Beam cleanup experiments for solid state laser with stochastic parallel gradient descent method

Sanhong Wang,Yonghui Liang,Qifeng Yu,Xiaojun Xu,Haotong Ma
DOI: https://doi.org/10.1117/12.835389
2009-01-01
Abstract:Thermal aberrations in high-power solid-state lasers degrade the beam quality and cause diffraction losses that reduce the output power. Adaptive optics system can be used to compensate for the wavefront distortions in real time to clean up the beam. But the conventional adaptive optics system based on wavefront sensing can't work well with the high-power laser beam cleanup because of the presence of branch points in the phase and the amplitude scintillation. However the other kind of adaptive optics system based on model-free optimization of a scalar beam-quality metric provides an attractive approach to perform the beam cleanup in smaller packages. Such an adaptive optics system with iteration rate of 100 Hz was built up. In this system a stochastic parallel gradient descent (SPGD) algorithm is implemented in a computer to control a deformable mirror for wavefront correction. Beam cleanup experiments with the SPGD AO system were conducted. The high-power laser beam was simulated by transmitting a low-power high-quality laser beam through an Nd: YAG-crystal power amplifier. The thermal aberrations of the beam could be changed by adjusting the pumping current of the power amplifier. The experimental results on the convergence and stability of this adaptive system for various conditions of dynamic aberrations are presented. © 2009 SPIE.
What problem does this paper attempt to address?