To understand the loss limitations of a splice between a hollow-core fiber and a conventional fiber, we use a numerical model to calculate the expected coupling loss between the NKT Photonics' HC-1550-02 fiber and a single-mode fiber (SMF) of arbitrary step-index profile. When the SMF parameters are optimized, the splice loss is predicted to be as low as~0.6 dB. This minimum is believed to be largely due to mode-shape mismatch. These predictions are confirmed experimentally by optimizing the splice loss between this photonic-bandgap fiber and five SMFs with different mode-field diameters (MFDs) and V numbers. With the SMF-28 fiber, the measured loss is 1:3 dB, in excellent agreement with theory. Using a SMF with parameters close to the optimum values (MFD ¼ 7:2 μm and V ¼ 2:16), this loss was reduced to a new record value of 0:79 dB. . A lower splice loss of 1 dB has also been demonstrated by applying longitudinal pressure to these two fibers prior to applying the arc [4]. In contrast, a simple overlap calculation using a Gaussian mode approximation based on the mode-field diameter of the two fibers (∼7:5 and ∼10:4 μm, respectively) predicts that the buttcoupling loss should be 0:46 dB [3]. Although this 0.5 to 1 dB difference has been tentatively explained by differences in the mode shapes, there is still a need to conduct thorough simulations to explain this measured loss. It is also important to explore the possibility of achieving a lower splice loss by selecting a solid-core fiber with a mode-field diameter (MFD) better matched to that of the hollow-core fiber.In this Letter, we use a numerical model previously reported [5] to calculate the exact fields of the HC-1550-02 fiber's fundamental mode, and with them we compute the expected coupling loss between this fiber and an arbitrary step-index SMF. This analysis predicts that, when the SMF parameters are optimized (MFD ≈ 7:25 μm and V ¼ 2:405), the splice loss should be as low as ∼0:6 dB. We confirm these predictions by optimizing the splice loss between this PBF and five SMFs with different MFDs and V numbers. With the SMF-28 fiber, we measured a loss of 1:3 dB, in good agreement with theory. Using a SMF with parameters close to the optimum values (MFD ¼ 7:2 μm and V ¼ 2:16), we reduced the splice loss to 0:79 dB. This is the lowest value reported to date for a splice between a hollow-core and a solid-core fiber using an arc splicer.The field transmission (t) and reflection (r) coefficients at a butt-coupled junction between a SMF and a PBF can be described in terms of the normalized vector electric and magnetic fields E i and H i of the HE 11 mode of the SMF, and the corresponding fields E t and H t of the PBF's fundamental mode. These modes are normalized to carry 1 W power, orwhere A is the cross-sectional area of the fiber. A similar expression applies for the SMF mode. Neglecting coupling into higher-order modes of the PBF, the continuity of the transverse fields at the interface between the fibers imposeswhere the subscript T represents the transvers...
Abstract-A full-vectorial finite-difference scheme utilizing the hexagonal Yee's cell is used in this paper to analyze the modes of photonic-bandgap fibers with C 6 symmetry. Because it respects the fiber's native symmetry, this method is free from any numerical birefringence. We also incorporate in it techniques for reducing the memory requirement (up to 3 to 4 times) and computational time, in particular by exploiting some of the symmetry properties of these fibers. Using sub-pixel averaging, we demonstrate quadratic convergence for the fundamental mode's effective index dependence on spatial resolution. We show that this method can be used to calculate the beat length of PBFs in which a birefringence is introduced by applying a small unidirectional stretch to the fiber cross section along one of its axes. Abrupt variations of the modeled fiber geometries with spatial resolution lead to oscillatory beat length convergence behavior. We can obtain a better estimate for beat length by averaging these oscillations. We apply a strong perturbation analysis to the fiber's unperturbed mode, calculated by our finite-difference method, to perform this averaging in a rigorous way. By fitting a polynomial to the predicted beat lengths as a function of grid spacing, we obtain an accurate estimate of the beat length at zero grid spacing. Reasonable convergence for the beat length is observed using a single processor with 8 GB of memory.
Abstract-Multiple-scale analysis is employed for the analysis of plane wave refraction at a nonlinear slab. It will be demonstrated that the perturbation method will lead to a nonuniformly valid approximation to the solution of the nonlinear wave equation. To construct a uniformly valid approximation, we will exploit multiplescale analysis.Using this method, we will derive the zerothorder approximation to the solution of the nonlinear wave equation analytically. This approximate solution clearly shows the effects of self-phase modulation (SPM) and cross-phase modulation (XPM) on plane wave refraction at the nonlinear slab. In fact, the obtained zeroth-order approximation is very accurate and there is not any need for derivation of higher-order approximations. As will be shown, the proposed method can be generalized to the rigorous study of nonlinear wave propagation in one-dimensional photonic band-gap structures.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.