Abstract
For the sake of mass production of a large and severe aspherical optics, we have developed a new stitch algorithm named ARSA (Approximated reference shape algorithm) which connects plural partial measurements and combines them into a whole shape. In the case of a free-form measurement machine utilizing a probe[6,7], measured data is sparse and three-dimensional points. So, it's difficult to precisely determine the overlapping area because of very limited number of points, and it's also difficult to use a conventional stitch algorithm based on the minimization of the mismatch in the overlapping area. The advantage of ARSA is that it is not influenced by the overlapping area. In the previous paper [8], we discussed the basic idea of the ARSA that utilizes approximated reference shape, and its performance to measure a high and mid-frequency range. This paper evaluates a performance to measure a low frequency by simulation. The simulated measurement data are three sets of fan-shaped partial measurement of a hyperboloid, full aperture of 1200 mm and a data pitch of 4 mm. In the case of quadratic surface, we found that the position error affects significantly to the reference shape, and also found that simple way of excluding Zernike 7 and 8 terms from the reference shape results in sufficiently small calculation error.