Options
2017
Conference Paper
Titel
Characterization and optimization of ultra-precision machine tools and processes by time resolved PSD signature of micro/nano structured optical surfaces
Abstract
Ultra-precision machines are the most accurate class of machine tools. The achievable precision is nowadays not limited by the available measuring technology, but rather by disturbance variables. The disturbances result primarily from thermal effects and vibrations as well as pulsations. Particularly in the case of the production of diffraction gratings with a structural height h < 100 nm, the ultra-precision machining gets to its limits. The challenge is growing with regard to the production of diffractive structures on curved surfaces, due to the necessary degrees of freedom and the number of involved machine axes. This limit can be shifted by means of targeted measurement and an extended knowledge of the disturbance variables. Thermal effects generate shape deviations with a low spatial frequency of several microns and a shape deviation up to several 100 nm and limit the absolute accuracy. Regarding shaping processes, vibrations transmitted into the work pieces by the tool generate ripple with a height of usually h < 20 nm with a mid/high spatial frequency in the range of k < 0.125 mm-1 (cutting speed f = 500 mm/min) and essentially limit the achievable minimum structure size. With smaller structures, the importance of vibration reduction increases. The causal clarification of the ripple is often not trivial, but can be supported by the use of PSD (power spectral density). The actual resulting spatial frequency caused by vibration is directly related to the hertzian frequency depending on the cutting speed. A computation of a time resolved PSD from the usually spatial resolved PSD simplifies the causal clarification. Each machine has a very specific PSD signature, which can be determined by shaping processes and WLI (white light interferometry) measurement. In the case of machines with a variety of axes, each axis configuration may have its own PSD signature. A compromise of possible degrees of freedom and resulting PSD function allows the selection of the most suitable configuration. Changes to this specific signature allow the detection of machine faults. Using the example of blazed grating structures, it is shown that this method is a powerful tool for the evaluation and optimization of ultra-precision machines. The surface-dominant frequency bands can be assigned to mechanical and electromagnetic sources, which can be eliminated in some cases. In order to eliminate mechanical and electromagnetic vibrations the method is explained on the basis of selected examples. The remaining time resolved PSD signature of a machine enables statements on the suitability of the machine for the production of specific optics. Regarding diffractive structures, it has been shown that the production of efficient blazed gratings is limited by the machine's natural frequencies, besides thermal effects. In addition, the knowledge about the frequencies gained from the time resolved PSD can also be used to shift the spatial frequencies by use of an appropriate cutting speed. The experiments were carried out on a modified, highly temperature stabilized and active vibration-damped ultra-precision machine LT-Ultra MMC1100 and represent the limitations and possibilities of the production of imaging diffractive optics for spectroscopic applications.