Fig. 1. The average power versus soliton order N of 1.5-m high-power ultrafast fiber lasers. Triangles and hexagrams respectively denote CPA- and NCPA-based fiber lasers. A more comprehensive survey of related references is provided in Table 3, Appendix A. The dashed-dotted line and dashed line correspond to a 10-m-core double-cladding fiber laser system (, ) and a 25-m-core large-mode-area fiber laser system (, ), respectively (assuming ). CPA, chirped pulse amplification; NCPA, nonlinear chirped pulse amplification; TMI, transverse mode instability.

Fig. 2. Numerical simulations of the amplified signals with different pre-chirping group delay dispersions (GDDs). (a) The contour plot of optical spectra with varying pre-chirping dispersion. SSFS, soliton self-frequency shift. (b) The corresponding pulsewidth variation. Regimes I, II and III are designated according to the spectral-temporal characteristics, and the pulse amplifications governed by the Raman effect, the soliton effect and weak nonlinearity, respectively, are identified.

Fig. 3. Schematic diagram of the experimental setup. Dispersion-compensation fiber (DCF) is employed to perform pre-chirping dispersion management. SESAM, semiconductor saturable absorber mirror; DF, dielectric film; EYDF, Er-Yb-doped fiber; PC, polarization controller; WDM, wavelength-division multiplexer; SM-LD, single-mode laser diode; ISO, isolator; EDF, Er-doped fiber; MM-LD, multimode laser diode; SPC, signal-pump combiner; DC-EYDF, double-cladding EYDF; OC, optical coupler; PM-DC-EYDF, polarization-maintaining DC-EYDF; PLMA-DC-EYDF, polarization-maintaining large-mode-area DC-EYDF; QBH, quartz block head; PM, polarization-maintaining.

Fig. 4. The characterization of the seed. (a) The optical spectrum. (b) The radio-frequency (RF) spectrum measured at a resolution bandwidth (RBW) of 10 Hz. (c) The RF spectrum measured at a 25-GHz span at an RBW of 30 kHz. (d) The oscilloscopic trace. Here, the pulse train at a 10.6-GHz repetition rate is viewed as a sinusoidal waveform due to the limitation of the electrical bandwidth. The inset shows the pulse trace in a wider span of 10 s.

Fig. 5. (a) The output power of the main fiber amplifier as a function of the pump power. (b) The autocorrelation trace measured at the maximum output power of 106.4 W when using a 32-m-long DCF.

Fig. 6. The operation regimes of the high-power fiber laser system by employing different lengths of DCFs. (a) Thirty simulated optical spectra operated in the Raman-effect-dominated regime (regime I) with different random Raman noise (grey curves), the average simulated optical spectrum (black curve) and the degree of coherence (orange curve). (b) Experimental optical spectrum operated in regime I. The inset shows an approximately 4-dB spectral fringe contrast suggesting a degraded coherence (~0.43) at the central spectral region of the signal. (c) The degree of coherence in the soliton-effect-dominated regime (regime II), wherein the central spectral region of the signal shows a good quality of coherence. (d) Experimental optical spectrum operated in regime II. (e) Experimental optical spectrum and autocorrelation trace operated in the weakly nonlinear regime (regime III). (f) The autocorrelation trace operated in regime III. The less-broadened optical spectrum and ps-level pulsewidth indicate weak nonlinearity that is not sufficient for soliton-effect compression.

Fig. 7. The intermodal modulation instability (IM-MI) that potentially existed in the LMA fiber-based main fiber amplifier. (a) The transverse modes supported by the 25-m-core LMA gain fiber, that is, LP₀₁, LP₁₁ and LP₂₁ in this case. In the calculation, the refractive index difference between the core and cladding is set to 0.0035. (b) The calculated first- and second-order dispersion curves for different linearly-polarized modes. (c) The optical spectra at the average powers of 80 and 100 W. (d) The calculated gain spectra of MI and IM-MIs resulting from the nonlinear interactions between the LP₀₁–LP₁₁ and LP₀₁–LP₂₁ mode pairs.

Fig. 8. The influence of intermodal four-wave mixing (IM-FWM) on the output performance of the high-power fiber laser system. (a) The IM-FWM-mediated energy transfer from transverse mode LP₀₁ to LP₂₁. With the presence of the modal dispersion, the pulses of transverse modes LP₀₁ and LP₂₁ walk off from each other, and form a pulse doublet separated by through the mode coupling. The relevant autocorrelation trace is provided as an inset on the right-hand side. (b) The optical spectrum measured with the maximum output power if an inappropriate coiling scheme is used in the experiment. (c) Closeup of the intrinsic longitudinal mode (left) and spectral structure resulting from the pulse doublet pattern (right).