Null Hartmann test for the fabrication of large aspheric surfaces

Most aspheric mirrors have been tested by the null lens or computer-generated hologram method. This approach, however, requires that the shape of the surface be similar to the target shape; otherwise testing may not be possible or correct. The Hartmann test has an advantage in that it has a larger dynamic range than a general interferometer, which means that the surface can be tested beginning at an early stage of the polishing process. We suggest use of the null Hartmann test in conjunction with a phase-shifting interferometer for the measurement of a 0.9-m aspheric concave mirror. This setup was able to measure the surface with a large surface form error as well as with a small error without sacrificing any measurement accuracy. Using this setup, we have successfully polished a surface to remove approximately 1 microm of peak-to-valley wave-front error of a total of 39 microm of error during 1 month of polishing.