| Popis: |
Ultra-stable laser cavities are the core components of today’s most precise measurement instruments. To design and realize an ultra-stable laser cavity, one must suppress all kinds of noise below the desired frequency stability. The thermal noise of the cavity spacer sets a fundamental limit for the possible stability, especially for cavities operating at room temperature. We present a comprehensive numerical study of thermal cavity spacer noise, accounting for all relevant geometrical parameters. Accelerated by a vast amount of simulation data, a neural network is trained to predict spacer noise precisely for a wide range of all relevant cavity parameters. Based on this neural network, we identified design rules to reduce spacer noise by more than 50% by only changing the cavity geometry. Optimizing the cavity geometry and the material combination can achieve a noise reduction of more than two orders of magnitude without the infrastructural effort of a cryogenic environment. |
