Abstract

The inhomogeneity of a window material can be observed by immersing the window into a liquid between two planes, or polishing both surfaces of the window to flats much better than the amount of inhomogeneity. These two methods are inconvenient for most applications. A new method which can measure the inhomogeneity independent of the quality of the optical surfaces of the window will be presented. This technique requires four measurements. First the two surfaces of the window are measured, next the window is measured in transmission followed by a reference flat, and finally the window is removed and a measurement is taken with the reference flat. These measurements are then combined to find the inhomogeneity of the window. The method requires that both surfaces be polished; however, there can be several fringes of deviation from flatness. A small wedge between the two surfaces will eliminate effects due to spurious reflections. One advantage of this technique is the removal of system error due to the interferometer. If the rms error due to random noise (i.e. repeatability) is Δ for a phase measurement, the resulting error in the homogeneity measurement is 1.6Δ for an index of refraction of 2. This pa per will describe this technique in detail and show experimental results.

© 1990 Optical Society of America