Modeling Coherent Beam Combining Feasibility with Supercontinuum Sources

Abstract

Using numerical modeling methods, this thesis investigates the feasibility of coherent beam combining (CBC) using supercontinuum light generated by independent all-normal dispersion photonic crystal fibers (ANDi PCFs). Such a system would require a high degree of temporal coherence and near zero relative phase error over a large spectral bandwidth between multiple channels. Coherent supercontinuum can be achieved when pumping ANDi PCFs with a low noise mode locked source emitting very short pulses (∼200 fs or less). These conditions prevent noise-seeded processes like modulation instability and result in a spectrum generated solely through coherent effects such as self phase modulation. The generalized nonlinear Schrödinger equation is used to calculate the spectral and temporal pulse evolution through ANDi PCFs with realistic noise applied. The resulting shot-to-shot coherence and relative phase are calculated and used to evaluate the expected coherent combining efficiency. It was found that >90% combining efficiency over a 850-1250 nm bandpass could theoretically be achieved between two channels when the relative intensity noise limited (RIN) to <1% and all other parameters are identical. Similar performance can be achieved over a smaller bandpass of 450-600 nm when pumping at 515 nm in the presence of 2% RIN. Additionally, tolerances are established for static differences in peak powers, pulse durations, and fiber lengths between channels. Finally, relative phase compensation after supercontinuum generation is explored by using bulk glass dispersion compensation. Least squares optimizations are performed to determine the ideal thicknesses of various glass types to provide the best dispersion compensation for a particular relative phase shape. Results are reported for seeding the fiber near the minimum dispersion wavelength at 1030 nm and for visible supercontinuum generation at 515 nm.