measurement. However, there is high random linear birefringence inside the sensing coil, which seriously affects the sensitivity of the sensor. The spun fiber is widely used in the field of current sensing to overcome the influence of random linear birefringence on sensitivity. Moreover, the method of adding a Faraday rotator mirror (FRM) to the end of the sensor's induction coil can also be used to analyze the effect of random linear birefringence. An FRM induces light reflection and polarization rotation of 90° so that the outgoing light is orthogonally polarized to the incoming light. Polarization modulation can be eliminated when orthogonally polarized light undergoes reciprocal birefringence. To remove the influence of random linear birefringence on the measurement of fibers' magneto-optical properties, we prepare spun fibers with four helical pitches and unspun fibers and build an FRM-based testing system for magneto-optical properties of fibers.

The optical fiber samples used are homemade low-birefringence spun fibers. During the fiber drawing process, the fiber preforms are rotated at 1000 r·min^-1, 667 r·min^-1, 500 r·min^-1, and 333 r·min^-1, and spun fibers with a pitch of 1.0 mm, 1.5 mm, 2.0 mm, and 3.0 mm are obtained, respectively. The wavelength of the light source used in the experiment is 1310 nm. A magneto-optical property testing system based on FRM is built to accurately measure the extinction ratio and Verdet constant of fiber samples. Specifically, the extinction ratio characteristics of spun fibers with different helical pitches and unspun fibers without and with the FRM are measured. Afterward, the Faraday rotation angles and Verdet constants of spun fibers with different helical pitches and unspun fibers without and with the FRM are tested and compared with the theoretical values.

In the case of a light source of 1310 nm and no FRM, the Verdet constant of the spun fiber is larger as its helical pitch decreases. In particular, when the pitch is 1.0 mm, its Verdet constant [0.8304 rad·(T?m)^-1] is about 3.43% higher than that of the unspun fiber [0.8029 rad·(T?m)^-1]. When an FRM is applied, the measured Verdet constants of different fiber samples are improved to a certain extent, especially for the spun fiber with a pitch of 1.0 mm, whose Verdet constant is improved by 7.50%. In addition, the mean square deviation of the measured Verdet constants of different fibers with FRM is 0.61%, which is less than 0.99% in the case of no FRM. It is indicated that the introduction of FRM can cancel the random linear birefringence in the fiber and improve the measurement stability of the Verdet constant of fibers with different helical pitches. The FRM measurements of the Verdet constant for spun fibers with pitches of 1.0 mm, 1.5 mm, 2.0 mm, and 3.0 mm are improved by 3.94%, 4.72%, 4.24%, and 5.63%, respectively. The addition of FRM can reduce the influence of random linear birefringence on the measurement of the magneto-optical properties of doped fibers and further increase the measurement accuracy of the Verdet constant of doped fibers.