Fiber splicing is a fundamental procedure in constructing all-fiber communication networks or fiber laser systems. Mode coupling occurring at the splicing point is a major cause of modal crosstalk and beam degradation in few-mode fibers. The performance of the splicing equipment significantly influences the mode coupling characteristics. Testing mode degradation at the splicing point is essential for determining optimal splicing parameters. In few-mode fiber splicing, mode coupling occurs simultaneously at the launching end and the splicing point, leading to a reciprocal mode coupling regime along the cascade fiber path. The traditional time-domain measurement methods require long transmission paths to accumulate total time delays of different modes and cannot distinguish or determine high-order modes, making it challenging to analyze mode coupling at the splicing point. Currently, fiber laser systems frequently involve few-mode fiber splicing, making it difficult to differentiate multiple-mode coupling events. To achieve a comprehensive understanding of mode properties at the splicing point, it is necessary to explore a modified method that can decouple and quantify mode coupling characteristics in few-mode fiber architectures.

We propose a modified spatially and spectrally (M-S²) resolving method that accounts for the reciprocal mode coupling regime in few-mode fiber splicing. This method allows for the separation and decoupling of different origins of mode coupling at the splicing point, facilitating quantifiable analysis of specific high-order modes. Numerical simulations are conducted with different mode coupling events and modal weights are conducted to evaluate the method’s effectiveness, showing strong potential for analyzing mode coupling characteristics at the splicing point. An experimental setup, as shown in Fig. 4, is constructed to investigate splicing. A tunable source, operating between 1070 nm and 1090 nm, injects single-mode laser beams into the fibers under test with a 0.05 nm wavelength interval. Each wavelength’s excited modes propagate through the fiber path and experience mode coupling at the splicing point. A 10 bit camera records the output field at each wavelength, generating an image sequence for mode property analysis. Finally, the effects of splicing parameters (e.g., arc power or duration) and fiber cleaving angles on mode coupling properties are studied for various fiber types. Investigated samples include fibers with identical or different cladding dimensions: 1) Splicing of 125 μm fibers, with dimensions of 10/125 μm and 15/125 μm and numerical apertures of 0.080 and 0.076. 2) Splicing of 125 μm and 250 μm fibers, with dimensions of 10/125 μm and 20/250 μm and numerical apertures of 0.080 and 0.112. 3) Splicing of 125 μm and 400 μm fibers, with dimensions of 10/125 μm and 20/400 μm and numerical apertures of 0.080 and 0.065. 4) Splicing of 400 μm fibers with identical parameters, including dimensions of 20/400 μm and a numerical aperture of 0.065.

Numerical simulations demonstrate that the M-S² resolving method effectively decouples and quantifies mode coupling at the splicing point. In the first case, the modal weights of E₁, E₂ and E₄are 99.0%, 0.5%, and 0.5%, respectively, corresponding to a theoretical multi-path interference (MPI) of -23.0 dB for E₂ and E₄. The MPI tested by the M-S² method is -24.8 dB. In the second case, the modal weights of E₁, E₂, and E₄are 99.0%, 0.8%, and 0.2%, respectively, corresponding to theoretical MPIs of -20.9 dB and -27.0 dB for E₂ and E₄. The MPIs tested by the M-S² resolving method are -22.8 dB and -28.7 dB, showing good agreement with theoretical values. For 125 μm fibers, the LP₁₁ mode content at the splicing point correlates strongly with ARC duration, increasing from -20.4 dB to -15.8 dB as arc duration extends from 1000 ms to 6000 ms. For 125 μm/250 μm fibers and 125 μm/400 μm fibers, the excited contents of azimuthal or radial modes show little correlation with arc parameters, while LP₁₁ mode evolution is highly correlated with the cleaving angle of the 125 μm fiber. For 400 μm fibers, mode coupling characteristics remain stable across a wide range of arc parameters, showing minimal sensitivity to arc parameter variation.

Using the modified spatially and spectrally resolving method, we investigate the influences of splicing parameters on mode coupling characteristics at the splicing point for various fiber types. By altering arc parameters at equal intervals, we quantitatively assess the key factors affecting mode coupling. Results indicate that for fibers with thin cladding thicknesses, mode coupling is significantly influenced by arc duration, which positively correlates with high-order mode content. For fibers with different cladding parameters, angular mode coupling is mainly affected by the cleaving angle of the thin fiber, while radial mode coupling is less affected by splicing parameters. Due to the offset of the discharge center and the fluctuation of arc behavior, the thermal effects on splicing point mode characteristics show random properties and greater cladding size differences lead to poorer splicing consistency. In contrast, thick fibers with identical cladding parameters exhibit less sensitivity to the variation of arc parameters, with stable mode coupling characteristics across a wide range of arc parameters. This work is of great significance for decoupling and quantitatively evaluating mode coupling characteristics at fiber splicing points and is particularly useful for mode properties analysis and beam quality optimization in high-power fiber laser systems.