Bend losses are a frequently encountered problem in the context of waveguides, and in particular in fiber optics, since fibers can be easily bent. Bend losses mean that optical fibers exhibit additional propagation losses by coupling light from core modes (guided modes) to cladding modes when they are bent. Typically, these losses rise very quickly once a certain critical bend radius is reached. This critical radius can be very small (a few millimeters) for fibers with robust guiding characteristics (high numerical aperture), whereas it is much larger (often tens of centimeters) for single-mode fibers with large mode areas.
Generally, bend losses increase strongly for longer wavelengths, although the wavelength dependence is often strongly oscillatory due to interference with light reflected at the cladding/coating boundary, and/or at the outer coating surface. The increasing bend losses at longer wavelengths often limit the usable wavelength range of a single-mode fiber. For example, a fiber with a single-mode cut-off wavelength of 800 nm, as is suitable for operation in the 1-μm region, may not be usable at 1500 nm because they would exhibit excessive bend losses. Note that even without macroscopic bending of a fiber, bend losses can occur as a result of microbends, i.e., microscopic disturbances in the fiber, which can be caused by imperfect fabrication conditions.
In multimode fibers, bend losses are usually strongly mode-dependent. The critical bend radius is typically larger for higher-order transverse modes. By properly adjusting the bend radius, it is possible to introduce significant losses for higher-order modes without affecting the lowest-order mode. This can be useful e.g. for the design of high-power fiber amplifiers and fiber lasers where a higher effective mode area can be achieved when using a fiber with multiple transverse modes.
Photonic crystal fibers can have very low bend losses even far beyond the single-mode cutoff wavelength. Therefore, they can be “endlessly single-mode”, i.e., they exhibit usable single-mode characteristics over a very large wavelength range.
Estimating Bend Losses
For estimating the magnitude of bend loss, the equivalent index method  can be used. The basic idea behind this technique is to calculate the mode distributions for an effective index which contains a term accounting for the modified path lengths at different transverse positions. An elasto-optic correction term (taking into account local modifications of the refractive index by mechanical stress) leads to an effectively weaker “tilt” of the refractive index profile than when considering the geometrical effect alone [3, 9].
Such a method of calculating bend losses is convenient and usually a good approximation, provided that there is no light reflected e.g. from the outer cladding surface back to the fiber core. More sophisticated models (see e.g. Ref. ) can include such effects, and thus predict the full wavelength dependence, but are complicated to handle.
In some application areas of fibers, and particularly in optical fiber communications, it is of interest to have a relatively bend-insensitive fibers. This is particularly important in the area of fiber to the home (FTTH), where it is problematic to demand that any tight bending is avoided in the fiber installation. The G.657 standard for telecom fibers defines the characteristics of such fibers, with the main categories A for use in access networks and B for short distances, but with even stronger bending tolerance – in category B3 allowing a bend radius down to only 5 mm. For comparison, the more general category G.652 allows a bend radius only down to 30 mm.
Bend-insensitive fibers can be made with different designs. It is common to have trench-assisted fiber designs, containing a ring (trench) with relatively low refractive index around the fiber core. Another possibility are certain designs of photonic crystal fibers, which may be called hole-assisted.
Note that there are trade-offs between bend insensitivity and other properties, such as low propagation losses in the straight form. However, lowest propagation losses are not essential e.g. for FTTH, where only a relatively short transmission distances need to be realized.
Bend Losses in Photonic Integrated Circuits
Bend losses are important not only in fiber optics, but also in the context of photonic integrated circuits. Compact circuits designs often require strong bending of waveguides on such chips, with desired values of the bend radius of a few microns only, but limited by the issue of bend losses. For strong bending and thus particularly compact designs, waveguides with a relatively high numerical aperture, i.e., with a large refractive index contrast (i.e., a high numerical aperture), are needed, e.g. based on silicon-on-insulator (SOI) technology. Those, however, are tentatively more sensitive to propagation losses by scattering at non-perfect boundaries between core and cladding.
|||D. Marcuse, “Curvature loss formula for optical fibers”, J. Opt. Soc. Am. 66 (3), 216 (1976); https://doi.org/10.1364/JOSA.66.000216|
|||D. Marcuse, “Field deformation and loss caused by curvature of optical fibers”, J. Opt. Soc. Am. 66 (4), 311 (1976); https://doi.org/10.1364/JOSA.66.000311|
|||R. Ulrich et al., “Bending-induced birefringence in single-mode fibers”, Opt. Lett. 5 (6), 273 (1980); https://doi.org/10.1364/OL.5.000273|
|||D. Marcuse, “Influence of curvature on the losses of doubly clad fibers”, Appl. Opt. 21 (23), 4208 (1982); https://doi.org/10.1364/AO.21.004208|
|||S. J. Garth, “Birefringence in bent single-mode fibers”, IEEE J. Lightwave Technol. 6 (3), 445 (1988); https://doi.org/10.1109/50.4022|
|||L. Faustini and G. Martini, “Bend loss in single-mode fibers”, IEEE J. Lightwave Technol. 15 (4), 671 (1997); https://doi.org/10.1109/50.566689|
|||J. M. Fini, “Bend-resistant design of conventional and microstructure fibers with very large mode area”, “Opt. Express 14 (1), 69 (2006)”; https://doi.org/10.1364/OPEX.14.000069|
|||R. W. Smink et al., “Bending loss in optical fibers – a full-wave approach”, J. Opt. Soc. Am. B 24 (10), 2610 (2007); https://doi.org/10.1364/JOSAB.24.002610|
|||R. T. Schermer, “Improved bend loss formula verified for optical fiber by simulation and experiment”, IEEE J. Quantum Electron. 43 (10), 899 (2007); https://doi.org/10.1109/JQE.2007.903364|
|||R. T. Schermer, “Mode scalability in bent optical fibers”, Opt. Express 15 (24), 15674 (2007); https://doi.org/10.1364/OE.15.015674|
|||A. Argyros et al., “Bend loss in highly multimode fibres”, Opt. Express 16 (23), 18590 (2008); https://doi.org/10.1364/OE.16.018590|
|||J. M. Fini, “Large mode area fibers with asymmetric bend compensation”, Opt. Express 19 (22), 21866 (2011); https://doi.org/10.1364/OE.19.021866|
|||J. M. Fini, “Intuitive modeling of bend distortion in large-mode-area fibers”, Opt. Lett. 32 (12), 1632 (2007); https://doi.org/10.1364/OL.32.001632|
|||C. Schulze et al., “Mode resolved bend loss in few-mode optical fibers”, Opt. Express 21 (3), 3170 (2013); https://doi.org/10.1364/OE.21.003170|
|||R. Paschotta, tutorial on "Passive Fiber Optics", Part 7: Propagation Losses|
|||R. Paschotta, case study on bend losses of a fiber|
Questions and Comments from Users
Here you can submit questions and comments. As far as they get accepted by the author, they will appear above this paragraph together with the author’s answer. The author will decide on acceptance based on certain criteria. Essentially, the issue must be of sufficiently broad interest.
Please do not enter personal data here; we would otherwise delete it soon. (See also our privacy declaration.) If you wish to receive personal feedback or consultancy from the author, please contact him, e.g. via e-mail.
By submitting the information, you give your consent to the potential publication of your inputs on our website according to our rules. (If you later retract your consent, we will delete those inputs.) As your inputs are first reviewed by the author, they may be published with some delay.