Fabrication and testing of large flats

Abstract

Flat mirrors of around 1 meter are efficiently manufactured with large plano polishers and measured with Fizeau interferometry. We have developed technologies and hardware that allow fabrication and testing of flat mirrors that are much larger. The grinding and polishing of the large surfaces uses conventional laps driven under computer control for accurate and systematic control of the surface figure. The measurements are provided by a combination of a scanning pentaprism test, capable of measuring power and low order irregularity over diameters up to 8 meters, and subaperture Fizeau interferometry. We have developed a vibration insensitive Fizeau interferometer with 1 meter aperture and software to optimally combine the data from the subaperture tests. These methods were proven on a 1.6 m flat mirror that was finished to 6 nm rms irregularity and 11 nm rms power.

Influence of both angle and position error of pentaprism on accuracy of pentaprism scanning system

“…The rotation take the order Yaw→Roll→Pitch. Assuming that the tolerances of 8 angle parameters take the same value ±5′, t1, t2, t5, t6, t7, t8, t9, t10 take a random, a uniformly selected value in their tolerance range respectively.Yaw3= unifrnd(-5/60, 5/60)°, Roll3= unifrnd(-5/60, 5/60)°, Pitch3= unifrnd(-5/60, 5 Next we analyze the weight of parameters, we narrow the tolerance range down gradually and observe the effect on the measurement results and for examples, we gradually narrow the tolerance range of Yaw1, Roll1 and Pitch1 respectively, as shown in Figure 18, 19 and 20. We can use this method to estimate and calculate the measurement accuracy of the pentaprism scanning system when the tolerances of its error parameters are given and analyze the influence of each error parameter and the synergy of multiple error parameters.…”