An optical system that utilizes visible light at nearly perpendicular incidence has been invented for use in testing the surface figures of nominally axisymmetric (paraboloidal, ellipsoidal, or conical) mirrors designed to function at grazing incidence. Such mirrors can be used as the focusing optical elements of x-ray telescopes and x-ray cameras. As explained below, the present system offers advantages over prior systems used to test such mirrors.
It is possible to test a grazing-incidence x-ray mirror by use of visible light at grazing incidence, but diffraction at visible wavelengths limits the achievable angular resolution to the order of one arc minute, whereas a resolution of 15 arc seconds or finer is typically needed for proper diagnosis of the surface figure. It is possible to test such a mirror at finer resolution by use of x rays at grazing incidence in the intended operational configuration, but testing in this way is of limited diagnostic value. Moreover, to prevent excessive absorption of x rays, x-ray testing must be performed in a vacuum chamber: this makes it difficult and time-consuming to manipulate the mirror and test equipment, thereby making testing expensive. The present system utilizes nearly perpendicular incidence to overcome the deleterious effect of diffraction, making it possible to obtain angular resolution of a few arc seconds with visible light and, hence, without need for a vacuum system.
The system (see Figure 1) includes the following components:
- A source of a parallel visible light beam of sufficient diameter,
- The nominally axisymmetric mirror under test,
- A prism that has (1) reflective conical faces with angles and diameter suitably matched to the mirror under test and (2) an axis of symmetry nominally parallel to the light beam,
- A focusing lens, and
- A screen or a charge-coupled-device video camera, located at the nominal focal plane, that captures the image formed by the prism, the mirror under test, and the focusing lens.
The image on the focal plane consists of three components: Component 1 is nominally a point image formed by focusing of that part of the incident parallel light beam that passes directly through the lens. Component 2 is nominally a circular image formed by reflection of the incident light beam from the first conical surface of the prism, followed by reflection from the mirror under test, followed by reflection from the second conical surface of the prism, followed by focusing via the lens.
Component 3 nominally consists of one or more point image(s) formed by passage of part of the incident parallel beam through the central part of the prism and then through the lens. The precise nature of component 3 depends on the specific design of the entrance and exit faces of the central part of the prism: If, for example, the prism is made of optical glass with flat entrance and exit surfaces, then depending on the precision of parallelarity between these surfaces, the resulting single point image may or may not coincide with the point image of beam 1. If, for another example, the prism is made of a metal, then the central core of the prism can be drilled out and a glass refractor with, say, three facets can be inserted to refract the incident parallel beam into three detected beams that, in turn, are focused to three point images on the screen (see Figure 2). These point images, in conjunction with the component-1 point image, can be used as fiducial marks for monitoring the orientation of the prism with respect to the incident parallel beam.
Among the most important features of this system are that the parallel light beam is the sole reference for the entire setup and the system inherently provides fiducial marks for its own alignment: The parallel light beam defines the direction of the optical axis. Once the axis of symmetry of the conical prism has been made parallel to the incident light beam and its position fixed, then the optical axis of the entire setup is fixed. The positions of the three component-3 point images with respect to the single component-1 point image serve as real-time indications of the alignment of the axis of symmetry of the prism.
When the optical axes of the conical prism and the mirror under test and the conical prism are perfectly aligned with each other, and provided that the mirror under test is perfect, then the image on the focal plane is as shown in the upper part of Figure 2. If these two optical axes do not coincide and/or if the mirror under test is not perfect, then the image becomes distorted, as shown by example in the lower part of Figure 2.
This work was done by William W. Zhang of Goddard Space Flight Center.