A mount has been devised to satisfy a requirement to suspend a highly precise, flat, circular, low-thermal-expansion glass mirror in a horizontal plane with its reflective side down while keeping the reflective mirror surface flat to within a peak-to-valley depth of less 50 nm. The difficulty of the suspension problem and the significance of the mount conceived as the solution of the problem arise from the large size and weight of the mirror (diameter 101.7 cm, thickness 18.8 cm, mass 385 kg).

Figure 1. The Nine-Point Hindle Mount includes six supports on a circle with a radius of 0.408D and three supports on a circle with a radius of 0.204D,where D is the outer diameter of the mirror. The supports in each group of three are connected to a triangular plate supported at its centroid, which lies on a circle with radius 0.309D. When the three triangular plates are picked up at their centroids, equal weight is borne by all nine supports.

The mount is an inverted version of the Hindle mount, which was first published in 1945 and is named after its inventor (Dr. J. H. Hindle). The Hindle mount supports a flat, circular mirror on multiple pads that are located so that all of the pads bear the same proportion of the total weight and the sag of the mirror under its own weight between the supports is minimized. The number of support pads in a Hindle mount is an integer multiple of three. The sag can be reduced by increasing the number of support pads. The present mount is based on a Hindle-mount geometry of nine support points grouped into three triangles (see Figure 1).

Figure 2. Parts of the Inverted Hindle Mount are attached to an aluminum disk of the same size as (and slightly heavier than) the glass mirror that the mount is intended to support.(The aluminum disk was used during development so as not to risk damage to the expensive glass mirror.)

The original Hindle-mount concept applies to a mirror suspended from its lower face with its reflective side up; the weight of the mirror presses it against the support pads, and the resulting friction often suffices to hold the mirror in place on the pads. In the present inverted Hindle mount, the mirror is suspended from its upper face; consequently, the weight of the mirror tends to pull it away from its supports, making it necessary to fasten the mirror to the supports in some fashion.

Figure 2 depicts some parts of the present inverted Hindle mount. The nine supports are made of Invar — a low-thermal-expansion Fe/Ni alloy. To distinguish them from the compression-loaded support pads of a traditional Hindle mount, these tensile*loaded supports are called “anchors.” The lower end of each anchor has an area of 3 in.2 (≈19.4 cm2) and is bonded to the face of the mirror by use of 3M’s 2216 (or equivalent) clear high-performance epoxy. The upper end of each anchor is tapped to accept a bolt. The upper ends of the anchors in each group of three are bolted loosely, with spherical washers, to a triangular aluminum plate 1 in. (2.54 cm) thick. The loose bolting and spherical washers allow some limited rotation and translation, thereby helping to prevent the buildup of bending moments that could cause the anchors to peel away from the mirror surface at the epoxy joints.

Each plate is supported from above by means of a clevis/ball-joint coupling attached to the plate at the centroid of the triangle defined by the three support points. The clevis is, in turn, supported from above by a threaded rod that passes through a massive overhead frame and engages a nut on top of the frame. The freedom of rotation of the clevis/ball joints contributes to the relief of any bending moments that could, if allowed to build up, cause peeling at the epoxy joints.

Experience in the development of the inverted Hindle mount revealed that it can be difficult to form reliable epoxy bonds between the anchors and the top surface of the mirror. Therefore, the mount must be augmented by handling fixtures with safety features to prevent the mirror from falling in the event of failure of one or more epoxy bonds.

This work was done by David W. Robinson of Goddard Space Flight Center. GSC-14316