A standard way of tolerancing optical elements or systems is to perform a Monte Carlo based analysis within a common optical design software package. Although, different weightings and distributions are assumed they are all counting on statistics, which usually means several hundreds or thousands of systems for reliable results. Thus, employing these methods for small batch sizes is unreliable, especially when aspheric surfaces are involved. The huge database of asphericon was used to investigate the correlation between the given tolerance values and measured data sets. The resulting probability distributions of these measured data were analyzed aiming for a robust optical tolerancing process.
For a lot of applications like spectrometer and high power laser roughness as an important parameter has been discussed over and over again. Especially for high power systems the surface quality is crucial for determining the damage threshold and therefore the field of application. Above that, it has often been difficult to compare roughness measurements because of different measurement methods and the usage of filters and surface fits. Measurement results differ significantly depending on filters and especially on the measured surface size. Insights will be given how values behave depending on the quality of surface and the size of measured area.
Many applications require a high quality of roughness in order to reduce scattering. Some of them in order to prevent from damage like high power laser applications. Others like spectrometers seek to increase the signal-to-noise ratio. Most of them have already been built with spherical surfaces. With higher demands on efficiency and more sophisticated versions aspherical surfaces need to be employed. Therefore, the high requirement in roughness known from spherical surfaces is also needed on aspherical surfaces. For one thing, the constant change of curvature of an aspherical surface accounts for the superior performance, for another thing, it prevents from using classical polishing technics, which guarantied this low roughness. New methods need to be qualified. In addition, also results of a new manufacturing process will be shown allowing low roughness on aspheric even with remarkable departure from the best fit sphere.