Multicore optical fiber alignment method, optical device manufacturing method, optical module manufacturing method, and multicore optical fiber alignment device

Abstract

This alignment method comprises: a step for calculating the amount of core positional deviation from an ideal core position, which is a design core position as designed or a nominal core position, for at least two of a plurality of cores included in a multicore optical fiber, assuming that a calculated positional deviation of a center position is zero; and a step for performing rotational alignment of the multicore optical fiber such that the sum of the m-th powers (where m is a number equal to or greater than 3) of the amounts of core positional deviation for the at least two cores, or the maximum amount of core positional deviation from among the amounts of core positional deviation for the at least two cores, is minimized.

Description

Method for centering multicore optical fibers, method for manufacturing optical devices, method for manufacturing optical modules, and centering device for multicore optical fibers

[0001] This disclosure relates to a method for centering multicore optical fibers, a method for manufacturing an optical device, a method for manufacturing an optical module, and a centering device for multicore optical fibers. This application claims priority under Japanese application No. 2024-170958, filed on 30 September 2024, and incorporates all the provisions of the said Japanese application.

[0002] Patent Document 1 discloses a method for aligning an optical waveguide component having multiple optical waveguides and an optical fiber array component having multiple optical fibers arranged in a row. In this alignment method, the amount of pitch deviation of each optical fiber facing each optical waveguide is determined with respect to the pitch of the optical waveguide as a reference. Then, at least one of the optical waveguide component and the optical fiber array component is moved relative to each other so that the average value of the determined pitch deviation amounts becomes zero.

[0003] Patent Document 2 discloses an optical fiber alignment device and alignment method for optical fibers having two or more cores, such as multicore fibers. In this alignment method, images of the end faces of two optical fibers having two or more cores are taken, and the position coordinates of the two or more cores at the end face are determined for each of the two optical fibers. Then, the position coordinates of the two or more cores are substituted into a theoretical formula that represents the sum of the misalignment losses when the two or more cores are connected to each other. From the theoretical formula, the positional relationship of the end faces of the two optical fibers is determined so as to minimize the sum of the misalignment losses. The two optical fibers are then arranged to satisfy the determined positional relationship.

[0004] Japanese Patent Publication No. 08-304667, International Publication No. 2015 / 025629

[0005] D. Marcuse, “Loss Analysis of Single-Mode Fiber Splices,” Bell Syst. Tech. J., vol. 56, no. 5, pp. 703-718, Jun. 1977.T. Hayashi et al., “Effective area measurement of few-mode fiber using far field scan technique with Hankel transform generalized for circularly-asymmetric mode,” Opt. Express, OE, vol. 26, no. 9, pp. 11137-11146, Apr. 2018.T. Hayashi et al., "Ultra-High-Density Microduct Cable with Uncoupled 12-Core Fibers with Standard 250-μm Coating," Optical Fiber Communication Conference, Tu2C.2 (2023)

[0006] A method for centering a multicore optical fiber according to one embodiment of the present disclosure comprises a determination step, a calculation step, and a rotational alignment step. In the determination step, the calculated center position is determined based on the position of at least one of the multiple different refractive index elements of the multicore optical fiber that have different refractive indices from the cladding, or based on the position of the cladding center. In the calculation step, assuming that the displacement of the calculated center position is zero, the amount of core displacement from the ideal core position, which is the design core position or the nominal core position, or the amount of connection loss obtained from the amount of core displacement is calculated for at least two of the multiple cores of the multicore optical fiber. In the rotational alignment step, the rotational alignment of the multicore optical fiber is performed so as to minimize the sum of the m-th power of the core displacement amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core displacement amount among the core displacement amounts of the at least two cores.

[0007] Figure 1 is a diagram showing the end face of a multicore optical fiber. Figure 2 is a flowchart showing the alignment method according to the first embodiment. Figure 3 is a schematic diagram showing the configuration of the alignment device according to the first embodiment. Figure 4 is a schematic diagram showing the optical device according to the first embodiment. Figure 5A is a schematic diagram showing the optical module according to the first embodiment. Figure 5B is a schematic diagram showing the optical module according to the first embodiment. Figure 6 is a diagram showing the end face of a first multicore optical fiber. Figure 7 is a diagram showing the end face of a second multicore optical fiber. Figure 8 is a diagram showing the butt surface when the end faces of the first multicore optical fiber and the end faces of the second multicore optical fiber are butted together. Figure 9 is a flowchart showing the alignment method according to the second embodiment. Figure 10 is a schematic diagram showing the configuration of the alignment device according to the second embodiment. Figure 11 is a schematic diagram showing the optical device according to the second embodiment. Figure 12 is a graph showing the relationship between core misalignment and probability density. Figure 13 is a graph showing the complementary cumulative distribution of core misalignment. Figure 14 is a graph showing the complementary cumulative distribution of connection loss. Figure 15 is a graph showing the complementary cumulative distribution of connection loss. Figure 16 is a graph showing the complementary cumulative distribution of connection loss. Figure 17 is a graph showing the complementary cumulative distribution of connection loss. Figure 18 is a graph showing the complementary cumulative distribution of connection loss. Figure 19 is a graph showing the complementary cumulative distribution of connection loss. Figure 20 is a graph showing the complementary cumulative distribution of connection loss. Figure 21 is a graph showing the complementary cumulative distribution of connection loss. Figure 22 is a graph showing the complementary cumulative distribution of connection loss. Figure 23 is a graph showing the complementary cumulative distribution of connection loss. Figure 24 is a graph showing the complementary cumulative distribution of connection loss. Figure 25 is a graph showing the complementary cumulative distribution of connection loss.

[0008] When connecting multicore optical fibers to optical components, or to other multicore optical fibers, rotational alignment of the multicore optical fibers is performed. During this process, misalignment between corresponding cores leads to increased connection loss. Therefore, it is desirable to reduce connection loss by minimizing core misalignment. Specifically, it is desirable to reduce the maximum core misalignment among multiple core misalignments to reduce the maximum connection loss.

[0009] The present disclosure aims to provide a method for aligning multicore optical fibers, a method for manufacturing an optical device, a method for manufacturing an optical module, and a multicore optical fiber aligning device that can reduce the maximum amount of core misalignment when aligning multicore optical fibers and optical components, or multicore optical fibers with each other.

[0010] According to this disclosure, it is possible to provide a method for aligning a multicore optical fiber, a method for manufacturing an optical device, a method for manufacturing an optical module, and a multicore optical fiber aligning device that can reduce the maximum amount of core misalignment when aligning a multicore optical fiber and an optical component, or multicore optical fibers with each other.

[0011] The contents of this disclosed embodiment will be explained.

[0012] [1] A method for centering a multicore optical fiber according to one embodiment of the present disclosure comprises a calculation step and a rotational alignment step. In the calculation step, assuming that the displacement of the calculated center position is zero, the amount of core displacement from the ideal core position, which is the design core position or the nominal core position, or the connection loss amount obtained from the amount of core displacement is calculated for at least two of the multiple cores of the multicore optical fiber. In the rotational alignment step, the rotational alignment of the multicore optical fiber is performed so as to minimize the sum of the m-th powers of the core displacement amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th powers of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core displacement amount among the core displacement amounts of the at least two cores. Furthermore, a multicore optical fiber centering device according to one embodiment of the present disclosure comprises a calculation unit and a centering control unit. The calculation unit assumes that the displacement of the calculated center position is zero and calculates the amount of core displacement from the ideal core position, which is the design core position or the nominal core position, or the connection loss amount obtained from the amount of core displacement, for at least two of the multiple cores of the multicore optical fiber. The centering control unit rotates and centers the multicore optical fiber so that the sum of the m-th power of the core displacement amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core displacement amount among the core displacement amounts of the at least two cores is minimized.

[0013] [2] A method for aligning a multicore optical fiber according to one embodiment of the present disclosure is a method for aligning a first multicore optical fiber and a second multicore optical fiber with respect to each other, comprising the steps of calculation and rotational alignment. In the calculation step, assuming that the calculated center position misalignment between the first multicore optical fiber and the second multicore optical fiber is zero, the amount of core misalignment between corresponding cores of the first multicore optical fiber and the second multicore optical fiber, or the connection loss amount obtained from the amount of core misalignment, is calculated for at least two of the multiple cores that each of the first multicore optical fiber and the second multicore optical fiber has. In the rotational alignment step, rotational alignment is performed on at least one of the first multicore optical fiber and the second multicore optical fiber such that the sum of the m-th power of the core misalignment amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core misalignment amount of the at least two cores is minimized. A multicore optical fiber alignment device according to one embodiment of the present disclosure is a device for aligning a first multicore optical fiber and a second multicore optical fiber with respect to each other, and comprises a calculation unit and an alignment control unit. The calculation unit assumes that the calculated center position misalignment between the first multicore optical fiber and the second multicore optical fiber is zero, and calculates the amount of core misalignment between corresponding cores of the first multicore optical fiber and the second multicore optical fiber, or the connection loss amount obtained from the amount of core misalignment, with respect to at least two of the multiple cores that each of the first multicore optical fiber and the second multicore optical fiber has. The alignment control unit performs rotational alignment of at least one of the first multicore optical fiber and the second multicore optical fiber so as to minimize the sum of the m-th power of the core misalignment amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core misalignment amount among the core misalignment amounts of the at least two cores.

[0014] In the centering methods and centering devices of [1] and [2] described above, rotational centering of the multi-core optical fiber is performed such that the sum of the m-th powers (m is a number of 3 or more) of the core misalignment amounts of at least two cores is minimized. In this case, for example, compared with the case where m is 2 or less, the ratio of the maximum core misalignment amount among the plurality of core misalignment amounts contributing to the sum of the m-th powers becomes larger, so the maximum core misalignment amount becomes smaller. Therefore, the maximum connection loss amount can be reduced. Alternatively, in the centering methods and centering devices of [1] and [2] described above, rotational centering of the multi-core optical fiber is performed such that the sum of the n-th powers (n is a number of 2 or more) of the connection loss amounts of at least two cores is minimized. As a result, compared with the case where n is 1, that is, compared with the case where rotational centering is performed such that the sum of the connection loss amounts of at least two cores is minimized, the ratio of the maximum connection loss amount among the connection loss amounts of the plurality of cores contributing to the sum of the n-th powers becomes larger, so the maximum connection loss amount becomes smaller. In other words, the maximum core misalignment amount becomes smaller. Alternatively, in the centering methods and centering devices of [1] and [2] described above, rotational centering of the multi-core optical fiber is performed such that the maximum core misalignment amount among the core misalignment amounts of at least two cores is minimized. Also in this case, since the maximum core misalignment amount becomes smaller, the maximum connection loss amount can be reduced.

[0015] When determining the quality of centering of the multi-core optical fiber, regarding the core misalignment amount, a smaller value is desirable, so it is mainly evaluated by referring to the maximum core misalignment amount. Therefore, by reducing the maximum core misalignment amount, it can be evaluated as a desirable centering result with a small core misalignment amount. According to the centering methods and centering devices of [1] and [2] described above, when performing rotational centering of the multi-core optical fiber and an optical component, or between multi-core optical fibers, the maximum core misalignment amount can be made smaller.

[0016] [3] The centering method of [1] above may further include a step of determining a computationally determined center position before the calculating step. In the determining step, based on the position of at least one of the plurality of different refractive index elements having a refractive index different from that of the cladding in the multi-core optical fiber, the position of the cladding center, or the arrangement of the elements of the optical fiber connection component that holds the tip of the multi-core optical fiber for connection with other optical components, the computationally determined center position is determined. In the centering device of [1] above, the calculation unit may determine the computationally determined center position based on the position of at least one of the plurality of different refractive index elements having a refractive index different from that of the cladding in the multi-core optical fiber, the position of the cladding center, or the arrangement of the elements of the optical fiber connection component that holds the tip of the multi-core optical fiber for connection with other optical components. In this case, the maximum core position deviation amount becomes smaller, and the maximum connection loss amount can be further reduced.

[0017] [4] The centering method of [2] above may further include a step of determining a computationally determined center position before the calculating step. In the determining step, in each of the first multi-core optical fiber and the second multi-core optical fiber, based on the position of at least one of the plurality of different refractive index elements having a refractive index different from that of the cladding or the position of the cladding center, the computationally determined center position is determined. In the centering device of [2] above, the calculation unit may determine the computationally determined center position based on the position of at least one of the plurality of different refractive index elements having a refractive index different from that of the cladding or the position of the cladding center in each of the first multi-core optical fiber and the second multi-core optical fiber. In this case, the maximum core position deviation amount becomes smaller, and the maximum connection loss amount can be further reduced.

[0018] [5] In the centering method and centering device of [1] to [4] above, m may be 4 or more. In this case, since the ratio of the contribution of the maximum core position deviation amount to the sum of the mth powers becomes even larger, the maximum core position deviation amount can be made even smaller.

[0019] [6] In the core alignment methods and apparatus described in [1] to [4] above, m may be 4. In this case, the proportion of the maximum core misalignment that contributes to the m-th sum is appropriately large, and the other core misalignments, excluding the maximum core misalignment, also contribute appropriately to the m-th sum. Thus, the maximum core misalignment is appropriately reduced, and it is possible to prevent the other core misalignments from becoming excessively large.

[0020] [7] In the calculation step of the alignment method described in [1] to [6] above, the amount of core misalignment or connection loss may be calculated for all of the multiple cores. Then, in the step of performing rotational alignment, rotational alignment may be performed so as to minimize the sum of the m-th power of the core misalignment amounts of all cores, the sum of the n-th power of the connection loss amounts of all cores, or the largest core misalignment amount among all cores. Similarly, the calculation unit of the alignment device described in [1] to [4] above may calculate the amount of core misalignment or connection loss for all of the multiple cores. Then, the alignment control unit may perform rotational alignment so as to minimize the sum of the m-th power of the core misalignment amounts of all cores, the sum of the n-th power of the connection loss amounts of all cores, or the largest core misalignment amount among all cores. In this case, since rotational alignment is performed based on the amount of core misalignment or connection loss of all cores, the maximum core misalignment amount can be made even smaller.

[0021] [8] In the centering methods and centering devices of [3] and [4] above, the at least one element with different refractive indices may include at least one core from among a plurality of cores. In this case, the calculated center position can be determined with high accuracy.

[0022] [9] In the centering methods and centering devices of [3] and [4] above, the at least one element with different refractive indices may include all of the cores among the multiple cores. In this case, the calculated center position can be determined with greater accuracy.

[0023]

[10] A method for manufacturing an optical device according to one embodiment of the present disclosure is a method for manufacturing an optical device comprising a multicore optical fiber and an optical component connected to the multicore optical fiber, comprising the alignment method of [1] or [3] above. According to this method for manufacturing an optical device, an optical device with a small maximum connection loss can be manufactured.

[0024]

[11] A method for manufacturing an optical device according to one embodiment of the present disclosure is a method for manufacturing an optical device comprising a first multicore optical fiber and a second multicore optical fiber connected to each other, and includes the alignment method of [2] or [4] above. According to this method for manufacturing an optical device, an optical device with a small maximum connection loss can be manufactured.

[0025]

[12] A method for manufacturing an optical module according to one embodiment of the present disclosure is a method for manufacturing an optical module comprising a multicore optical fiber and an optical fiber connector. The optical fiber connector holds the tip of the multicore optical fiber for connection with other optical components. The manufacturing method includes a method for centering a multicore optical fiber as described in [1] or [3] above. According to this method for manufacturing an optical module, an optical module with a small maximum connection loss can be manufactured.

[0026] [Details of Embodiments of the Disclosure] Specific examples of the Disclosure will be described below with reference to the drawings. The present invention is not limited to these examples, but is indicated by the claims, and all modifications within the meaning and scope of the equivalents of the claims are intended to be included. In the following description, the same elements in the description of the drawings are denoted by the same reference numerals, and redundant descriptions are omitted.

[0027] [First Embodiment] In the first embodiment, a case is described in which a multicore optical fiber is centered so that each of the multiple core positions of the multicore optical fiber approaches the ideal core position. Figure 1 shows the end face 11 of a multicore optical fiber (MCF) 10. The MCF 10 has a plurality of cores 12 dispersed on the end face 11 and a cladding 13 surrounding the plurality of cores 12. In the illustrated example, there are four cores 12. The refractive index of the plurality of cores 12 is greater than the refractive index of the cladding 13. The MCF 10 may further have one or more markers. The refractive index of the markers is greater than or less than the refractive index of the cladding 13. Each of the plurality of cores 12 and one or more markers is a different refractive index element with a refractive index different from that of the cladding 13.

[0028] The end face 11 has a circular shape, and the shapes of each core 12 and marker on the end face 11 are also circular. The end face 11 has a calculated center 11a. The center 11a may be determined based on the center of the cladding 13 (cladding center), or it may be determined based on the position of at least one of the multiple different refractive index elements. In one example, the center 11a is the cladding center. In another example, the center 11a is the center of the multiple different refractive index elements located at the cladding center. In yet another example, the center 11a is the geometric center of at least two of the multiple different refractive index elements. Each of the multiple cores 12 has a center point 12a, which is the center of a circle.

[0029] Figure 1 shows a Cartesian coordinate system with the center 11a as the origin. In the following explanation, the core position of each core 12 refers to the position of the center point 12a in the Cartesian coordinate system (XY Cartesian coordinate system). The position of the marker refers to the position of the center point of each marker in the Cartesian coordinate system.

[0030] Figure 1 shows multiple ideal core positions 12b. Each ideal core position 12b is the ideal core position for each core 12 and is used as a reference core position without misalignment when rotating the MCF 10. The multiple ideal core positions 12b have a relative positional relationship between multiple cores 12 in design or nominal terms. In the illustrated example, four ideal core positions 12b are shown that are equidistant from the ideal center point and equally spaced from each other. In other words, the four ideal core positions 12b are the four vertices of a square Q centered on the ideal center point. The center point 12a of each core 12 is shifted by a core misalignment amount D from the corresponding ideal core position 12b. The smaller the core misalignment amount D after rotational alignment, the smaller the connection loss with other multicore optical fibers or optical components. In many cases, the connection loss of the MCF 10 is evaluated based on the maximum connection loss among the connection losses of the multiple cores 12. Therefore, it is desirable to perform rotational alignment so that the maximum core misalignment amount D is as small as possible.

[0031] Figure 2 is a flowchart showing a centering method according to the first embodiment of this disclosure. This centering method is for bringing the core position of each core 12 closer to the corresponding ideal core position 12b, and includes an acquisition step ST11, a determination step ST12, a calculation step ST13, and a rotational centering step ST14. In the acquisition step ST11, the core positions of the multiple cores 12 of the MCF 10, the cladding center position, and the marker position are acquired. These positions may be acquired by photographing the end face 11 with a camera and analyzing the image. Alternatively, data regarding these positions acquired by analyzing an image of the end face 11 may be prepared. Alternatively, data regarding these positions that has already been digitized may be acquired.

[0032] In the determination step ST12, the position of the center 11a is determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12 and one or more markers). Alternatively, the position of the center 11a is determined based on the position of the cladding center. In one example, the position of the center 11a is determined to be the position of the cladding center. In another example, the position of the center 11a is determined to be the position of the center of the different refractive index element (core or marker) located at the cladding center among the multiple different refractive index elements. In yet another example, the position of the center 11a is determined to be the position of the geometric centers of at least two of the multiple different refractive index elements. The geometric centers of at least two different refractive index elements are calculated from the center positions of each of the at least two different refractive index elements.

[0033] If the cladding center is not used in the decision step ST12, the acquisition of the cladding center's position in the acquisition step ST11 described above can be omitted. If the marker's position is not used in the decision step ST12, the acquisition of the marker's position in the acquisition step ST11 described above can be omitted.

[0034] Alternatively, in the determination step ST12, the position of the center 11a is determined based on the arrangement of elements of the optical fiber connector that holds the tip of the MCF 10 for connection with other optical components. For example, if the optical fiber connector is a single-core ferrule having a single optical fiber insertion hole in the center of a cylindrical ceramic member, the position of the center 11a is set to the center of the outer diameter of the single-core ferrule. Then, with the center 11a as the coordinate origin, the X and Y axes are set so that the X or Y axis is parallel to, for example, one side of the flange attached to the rear end of the cylindrical ceramic member, or a straight line connecting two protruding points, and the ideal core position 12b is determined. If the optical fiber connector is a multi-core ferrule such as an MT ferrule, for example, the position of the center 11a is set to the center of one of the two guide pin holes provided between multiple optical fiber insertion holes, or the geometric center of the two guide pin holes is set to the position of the center 11a. Then, with the center 11a as the coordinate origin, the X and Y axes are determined such that, for example, the X axis or Y axis is parallel to the line connecting the centers of the two guide pin holes, and the ideal core position 12b is determined. If the optical fiber connection component is, for example, a substrate having an optical fiber housing groove (for example, a V groove), the position of the center 11a is determined based on the position of the optical fiber housing groove. Then, the X and Y axes are determined such that, for example, the X axis or Y axis is parallel to the top surface or side surface of the substrate, and the ideal core position 12b is determined.

[0035] In calculation step ST13, it is first assumed that the displacement of center 11a from the ideal center point is zero. In other words, center 11a is considered the ideal center point. Then, each of the multiple ideal core positions 12b is associated with each of the multiple cores 12, and the amount of core displacement D from the ideal core position 12b is calculated for at least two of the multiple cores 12. In one example, the amount of core displacement D from the ideal core position 12b is calculated for all cores 12.

[0036] In the rotational alignment step ST14, with the origin of coordinates as the center 11a, while calculatingly moving (i.e., rotating) the angular positions of the center points 12a of the plurality of cores 12 around the center 11a (i.e., the origin of coordinates), the m-th power sum of the core position deviation amounts D of the at least two cores 12 is obtained over a plurality of times. However, m is a number of 3 or more or 4 or more. m may be an integer or a real number. In one example, the m-th power sum of the core position deviation amounts D of all the cores 12 is obtained over a plurality of times. If the core position deviation amounts D of the four cores 12 are respectively D １ , D ２ , D ３ , D ４ , then the m-th power sum S1 is expressed as follows. S1 = D １ ｍ + D ２ ｍ + D ３ ｍ + D ４ ｍ In the rotational alignment step ST14, the angular positions of the plurality of cores 12 when the m-th power sum S1 becomes minimum are obtained, and the MCF10 is actually rotated so that the plurality of cores 12 approach those angular positions. In this specification, the "m-th power sum of the core position deviation amounts" includes not only the m-th power sum of the core position deviation amounts themselves but also the m-th power sum of values proportional to the core position deviation amounts.

[0037] Alternatively, in the rotational alignment step ST14, with the origin of coordinates as the center 11a, while calculatingly moving the angular positions of the center points 12a of the plurality of cores 12 around the center 11a, the maximum core position deviation amount D among the core position deviation amounts D of the at least two cores 12 is obtained over a plurality of times. In one example, the maximum core position deviation amount D among the core position deviation amounts D of all the cores 12 is obtained over a plurality of times. At this time, when the angular positions are different, the cores 12 having the maximum core position deviation amount D may be different from each other. In the rotational alignment step ST14, the angular positions of the plurality of cores 12 when the maximum core position deviation amount D becomes minimum are obtained, and the MCF10 is actually rotated so that the plurality of cores 12 approach those angular positions.

[0038] In calculation step ST13, for at least two of the multiple cores 12, the decibel value of the connection loss may be calculated instead of the core displacement amount D. In that case, in rotational centering step ST14, with the coordinate origin as the center 11a, the sum of the nth powers of the decibel values ​​of the connection loss for at least two of the cores 12 is calculated multiple times while calculating the angular position of the center points 12a of the multiple cores 12 around the center 11a. However, n is a number of 2 or more. n may be an integer or a real number. In one example, the decibel value of the connection loss is calculated for all cores 12, and the sum of the nth powers of the decibel values ​​of the connection loss is calculated multiple times. The decibel values ​​of the connection loss for the four cores 12 are each L １ , L ２ , L ３ , L ４ Therefore, the sum of n powers S² can be expressed as follows: S² = L １ ｎ +L ２ ｎ +L ３ ｎ +L ４ ｎ Then, the angular positions of the multiple cores 12 when the sum of the nth powers S2 is minimized are determined, and the MCF 10 is actually rotated so that the multiple cores 12 approach that angular position. In this specification, "connection loss amount" includes not only the connection loss amount itself, but also values ​​proportional to the connection loss amount.

[0039] The decibel value of the connection loss of each core 12 is calculated based on the core displacement D. If the electric field distribution of the waveguide mode of each core 12 can be approximated by a Gaussian distribution, the spot size of the core 12 is w １ , the ideal spot size lol ２ Therefore, the decibel value α of the connection loss can be expressed by the following formula (1) (see Non-Patent Document 1). For the spot size, for example, half the mode field diameter (MFD) as defined in section 3.4.1.4 of ITU-T G. 650.1 (10 / 2020) may be used. This value can be used as a good approximation of the spot size even if the electric field distribution of the waveguide modes of each core 12 does not perfectly match the Gaussian distribution.

[0040] Alternatively, the electric field distribution of the core 12 on the end face 11 is E １ , the ideal electric field distribution E ２ Therefore, the decibel value α of the connection loss can also be expressed by the following formula (2). The double integral sign in equation (2) represents the surface integral over an appropriate range on the end face 11. When considering a certain core 12, the integral should be performed over the range from the center point 12a of the core 12 until the electric field is sufficiently attenuated. At this time, the electric field distribution outside the core 12 is affected by the measurement, resulting in the electric field distribution E １ , E ２ If included in the above, integration should be performed within a range where the influence of electric field distributions other than core 12 can be ignored.

[0041] Electric field distribution E １ , E ２ This may be calculated from the refractive index distribution. Electric field distribution E １ , E ２ This can also be determined by measuring the intensity distribution of light emitted from the actual core 12 in the near field and calculating the square root of the linear value of the intensity. Electric field distribution E １ , E ２ The electric field distribution may be determined by measuring the intensity distribution of light emitted from the actual core 12 in the far field, calculating the square root of the linear value of the intensity to obtain the electric field distribution in the far field, and then applying an appropriate Hankel transform or two-dimensional Fourier transform to this electric field distribution (see Non-Patent Literature 2). When determining the electric field distribution in the far field, the electric field may cross zeros and its sign may change before and after the zeros, but this does not appear in the intensity distribution, so the sign of the electric field may be reversed before and after the minimum value of the intensity distribution. By appropriately reversing the sign, the electric field distribution in the near field can be determined more accurately.

[0042] Alternatively, the relationship between the core misalignment amount D and the decibel value α of the connection loss can be determined in advance, and when performing rotational alignment, this relationship can be referenced to calculate the decibel value α of the connection loss using interpolation.

[0043] In the above explanation, the sum of m-th powers S1, the sum of n-th powers S2, or the maximum core displacement D is calculated multiple times. However, if possible using a mathematical algorithm, the angular position of each core 12 that minimizes the sum of m-th powers S1, the sum of n-th powers S2, or the maximum core displacement D may be calculated directly in a single calculation.

[0044] In the rotational alignment step ST14, the angular positions of the multiple cores 12 are determined when the sum of the m-th power S1, the sum of the n-th power S2, or the maximum core displacement D is minimized. After actually rotating the MCF 10 so that the multiple cores 12 approach that angular position, the alignment of the center 11a with respect to the ideal center position (XY alignment) may be further performed. Alternatively, the rotation of the MCF 10 and XY alignment may be performed simultaneously to bring the sum of the m-th power S1, the sum of the n-th power S2, or the maximum core displacement D closer to its minimum value.

[0045] Figure 3 is a schematic diagram showing the configuration of the core alignment device 20 according to this embodiment. The core alignment device 20 comprises a storage unit 21, a calculation unit 22, a core alignment unit 24, and a core alignment control unit 25. The storage unit 21 stores the core positions (positions of the center point 12a) of the multiple cores 12 of the MCF 10. The core positions of the multiple cores 12 may be obtained by photographing the end faces 11 with a camera provided in the core alignment device 20 and analyzing the images. Alternatively, the storage unit 21 may pre-store data regarding core positions obtained by analyzing images of the end faces 11. The calculation unit 22 determines the position of the calculated center 11a by performing the determination step ST12 described above. The calculation unit 22 calculates the decibel value of the core position deviation amount D from the ideal core position 12b or the connection loss amount for at least two of the multiple cores 12 (or all of the cores 12) by performing the calculation step ST13 described above. The centering unit 24 rotates the MCF 10 while holding it. The centering control unit 25 controls the operation of the centering unit 24. By performing the rotational centering step ST14 described above, the centering control unit 25 operates the centering unit 24 to perform rotational centering of the MCF 10 so that the sum of the m-th powers of the core misalignment amounts D of at least two cores 12 (or all cores 12) S1, the sum of the n-th powers of the decibel values ​​of the connection loss amounts of at least two cores 12 (or all cores 12) S2, or the largest core misalignment amount D among the core misalignment amounts D of at least two cores 12 (or all cores 12) is minimized.

[0046] The rotation alignment device 20, specifically the storage unit 21, the calculation unit 22, and the rotation alignment control unit 25, may be configured as a computer including, for example, a processor (CPU), main memory such as ROM and RAM, and auxiliary storage such as a hard disk. The storage unit 21 is implemented by the main memory or auxiliary storage. The computer's processor can perform the calculation unit 22 and the rotation alignment control unit 25 according to the measurement program. The measurement program causes the computer's processor to execute the acquisition step ST11, the determination step ST12, the calculation step ST13, and the rotation alignment step ST14 described above. The measurement program is stored in an internal or external storage device or storage medium of the computer, such as an auxiliary storage device.

[0047] Figure 4 is a schematic diagram showing an optical device 40 according to this embodiment. The optical device 40 comprises an MCF 10 and optical components 41. The optical components 41 have a plurality of cores, and each core is connected to a corresponding core 12, thereby connecting to the MCF 10. The above-described alignment method is used when manufacturing the optical device 40.

[0048] Figures 5A and 5B are schematic diagrams showing an optical module 50 according to this embodiment. The optical module 50A shown in Figure 5A comprises a single MCF 10 and an optical fiber connector 51. The optical fiber connector 51 holds the tip of the MCF 10 for connection between the MCF 10 and other optical components. The optical fiber connector 51 is, for example, a single-core ferrule having a single optical fiber insertion hole in the center of a cylindrical ceramic member. When manufacturing the optical module 50A, the MCF 10 is inserted into the optical fiber insertion hole of a flanged ceramic member, rotated and centered using the centering method described above, and then the MCF 10 is adhesively fixed to the optical fiber insertion hole. Alternatively, the MCF 10 is rotated and centered using the centering method described above, and then the MCF 10 is inserted into the optical fiber insertion hole of a flanged ceramic member and adhesively fixed. Alternatively, the MCF10 may be inserted into the optical fiber insertion hole of a flangeless ceramic member and fixed in place by adhesive, then the MCF10 may be rotated and centered using the centering method described above, and then the flange may be attached to the ceramic member.

[0049] The optical module 50B shown in Figure 5B comprises a plurality of MCFs 10 and an optical fiber connecting component 52. The optical fiber connecting component 52 holds the ends of the plurality of MCFs 10 for connection between the plurality of MCFs 10 and other optical components (e.g., other plurality of MCFs). The optical fiber connecting component 52 is, for example, a multi-core ferrule such as an MT ferrule, or a substrate having a plurality of optical fiber accommodating grooves (e.g., V-grooves). When manufacturing the optical module 50B, each MCF 10 is inserted into each optical fiber insertion hole of the multi-core ferrule, or placed in each optical fiber accommodating groove of the substrate, and then each MCF 10 is rotated and aligned using the alignment method described above, and then each MCF 10 is bonded and fixed to each optical fiber insertion hole or each optical fiber accommodating groove. Alternatively, after rotating and aligning each MCF 10 using the alignment method described above, each MCF 10 is inserted into each optical fiber insertion hole of the multi-core ferrule, or placed in each optical fiber accommodating groove of the substrate, and then bonded and fixed.

[0050] The effects obtained by the alignment method and alignment device 20 of this embodiment, as described above, will now be explained along with the problems of the reference example. In the reference example, the sum of the core misalignment amounts D of the multiple cores 12 is calculated multiple times while the angular position of the center points 12a of the multiple cores 12 around the center 11a is moved in calculation form. Then, the MCF 10 is rotated to align its angular position with the angular position where the sum of the core misalignment amounts D is minimized. In this case, the core misalignment amount D may be overestimated, and the maximum connection loss may increase. Specifically, when minimizing the sum of the core misalignment amounts D, the contribution of both the core 12 with a small core misalignment amount D and the core 12 with a large core misalignment amount D to the sum of the core misalignment amount D is the same. Therefore, when minimizing the sum of the core misalignment amounts D, it is possible that the core misalignment amount D of the core 12 with a small core misalignment amount D will be further reduced, and the core misalignment amount D of the core 12 with a large core misalignment amount D will be further increased, and so on. The same or similar thing can happen when minimizing the sum of squares of core displacement amounts D. The same or similar thing can happen when minimizing the sum of connection losses, since connection losses are proportional to the square of core displacement amount D.

[0051] To address these problems, in this embodiment, the rotational alignment of the MCF 10 is performed so that the sum of the m-th powers of the core misalignment amounts D of at least two cores 12, S1 (where m is a number of 3 or more), is minimized. In this case, compared to the case where m is 2 or less, for example, the proportion of the maximum core misalignment amount D contributing to the m-th power sum S1 is larger, so the maximum core misalignment amount D becomes smaller. Therefore, the maximum connection loss can be reduced. Alternatively, in this embodiment, the rotational alignment of the MCF 10 is performed so that the sum of the n-th powers of the decibel values ​​of the connection loss amounts of at least two cores 12, S2 (where n is a number of 2 or more), is minimized. As a result, compared to the case where n is 1, that is, when the rotational alignment is performed so that the sum of the connection loss amounts of at least two cores 12 is minimized, the proportion of the maximum connection loss amount contributing to the n-th power sum S2 is larger, so the maximum connection loss becomes smaller. In other words, the maximum core misalignment amount D becomes smaller. Alternatively, in this embodiment, the rotational alignment of the MCF 10 is performed so that the maximum core misalignment amount D among the core misalignment amounts D of at least two cores 12 is minimized. In this case as well, the maximum core misalignment amount D is reduced, so the maximum connection loss can be reduced.

[0052] When determining the quality of alignment of the MCF 10, a smaller core misalignment amount D is desirable, so the evaluation is mainly based on the maximum core misalignment amount D. Therefore, by reducing the maximum core misalignment amount D, a desirable alignment result with a small core misalignment amount D can be evaluated. According to the alignment method and alignment device 20 of this embodiment, the maximum core misalignment amount D can be made even smaller when the MCF 10 is rotationally aligned with respect to the optical component to which it is connected.

[0053] The method of minimizing the sum of the nth powers of the decibel values ​​of connection loss, S2, allows for a more direct evaluation of connection loss than a method based solely on the core displacement D, especially when one or both of the spot size and MFD have variations, or when one or both of the spot size and MFD are designed to differ across multiple cores 12. Therefore, connection loss can be reduced with greater accuracy.

[0054] As in this embodiment, the alignment method may include a determination step ST12 before the calculation step ST13. In the determination step ST12, the position of the center 11a is determined based on the position of at least one of the multiple different refractive index elements in the MCF 10 that have different refractive indices from the cladding 13, or the position of the center 11a is determined based on the position of the cladding center. In this case, the maximum core misalignment amount D becomes smaller, and the maximum connection loss can be further reduced.

[0055] As mentioned above, m may be 4 or greater. In this case, the proportion that the maximum core displacement D contributes to the m-th power sum S1 becomes even larger, so the maximum core displacement D can be made even smaller, and the maximum connection loss can be reduced even further.

[0056] As mentioned above, m may also be 4. In this case, the proportion of the maximum core displacement D contributing to the m-th power sum S1 becomes appropriately large, and the other core displacements D, excluding the maximum core displacement D, also contribute appropriately to the m-th power sum S1. Therefore, the maximum core displacement D is appropriately reduced, and it is possible to prevent the other core displacements D from becoming excessively large.

[0057] As described above, in calculation step ST13, the core misalignment amount D or connection loss amount may be calculated for all of the multiple cores 12. Then, in rotational alignment step ST14, rotational alignment may be performed so that the sum of the m-th power of the core misalignment amounts D of all cores 12 S1, the sum of the n-th power of the decibel values ​​of the connection loss amount of all cores 12 S2, or the largest core misalignment amount D among all cores 12 is minimized. In this case, since rotational alignment is performed based on the core misalignment amount D or connection loss amount of all cores 12, the largest core misalignment amount D can be made even smaller, and the largest connection loss can be reduced even further.

[0058] As mentioned above, in addition to the rotational alignment of the first MCF 10A and the second MCF 10B, the center 11aa and the center 11ab may also be aligned (XY alignment). In this case, the maximum core misalignment amount D can be further reduced, and the maximum connection loss can be further reduced.

[0059] As mentioned above, in the determination step ST12, the position of the center 11a may be determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12 and one or more markers). In this case, alignment can be performed even if the coordinates of the cladding center cannot be accurately measured. In addition, generally, the measurement error is smaller for the center position of a circle with a smaller diameter than for the center position of a circle with a larger diameter. For example, if the diameter of the cladding 13 is 10 times the diameter of the core 12, the measurement error of the center position of the core 12 will be approximately one-tenth of the measurement error of the center position of the cladding 13. Therefore, since the measurement accuracy can be higher for the position of each different refractive index element compared to the cladding center position, XY alignment can be performed with greater accuracy. Thus, connection loss can be reduced.

[0060] In the determination step ST12, the position of the center 11a may be determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12 and one or more markers). The at least one different refractive index element may include at least one of the multiple cores 12. In this case, the calculated position of the center 11a can be determined with high accuracy. Alternatively, the at least one different refractive index element may include all of the multiple cores 12. In this case, the calculated position of the center 11a can be determined with even higher accuracy.

[0061] [Second Embodiment] In this embodiment, the case in which the first MCF and the second MCF are aligned with each other will be described. Figure 6 is a diagram showing the end face 11A of the first MCF 10A. The first MCF 10A has a plurality of cores 12A (four in the illustrated example) distributed on the end face 11A, and a cladding 13A surrounding the plurality of cores 12A. The end face 11A has a calculated center 11aa. Each of the plurality of cores 12A has a center point 12aa, which is the center of a circle. Figure 7 is a diagram showing the end face 11B of the second MCF 10B. The second MCF 10B has a plurality of cores 12B (four in the illustrated example) distributed on the end face 11B, and a cladding 13B surrounding the plurality of cores 12B. The end face 11B has a calculated center 11ab. Each of the plurality of cores 12B has a center point 12ab, which is the center of a circle. The first MCF10A and the second MCF10B may further have one or more markers.

[0062] Figure 8 shows the abutting surface 11C when the end face 11A of the first MCF 10A and the end face 11B of the second MCF 10B are butted together. The center point 12aa of each core 12A and the corresponding center point 12ab of the core 12B are offset from each other by a core displacement amount D. The smaller the core displacement amount D after rotational alignment, the smaller the connection loss between the first MCF 10A and the second MCF 10B. In most cases, the connection loss between the first MCF 10A and the second MCF 10B is evaluated based on the maximum connection loss among multiple sets of connection losses between the corresponding cores 12A and 12B. Therefore, it is desirable to perform rotational alignment so that the maximum core displacement amount D is as small as possible.

[0063] Figure 9 is a flowchart showing a centering method according to a second embodiment of the present disclosure. This centering method is for bringing the core positions of corresponding cores 12A and core 12B closer together, and includes an acquisition step ST21, a determination step ST22, a calculation step ST23, and a rotational centering step ST24. In the acquisition step ST21, the core positions of the multiple cores 12A of the first MCF 10A and the multiple cores 12B of the second MCF 10B, as well as the cladding centers and marker positions of the first MCF 10A and the second MCF 10B, are acquired. These positions may be acquired by photographing the end faces 11A and 11B with a camera and analyzing the images. Alternatively, data regarding these positions acquired by analyzing images of the end faces 11A and 11B may be prepared. Alternatively, data regarding these positions that has already been digitized may be acquired.

[0064] In the determination step ST22, the position of center 11aa is determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12A and one or more markers) of the first MCF 10A. The position of center 11ab is determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12B and one or more markers) of the second MCF 10B. Alternatively, the positions of centers 11aa and 11ab are determined based on the positions of the cladding centers of the first MCF 10A and the second MCF 10B, respectively. In one example, the positions of centers 11aa and 11ab are determined to be the positions of the cladding centers. In another example, the positions of centers 11aa and 11ab are determined to be the positions of the centers of the different refractive index elements (cores or markers) that are placed at the cladding centers among the multiple different refractive index elements. In yet another example, the positions of centers 11aa and 11ab are determined to be the positions of the geometric centers of at least two of the multiple different refractive index elements. The geometric centers of at least two elements with different refractive indices are calculated from the central positions of each of the two elements with different refractive indices.

[0065] If the cladding center position is not used in the decision step ST22, the acquisition of the cladding center position in the acquisition step ST21 described above can be omitted. If the marker position is not used in the decision step ST22, the acquisition of the marker position in the acquisition step ST21 described above can be omitted.

[0066] In calculation step ST23, it is first assumed that the relative positional displacement between the center 11aa of the first MCF10A and the center 11ab of the second MCF10B is zero. Then, each of the multiple cores 12A of the first MCF10A and each of the multiple cores 12B of the second MCF10B are associated with each other, and the core positional displacement amount D is calculated for at least two of the multiple pairs of cores 12A and cores 12B. In one example, the core positional displacement amount D is calculated for all pairs of cores 12A and cores 12B.

[0067] In the rotational centering step ST24, with the coordinate origin set as centers 11aa and 11ab, the relative angular positions of the center points 12aa of the multiple cores 12A and 12ab of the multiple cores 12B around the coordinate origin are calculated and moved (i.e., rotated), and the sum of the m-th powers of the core displacement amounts D of at least two pairs of cores is calculated multiple times. However, m is a number of 3 or more or 4 or more. m may be an integer or a real number. In one example, the sum of the m-th powers of the core displacement amounts D of all pairs of cores 12A and cores 12B is calculated multiple times. The core displacement amounts D of the four pairs are each D １ , D ２ , D ３ , D ４ Therefore, the sum of the mth powers S1 can be expressed as follows: S1 = D １ ｍ +D ２ ｍ +D ３ ｍ +D ４ ｍIn the rotational alignment step ST24, the relative angular positions of the cores 12A and 12B when the sum of the m powers S1 is minimized are determined, and one or both of the first MCF 10A and the second MCF 10B are actually rotated so that the cores 12A and 12B approach that angular position. The centers of rotation at this time are not limited to centers 11aa and 11ab. For example, in order to align the axes of the first MCF 10A and the second MCF 10B, the outer circumference of the first MCF 10A and the outer circumference of the second MCF 10B may be aligned.

[0068] Alternatively, in the rotational alignment step ST24, with the coordinate origin set as centers 11aa and 11ab, the maximum core displacement D among the core displacement D of at least two pairs is determined multiple times while calculating and moving the relative angular positions of the center points 12aa of the multiple cores 12A and the center points 12ab of the multiple cores 12B around the coordinate origin. In one example, the maximum core displacement D among the core displacement D of all pairs of cores 12A and cores 12B is determined multiple times. In this case, if the angular positions are different, the pair having the maximum core displacement D may be different from each other. In the rotational alignment step ST24, the relative angular positions of cores 12A and 12B when the maximum core displacement D is minimized are determined, and one or both of the first MCF 10A and the second MCF 10B are actually rotated so that cores 12A and 12B approach that angular position.

[0069] In calculation step ST23, for at least two of the multiple sets of cores 12A and 12B, the decibel value of the connection loss may be calculated instead of the core displacement amount D. In that case, in rotational centering step ST24, with the coordinate origin set to centers 11aa and 11ab, the relative angular position of the center point 12aa of the multiple cores 12A and the center point 12ab of the multiple cores 12B around the coordinate origin is calculated and the sum of the nth powers of the decibel values ​​of the connection loss for the above at least two sets is calculated multiple times. However, n is a number of 2 or more. n may be an integer or a real number. In one example, the decibel value of the connection loss is calculated for all sets of cores 12A and cores 12B, and the sum of the nth powers of the decibel values ​​of the connection loss is calculated multiple times. The decibel values ​​of the connection loss for the four sets are each L １, L ２ , L ３ , L ４ Therefore, the sum of n powers S² can be expressed as follows: S² = L １ ｎ +L ２ ｎ +L ３ ｎ +L ４ ｎ Then, the relative angular positions of the cores 12A and 12B when the sum of the nth powers S2 is minimized are determined, and one or both of the first MCF 10A and the second MCF 10B are actually rotated so that the cores 12A and 12B approach that angular position. The method for calculating the decibel value of the connection loss is the same as in the first embodiment.

[0070] In the above explanation, the sum of m-th powers S1, the sum of n-th powers S2, or the maximum core displacement D is calculated multiple times. However, if possible using a mathematical algorithm, the relative angular positions of cores 12A and 12B that minimize the sum of m-th powers S1, the sum of n-th powers S2, or the maximum core displacement D may be calculated directly in a single calculation.

[0071] In the rotational alignment step ST24, the relative angular positions of the cores 12A and 12B are determined when the sum of the m-th power S1, the sum of the n-th power S2, or the maximum core displacement D is minimized. After actually rotating one or both of the first MCF 10A and the second MCF 10B so that the cores 12A and 12B approach that angular position, the center 11aa of the first MCF 10A and the center 11ab of the second MCF 10B may be further aligned (XY alignment). Alternatively, the relative rotation and relative XY alignment of the first MCF 10A and the second MCF 10B may be performed simultaneously to bring the sum of the m-th power S1, the sum of the n-th power S2, or the maximum core displacement D closer to its minimum value.

[0072] Figure 10 is a schematic diagram showing the configuration of the centering device 30 according to this embodiment. The centering device 30 comprises a storage unit 31, a calculation unit 32, a centering unit 34, and a centering control unit 35. The storage unit 31 stores the core positions (position of the center point 12aa) of the multiple cores 12A of the first MCF 10A, and the core positions (position of the center point 12ab) of the multiple cores 12B of the second MCF 10B. The core positions of the cores 12A and 12B may be obtained by photographing the end faces 11A and 11B with a camera provided in the centering device 30 and analyzing the images. Alternatively, the storage unit 31 may pre-store data regarding the core positions obtained by analyzing images of the end faces 11A and 11B. The calculation unit 32 determines the calculated center positions 11aa and 11ab by performing the determination step ST22 described above. The calculation unit 32 performs the calculation step ST23 described above to calculate the decibel value of the core misalignment amount D or connection loss amount for at least two sets (or all sets) of the multiple sets of cores 12A and 12B. The centering unit 34 rotates one or both of the first MCF 10A and the second MCF 10B while holding the first MCF 10A and the second MCF 10B. The centering control unit 35 controls the operation of the centering unit 34. The centering control unit 35 performs the rotational centering step ST24 described above, and operates the centering unit 34 to perform rotational centering of the first MCF 10A and the second MCF 10B so as to minimize the sum of the m-th powers of the core misalignment amounts D of at least two pairs (or all pairs) of cores 12A and 12B, the sum of the n-th powers of the decibel values ​​of the connection loss amounts of at least two pairs (or all pairs), or the largest core misalignment amount D among the core misalignment amounts D of at least two pairs (or all pairs).

[0073] The gear alignment device 30, specifically the storage unit 31, the calculation unit 32, and the gear alignment control unit 35, may be configured as a computer including, for example, a processor (CPU), main memory such as ROM and RAM, and auxiliary storage such as a hard disk. The storage unit 31 is implemented by the main memory or auxiliary storage. The computer's processor can perform the calculation unit 32 and the gear alignment control unit 35 according to the measurement program. The measurement program causes the computer's processor to execute the acquisition step ST21, the determination step ST22, the calculation step ST23, and the rotational gear alignment step ST24 described above. The measurement program is stored in an internal or external storage device or storage medium of the computer, such as an auxiliary storage device.

[0074] Figure 11 is a schematic diagram showing an optical device 42 according to this embodiment. The optical device 42 comprises a first MCF 10A and a second MCF 10B connected to each other. The above-described alignment method is used when manufacturing the optical device 42.

[0075] The effects obtained by the alignment method and alignment device 30 of this embodiment, as described above, will now be explained. In this embodiment as well, the rotational alignment of the first MCF 10A and the second MCF 10B is performed so as to minimize the sum of the m-power sum of the core misalignment amounts D of at least two sets of cores 12A and 12B (where m is a number of 3 or more). In this case, compared to the case where m is 2 or less, for example, the proportion of the maximum core misalignment amount D contributing to the m-power sum S1 becomes larger, so the maximum core misalignment amount D becomes smaller. Therefore, the maximum connection loss can be reduced. Alternatively, in this embodiment, the rotational alignment of the first MCF 10A and the second MCF 10B is performed so as to minimize the sum of the n-power sum of the decibel values ​​of the connection loss amounts of at least two sets of cores 12A and 12B (where n is a number of 2 or more). As a result, compared to the case where n is 1, the proportion of the maximum connection loss contributing to the n-power sum S2 becomes larger, so the maximum connection loss becomes smaller. In other words, the maximum core misalignment D becomes smaller. Alternatively, in this embodiment, the rotational alignment of the first MCF 10A and the second MCF 10B is performed such that the maximum core misalignment D among at least two pairs of core misalignments D of cores 12A and 12B is minimized. In this case as well, the maximum core misalignment D becomes smaller, thus reducing the maximum connection loss.

[0076] When determining the quality of alignment of the first MCF 10A and the second MCF 10B, a smaller core misalignment amount D is desirable, and therefore, the evaluation is mainly based on the maximum core misalignment amount D. Accordingly, by reducing the maximum core misalignment amount D, a desirable alignment result with a small core misalignment amount D can be evaluated. According to the alignment method and alignment device 30 of this embodiment, the maximum core misalignment amount D can be made even smaller when the first MCF 10A and the second MCF 10B are rotationally aligned with respect to each other.

[0077] As in this embodiment, the alignment method may include a determination step ST22 before the calculation step ST23. In the determination step ST22, in the first MCF 10A and the second MCF 10B, the positions of the centers 11aa and 11ab are determined based on the position of at least one of the multiple elements with different refractive indices that have different refractive indices from the cladding 13A and 13B, or the positions of the centers 11aa and 11ab are determined based on the position of the cladding center. In this case, the maximum core misalignment amount D becomes smaller, and the maximum connection loss can be further reduced.

[0078] As described above, in calculation step ST23, the core misalignment amount D or connection loss amount may be calculated for all of the multiple sets of cores 12A and 12B. Then, in rotational alignment step ST24, rotational alignment may be performed so as to minimize the sum of the m-th powers of the core misalignment amounts D for all sets of cores 12A and 12B S1, the sum of the n-th powers of the decibel values ​​of the connection loss amounts for all sets of cores 12A and 12B S2, or the maximum core misalignment amount D among all sets of cores 12A and 12B. In this case, since rotational alignment is performed based on the core misalignment amount D or connection loss amount for all sets of cores 12A and 12B, the maximum core misalignment amount D can be made even smaller.

[0079] As in this embodiment, in the determination step ST22, the position of the center 11aa may be determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12A and one or more markers). The at least one different refractive index element may include at least one of the multiple cores 12A. In this case, the calculated position of the center 11aa can be determined with high accuracy. Alternatively, the at least one different refractive index element may include all of the multiple cores 12A. In this case, the calculated position of the center 11aa can be determined with even higher accuracy. The position of the center 11ab may be determined based on the position of at least one of the multiple different refractive index elements (multiple cores 12B and one or more markers). The at least one different refractive index element may include at least one of the multiple cores 12B. In this case, the calculated position of the center 11ab can be determined with high accuracy. Alternatively, the at least one different refractive index element may include all of the multiple cores 12B. In this case, the position of the calculated center 11ab can be determined with greater precision.

[0080] The inventors have verified the effects of the first and second embodiments described above through calculations. The verification results are described below. In the following description, comparative examples and examples 1 to 4 are defined as follows. The actual core positions are assumed to vary according to a normal distribution with a standard deviation of 0.25 μm. Comparative example: Rotational alignment of MCF10, or the first MCF10A and the second MCF10B, is performed so as to minimize the sum of the squares of the core misalignment D. Example 1: Rotational alignment of MCF10, or the first MCF10A and the second MCF10B, is performed so as to minimize the sum of the cubes of the core misalignment D. Example 2: Rotational alignment of MCF10, or the first MCF10A and the second MCF10B, is performed so as to minimize the sum of the fourth powers of the core misalignment D. Example 3: Rotational alignment of MCF10, or the first MCF10A and the second MCF10B, is performed so that the sum of the eighth powers of the core displacement amounts D is minimized. Example 4: Rotational alignment of MCF10, or the first MCF10A and the second MCF10B, is performed so that the maximum core displacement amount D among the core displacement amounts D of multiple cores 12, or multiple sets of cores 12A and 12B, is minimized.

[0081] [Examples] In a first embodiment in which multiple core positions of a single MCF 10 are aligned to approach the ideal core position, the case in which the cladding center is the calculated center 11a will be described. Figure 12 is a graph showing the relationship between the amount of core misalignment D and the probability density for comparative examples and Examples 1 to 4 in the first embodiment, where the four design core positions are (20 μm, 20 μm), (-20 μm, 20 μm), (-20 μm, -20 μm), and (20 μm, -20 μm) in a Cartesian coordinate system with the center 11a (or centers 11aa, 11ab) as the origin. The upper part of Figure 12 shows the probability density on the vertical axis linearly. The lower part of Figure 12 shows the probability density on the vertical axis logarithmically. Figure 13 is a graph showing the complementary cumulative distribution of the amount of core misalignment D for comparative examples and Examples 1 to 4 in the same case. The complementary cumulative distribution of core displacement D is the number obtained by subtracting the cumulative distribution from 1; in other words, it is the distribution that shows the probability that the core displacement D is greater than or equal to a certain value. The upper part of Figure 13 shows the complementary cumulative distribution on the vertical axis in a linear representation. The lower part of Figure 13 shows the complementary cumulative distribution on the vertical axis in a logarithmic representation.

[0082] Referring to Figures 12 and 13, in the comparative example, the tail of the probability distribution is distributed over a larger displacement amount, and the tail of the probability distribution is longer, compared to Examples 1 to 4. In other words, the maximum core displacement amount D is larger in the comparative example compared to Examples 1 to 4. To put it another way, according to Examples 1 to 4, the maximum core displacement amount D can be reduced. Referring to Figures 12 and 13, Example 2 has a shorter tail of the probability distribution than Example 1, so the maximum core displacement amount D can be reduced even further. Examples 3 and 4 have shorter tails of the probability distribution than Example 2, so the maximum core displacement amount D can be reduced even further. Here, reducing the maximum core displacement amount D is not merely a manipulation of apparent numbers. This means that the maximum core displacement amount D is actually reduced, which means that there is an appropriate rotation angle around the fiber center of the multicore optical fiber, and in the examples, alignment can be performed with an appropriate rotation angle.

[0083] Here, we consider the connection loss (butt coupling loss) when the mode field diameter (MFD) is set to 8.6 μm and MFD mismatch can be ignored. The upper part of Figure 14 is a graph showing the complementary cumulative distribution of connection loss (hereinafter referred to as master connection loss) caused by the core position misalignment between the ideal core position and the measured core position in the first embodiment. The lower part of Figure 14 is a graph showing the complementary cumulative distribution of connection loss (hereinafter referred to as random connection loss) when two rotationally aligned MCF10s are randomly selected, and while maintaining the two MCF10s at the adjusted rotation angle, the positions of their centers 11a are aligned, and then they are randomly butt coupled. Referring to Figure 14, the tail of the complementary cumulative distribution of connection loss can be made shorter in Examples 2 and 4 than in the comparative example. From this, it can be seen that Examples 2 and 4 can reduce the worst value of connection loss compared to the comparative example. In Example 2, the value of the complementary cumulative distribution becomes smaller than in the comparative example even at small values ​​of connection loss.

[0084] In the examples shown in Figures 12 to 14 above, the MFD is set to 8.6 μm. Even if the MFD is different, it is presumed that the same or similar results will be obtained, as only the scale of the horizontal axis of the graph will change.

[0085] In the examples shown in Figures 12 to 14 above, the four design core positions are (20 μm, 20 μm), (-20 μm, 20 μm), (-20 μm, -20 μm), and (20 μm, -20 μm) in a Cartesian coordinate system with the center 11a as the origin, but the core positions are not limited to these. Figure 15 is a graph showing the complementary cumulative distribution of connection loss when the four core positions are (12.5 μm, 12.5 μm), (-12.5 μm, 12.5 μm), (-12.5 μm, -12.5 μm), and (12.5 μm, -12.5 μm). The upper part of Figure 15 shows the connection loss to the master, and the lower part of Figure 15 shows the random connection loss. Figure 16 is a graph showing the complementary cumulative distribution of connection loss when the four core positions are (10 μm, 10 μm), (-10 μm, 10 μm), (-10 μm, -10 μm), and (10 μm, -10 μm). The upper part of Figure 16 shows the connection loss relative to the master, and the lower part of Figure 16 shows the random connection loss. Referring to Figures 15 and 16, the tails of the complementary cumulative distribution are shorter in Examples 2 and 4 compared to the comparative example. In Example 2, the value of the complementary cumulative distribution becomes smaller than that of the comparative example even at small connection loss values.

[0086] The number of cores is not limited to four. For example, in terms of design, eight cores may be arranged at equal intervals on a circle centered on the center 11a. Figure 17 is a graph showing the complementary cumulative distribution of connection loss when eight cores are arranged at equal intervals of 30 μm. The upper part of Figure 17 shows the connection loss to the master, and the lower part of Figure 17 shows the random connection loss. Referring to Figure 17, the tails of the complementary cumulative distribution are shorter in Examples 2 and 4 compared to the comparative example. In Example 2, the value of the complementary cumulative distribution becomes smaller than that of the comparative example even at small connection loss values.

[0087] The arrangement of multiple cores is not limited to those having two or more rotational symmetries, as in the examples described above. Figure 18 is a graph showing the complementary cumulative distribution of connection loss in a 12-core MCF without two or more rotational symmetries, as described in Non-Patent Literature 3, where the design core center spacing is 35 μm. The upper part of Figure 18 shows the connection loss to the master, and the lower part shows the random connection loss. Referring to Figure 18, the tails of the complementary cumulative distribution are shorter in Examples 2 and 4 compared to the comparative example. In Example 2, the value of the complementary cumulative distribution becomes smaller than that of the comparative example even at small connection loss values.

[0088] Alignment may be performed using only some (at least two) of the cores 12, rather than all of them. For example, eight of the twelve cores described above are located on the circumference closer to the outer edge relative to the other four cores. In this case, rotational alignment may be performed using only the eight outer cores. Figure 19 is a graph showing the complementary cumulative distribution of connection loss in such a case. The upper part of Figure 19 shows the connection loss relative to the master, and the lower part of Figure 19 shows the random connection loss. Referring to Figure 19, the tails of the complementary cumulative distribution are shorter in Examples 2 and 4 compared to the comparative example.

[0089] In the first embodiment, in which multiple core positions of a single MCF 10 are aligned to approach the ideal core position, the case in which the geometric center of the multiple cores 12 is set to the calculated center 11a will be described. Figure 20 is a graph showing the complementary cumulative distribution of connection loss for the comparative example and Examples 2 and 4, when the four design core positions are (20 μm, 20 μm), (-20 μm, 20 μm), (-20 μm, -20 μm), and (20 μm, -20 μm) in a Cartesian coordinate system with the center 11a as the origin. The upper part of Figure 20 shows the connection loss to the master, and the lower part of Figure 20 shows the random connection loss. Referring to Figure 20, the tails of the complementary cumulative distribution are shorter in Examples 2 and 4 compared to the comparative example. In Example 2, the value of the complementary cumulative distribution becomes smaller than that of the comparative example even at small connection loss values. Comparing Figure 20 with Figures 14 to 19, by using the geometric centers of the multiple cores 12 as the center 11a, the complementary cumulative distribution value becomes smaller than that of the comparative example, even among the values ​​of small connection loss, compared to the case where the cladding center is the center 11a. Even when the number and arrangement of the cores 12 differ from the above, the tails of the complementary cumulative distribution become shorter in Examples 2 and 4 compared to the comparative example.

[0090] The geometric centers of some (at least two) of the cores 12, rather than all of them, may be used as the calculated centers 11a. For example, four of the twelve cores described in Non-Patent Document 3 are located on a circumference close to the center relative to the other eight cores. In this case, the geometric centers of the four cores close to the center may be used as the calculated centers 11a. Hereinafter, the measurement error of the cladding center position will be assumed to be 0.3 μm in standard deviation, and the measurement error of the core center position will be assumed to be negligible compared to the variation in core position. Not limited to this case, if the measurement error of the geometric centers of some (at least two) of the cores 12 is smaller than the measurement error of the cladding center position, the geometric centers of some (at least two) of the cores 12 may be used as the calculated centers 11a.

[0091] The upper part of Figure 21 shows the complementary cumulative distribution of connection loss to the master when the core positions of the eight outer cores are aligned to the ideal core positions, and the cladding center is used as the calculated center 11a. The lower part of Figure 21 shows the complementary cumulative distribution of connection loss to the master when the core positions of the eight outer cores are aligned to the ideal core positions, and the geometric centers of the four cores closest to the center are used as the calculated center 11a. Referring to Figure 21, in both cases, the tails of the complementary cumulative distribution are shorter in the embodiment compared to the comparative example. In addition, even when the geometric centers of some of the cores 12, rather than all of them, are used as the center 11a, the tails of the complementary cumulative distribution are shorter compared to when the cladding center is used as the center 11a, thus reducing connection loss.

[0092] The upper part of Figure 22 shows the complementary cumulative distribution of random connection loss when the core positions of the eight outer cores are aligned to the ideal core positions, and the cladding center is used as the calculated center 11a. The lower part of Figure 22 shows the complementary cumulative distribution of random connection loss when the core positions of the eight outer cores are aligned to the ideal core positions, and the geometric centers of the four cores closest to the center are used as the calculated center 11a. Referring to Figure 22, in all cases, the tail of the complementary cumulative distribution is shorter in the examples compared to the comparative example. In Example 2, the value of the complementary cumulative distribution is smaller than that of the comparative example even at small connection loss values. Referring to the lower part of Figure 22, Example 4 appears to have a longer tail than the comparative example. If we plot down to the region where the probability of the complementary cumulative distribution is even lower, the tail of the complementary cumulative distribution is shortest in Example 4. In addition, even when the geometric centers of some cores, rather than all cores, are used as the center 11a, the tail of the complementary cumulative distribution is shorter and connection loss can be reduced compared to when the cladding center is used as the center 11a.

[0093] An example of a second embodiment will be described in which rotational alignment is performed so that each of the multiple core positions of the first MCF10A and each of the multiple core positions of the second MCF10B are brought closer to each other. First, the case in which the four design core positions are (20 μm, 20 μm), (-20 μm, 20 μm), (-20 μm, -20 μm), and (20 μm, -20 μm) in a Cartesian coordinate system with the origin at centers 11aa and 11ab, and the cladding centers are at centers 11aa and 11ab will be described. Figure 23 is a graph showing the complementary cumulative distribution of random connection loss for the comparative example and examples 2 and 4 in this case. Referring to Figure 23, the tails of the complementary cumulative distribution are shorter in the examples compared to the comparative example. The same or similar results can be obtained even if the number or positions of the cores of the first MCF10A and the second MCF10B are different.

[0094] Figure 24 is a graph showing the complementary cumulative distribution of random connection loss for comparative examples and Examples 2 and 4, where the geometric centers of multiple cores 12A are set to center 11aa and the geometric centers of multiple cores 12B are set to center 11ab. Referring to Figure 24, the tails of the complementary cumulative distribution are shorter in the Examples compared to the Comparative Example. The same or similar results can be obtained even if the number or position of cores in the first MCF 10A and the second MCF 10B are different. In addition, by setting the geometric centers of multiple cores to centers 11aa and 11ab, the tails of the complementary cumulative distribution are shorter than when the cladding centers are set to centers 11aa and 11ab, and connection loss can be reduced.

[0095] In the embodiments described above, the cladding center or the geometric center of the multiple cores was designated as center 11a (or centers 11aa, 11ab). However, regardless of whether the cladding center or the geometric center of the multiple cores is designated as center 11a, rotational alignment and translation (XY alignment) may be performed so that the core displacement amount D satisfies any of the above conditions (Embodiments 1 to 4). Figure 25 is a graph showing the complementary cumulative distribution of random connection loss for the comparative example and Examples 2 and 4 in such cases. Referring to Figure 25, the tails of the complementary cumulative distribution are shorter in the examples compared to the comparative example. The same or similar results can be obtained even if the number or position of the cores in the MCF 10 is different. The same or similar results can be obtained even if the number or position of the cores in the first MCF 10A and the second MCF 10B is different. In addition, by performing rotational and translational alignment (XY alignment) independently of the cladding center and the geometric centers of the multiple cores, the tails of the complementary cumulative distribution become shorter and connection losses can be reduced compared to when the geometric center of the multiple cores is center 11a (or centers 11aa, 11ab) and when the cladding center is center 11a (or centers 11aa, 11ab).

[0096] The method for centering a multicore optical fiber, the method for manufacturing an optical device, the method for manufacturing an optical module, and the centering device for a multicore optical fiber according to this disclosure are not limited to the embodiments described above, and various other modifications are possible. For example, the number and arrangement of cores in a multicore optical fiber, and the distribution of variations in the actual core positions are not limited to those described in the embodiments above.

[0097] 10...Multicore optical fiber (MCF) 10A...First MCF 10B...Second MCF 11, 11A, 11B...End face 11a, 11aa, 11ab...Center 11C...Butt face 12, 12A, 12B...Core 12a, 12aa, 12ab...Center point 12b...Ideal core position 13, 13A, 13B...Cladding 20, 30...Alignment device 21, 31...Storage unit 22, 32...Calculation unit 24, 34...Alignment unit 25, 35...Alignment control unit 40, 42...Optical device 41...Optical component 50A, 50B...Optical module 51, 52...Optical fiber connection component D...Core position misalignment amount Q...Square ST11, ST21...Acquisition step ST12, ST22...Determination step ST13, ST23...Calculation steps ST14, ST24...Rotation centering steps

Claims

1. A method for centering a multicore optical fiber, comprising:

1. Assuming that the calculated center position misalignment is zero, calculating the amount of core misalignment from the ideal core position, which is the design core position or the nominal core position, or the connection loss amount obtained from the amount of core misalignment, for at least two of the multiple cores of the multicore optical fiber; and 2. Rotating the multicore optical fiber such that the sum of the m-th power of the core misalignment amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core misalignment amount of the at least two cores is minimized.

2. The method for centering a multicore optical fiber according to claim 1, further comprising determining the calculated center position before performing the calculation, wherein the calculation determines the calculated center position based on the position of at least one of the multiple different refractive index elements having different refractive indices from the cladding, the position of the cladding center, or the arrangement of elements of an optical fiber connector that holds the tip of the multicore optical fiber for connection with other optical components.

3. A method for aligning a first multicore optical fiber and a second multicore optical fiber, comprising: assuming that the positional misalignment of the calculated center positions of the first multicore optical fiber and the second multicore optical fiber is zero, calculating the amount of core misalignment between corresponding cores of the first multicore optical fiber and the second multicore optical fiber, or the connection loss amount obtained from the amount of core misalignment, with respect to at least two of the multiple cores that each of the first multicore optical fiber and the second multicore optical fiber has; and performing rotational alignment of at least one of the first multicore optical fiber and the second multicore optical fiber so as to minimize the sum of the m-th power of the core misalignment amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core misalignment amount of the at least two cores.

4. The method for centering a multicore optical fiber according to claim 3, further comprising determining the calculated center position before performing the calculation, wherein in each of the first multicore optical fiber and the second multicore optical fiber, the calculated center position is determined based on the position of at least one element with a different refractive index from a plurality of elements with different refractive indices from the cladding, or the position of the cladding center.

5. A method for centering a multicore optical fiber according to any one of claims 1 to 4, wherein m is 4 or more.

6. A method for centering a multicore optical fiber according to any one of claims 1 to 4, wherein m is 4.

7. A method for centering a multicore optical fiber according to any one of claims 1 to 6, wherein, in the calculation, the amount of core misalignment or the amount of connection loss is calculated for all of the multiple cores, and in the rotational alignment, the rotational alignment is performed such that the sum of the m-th power of the amount of core misalignment of all the cores, the sum of the n-th power of the amount of connection loss of all the cores, or the largest amount of core misalignment of all the cores is minimized.

8. The method for centering a multicore optical fiber according to claim 2 or 4, wherein the at least one element with a different refractive index includes at least one core from the plurality of cores.

9. The method for centering a multicore optical fiber according to claim 2 or 4, wherein the at least one element with a different refractive index includes all of the cores of the plurality of cores.

10. A method for manufacturing an optical device comprising a multicore optical fiber and an optical component connected to the multicore optical fiber, the method comprising a method for centering a multicore optical fiber according to claim 1 or claim 2.

11. A method for manufacturing an optical module comprising a multicore optical fiber and an optical fiber connector component for holding the tip of the multicore optical fiber for connection between the multicore optical fiber and other optical components, the method comprising the method for centering a multicore optical fiber according to claim 1 or claim 2.

12. A method for manufacturing an optical device comprising a first multicore optical fiber and a second multicore optical fiber connected to each other, the method comprising the method for centering multicore optical fibers according to claim 3 or claim 4.

13. A multicore optical fiber alignment device comprising: a calculation unit that calculates the amount of core misalignment from the ideal core position, which is the design core position or the nominal core position, or the connection loss amount obtained from the amount of core misalignment, for at least two of the multiple cores of the multicore optical fiber, assuming that the calculated center position misalignment is zero; and a centering control unit that rotates and aligns the multicore optical fiber so that the sum of the m-th powers of the core misalignment amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th powers of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core misalignment amount of the at least two cores is minimized.

14. The centering device for a multicore optical fiber according to claim 13, wherein the calculation unit determines the calculated center position based on the position of at least one of the multiple different refractive index elements having different refractive indices from the cladding, the position of the cladding center, or the arrangement of elements of an optical fiber connector that holds the tip of the multicore optical fiber for connection with other optical components.

15. A device for aligning a first multicore optical fiber and a second multicore optical fiber, comprising: a calculation unit that assumes that the positional misalignment of the calculated center positions of the first multicore optical fiber and the second multicore optical fiber is zero, and calculates the amount of core misalignment between corresponding cores of the first multicore optical fiber and the second multicore optical fiber, or the amount of connection loss obtained from the amount of core misalignment, with respect to at least two of the multiple cores that each of the first multicore optical fiber and the second multicore optical fiber has; and a centering control unit that performs rotational centering on at least one of the first multicore optical fiber and the second multicore optical fiber so as to minimize the sum of the m-th power of the core misalignment amounts of the at least two cores (where m is a number of 3 or more), the sum of the n-th power of the connection loss amounts of the at least two cores (where n is a number of 2 or more), or the largest core misalignment amount of the at least two cores.

16. The centering device for a multicore optical fiber according to claim 15, wherein the calculation unit determines the calculated center position in each of the first multicore optical fiber and the second multicore optical fiber based on the position of at least one of a plurality of elements with different refractive indices that have different refractive indices from the cladding, or based on the position of the cladding center.

Images